id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.7k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-04-25 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2107.01807 | Rachmad Vidya Wicaksana Putra | Rachmad Vidya Wicaksana Putra, Muhammad Shafique | Q-SpiNN: A Framework for Quantizing Spiking Neural Networks | Accepted for publication at the 2021 International Joint Conference
on Neural Networks (IJCNN), July 2021, Virtual Event | null | 10.1109/IJCNN52387.2021.9534087 | null | cs.NE cs.AR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A prominent technique for reducing the memory footprint of Spiking Neural
Networks (SNNs) without decreasing the accuracy significantly is quantization.
However, the state-of-the-art only focus on employing the weight quantization
directly from a specific quantization scheme, i.e., either the post-training
quantization (PTQ) or the in-training quantization (ITQ), and do not consider
(1) quantizing other SNN parameters (e.g., neuron membrane potential), (2)
exploring different combinations of quantization approaches (i.e., quantization
schemes, precision levels, and rounding schemes), and (3) selecting the SNN
model with a good memory-accuracy trade-off at the end. Therefore, the memory
saving offered by these state-of-the-art to meet the targeted accuracy is
limited, thereby hindering processing SNNs on the resource-constrained systems
(e.g., the IoT-Edge devices). Towards this, we propose Q-SpiNN, a novel
quantization framework for memory-efficient SNNs. The key mechanisms of the
Q-SpiNN are: (1) employing quantization for different SNN parameters based on
their significance to the accuracy, (2) exploring different combinations of
quantization schemes, precision levels, and rounding schemes to find efficient
SNN model candidates, and (3) developing an algorithm that quantifies the
benefit of the memory-accuracy trade-off obtained by the candidates, and
selects the Pareto-optimal one. The experimental results show that, for the
unsupervised network, the Q-SpiNN reduces the memory footprint by ca. 4x, while
maintaining the accuracy within 1% from the baseline on the MNIST dataset. For
the supervised network, the Q-SpiNN reduces the memory by ca. 2x, while keeping
the accuracy within 2% from the baseline on the DVS-Gesture dataset.
| [
{
"version": "v1",
"created": "Mon, 5 Jul 2021 06:01:15 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Putra",
"Rachmad Vidya Wicaksana",
""
],
[
"Shafique",
"Muhammad",
""
]
] | TITLE: Q-SpiNN: A Framework for Quantizing Spiking Neural Networks
ABSTRACT: A prominent technique for reducing the memory footprint of Spiking Neural
Networks (SNNs) without decreasing the accuracy significantly is quantization.
However, the state-of-the-art only focus on employing the weight quantization
directly from a specific quantization scheme, i.e., either the post-training
quantization (PTQ) or the in-training quantization (ITQ), and do not consider
(1) quantizing other SNN parameters (e.g., neuron membrane potential), (2)
exploring different combinations of quantization approaches (i.e., quantization
schemes, precision levels, and rounding schemes), and (3) selecting the SNN
model with a good memory-accuracy trade-off at the end. Therefore, the memory
saving offered by these state-of-the-art to meet the targeted accuracy is
limited, thereby hindering processing SNNs on the resource-constrained systems
(e.g., the IoT-Edge devices). Towards this, we propose Q-SpiNN, a novel
quantization framework for memory-efficient SNNs. The key mechanisms of the
Q-SpiNN are: (1) employing quantization for different SNN parameters based on
their significance to the accuracy, (2) exploring different combinations of
quantization schemes, precision levels, and rounding schemes to find efficient
SNN model candidates, and (3) developing an algorithm that quantifies the
benefit of the memory-accuracy trade-off obtained by the candidates, and
selects the Pareto-optimal one. The experimental results show that, for the
unsupervised network, the Q-SpiNN reduces the memory footprint by ca. 4x, while
maintaining the accuracy within 1% from the baseline on the MNIST dataset. For
the supervised network, the Q-SpiNN reduces the memory by ca. 2x, while keeping
the accuracy within 2% from the baseline on the DVS-Gesture dataset.
|
2108.10015 | Xinghao Yang | Xinghao Yang, Weifeng Liu, James Bailey, Dacheng Tao, Wei Liu | Semantic-Preserving Adversarial Text Attacks | 12 pages, 3 figures, 10 tables | null | null | null | cs.CL stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep neural networks (DNNs) are known to be vulnerable to adversarial images,
while their robustness in text classification is rarely studied. Several lines
of text attack methods have been proposed in the literature, including
character-level, word-level, and sentence-level attacks. However, it is still a
challenge to minimize the number of word changes necessary to induce
misclassification, while simultaneously ensuring lexical correctness, syntactic
soundness, and semantic similarity. In this paper, we propose a Bigram and
Unigram based adaptive Semantic Preservation Optimization (BU-SPO) method to
examine the vulnerability of deep models. Our method has four major merits.
Firstly, we propose to attack text documents not only at the unigram word level
but also at the bigram level which better keeps semantics and avoids producing
meaningless outputs. Secondly, we propose a hybrid method to replace the input
words with options among both their synonyms candidates and sememe candidates,
which greatly enriches the potential substitutions compared to only using
synonyms. Thirdly, we design an optimization algorithm, i.e., Semantic
Preservation Optimization (SPO), to determine the priority of word
replacements, aiming to reduce the modification cost. Finally, we further
improve the SPO with a semantic Filter (named SPOF) to find the adversarial
example with the highest semantic similarity. We evaluate the effectiveness of
our BU-SPO and BU-SPOF on IMDB, AG's News, and Yahoo! Answers text datasets by
attacking four popular DNNs models. Results show that our methods achieve the
highest attack success rates and semantics rates by changing the smallest
number of words compared with existing methods.
| [
{
"version": "v1",
"created": "Mon, 23 Aug 2021 09:05:18 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 01:58:16 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Yang",
"Xinghao",
""
],
[
"Liu",
"Weifeng",
""
],
[
"Bailey",
"James",
""
],
[
"Tao",
"Dacheng",
""
],
[
"Liu",
"Wei",
""
]
] | TITLE: Semantic-Preserving Adversarial Text Attacks
ABSTRACT: Deep neural networks (DNNs) are known to be vulnerable to adversarial images,
while their robustness in text classification is rarely studied. Several lines
of text attack methods have been proposed in the literature, including
character-level, word-level, and sentence-level attacks. However, it is still a
challenge to minimize the number of word changes necessary to induce
misclassification, while simultaneously ensuring lexical correctness, syntactic
soundness, and semantic similarity. In this paper, we propose a Bigram and
Unigram based adaptive Semantic Preservation Optimization (BU-SPO) method to
examine the vulnerability of deep models. Our method has four major merits.
Firstly, we propose to attack text documents not only at the unigram word level
but also at the bigram level which better keeps semantics and avoids producing
meaningless outputs. Secondly, we propose a hybrid method to replace the input
words with options among both their synonyms candidates and sememe candidates,
which greatly enriches the potential substitutions compared to only using
synonyms. Thirdly, we design an optimization algorithm, i.e., Semantic
Preservation Optimization (SPO), to determine the priority of word
replacements, aiming to reduce the modification cost. Finally, we further
improve the SPO with a semantic Filter (named SPOF) to find the adversarial
example with the highest semantic similarity. We evaluate the effectiveness of
our BU-SPO and BU-SPOF on IMDB, AG's News, and Yahoo! Answers text datasets by
attacking four popular DNNs models. Results show that our methods achieve the
highest attack success rates and semantics rates by changing the smallest
number of words compared with existing methods.
|
2108.10271 | Rachmad Vidya Wicaksana Putra | Rachmad Vidya Wicaksana Putra, Muhammad Abdullah Hanif, Muhammad
Shafique | ReSpawn: Energy-Efficient Fault-Tolerance for Spiking Neural Networks
considering Unreliable Memories | To appear at the 40th IEEE/ACM International Conference on
Computer-Aided Design (ICCAD), November 2021, Virtual Event | null | 10.1109/ICCAD51958.2021.9643524 | null | cs.AR cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spiking neural networks (SNNs) have shown a potential for having low energy
with unsupervised learning capabilities due to their biologically-inspired
computation. However, they may suffer from accuracy degradation if their
processing is performed under the presence of hardware-induced faults in
memories, which can come from manufacturing defects or voltage-induced
approximation errors. Since recent works still focus on the fault-modeling and
random fault injection in SNNs, the impact of memory faults in SNN hardware
architectures on accuracy and the respective fault-mitigation techniques are
not thoroughly explored. Toward this, we propose ReSpawn, a novel framework for
mitigating the negative impacts of faults in both the off-chip and on-chip
memories for resilient and energy-efficient SNNs. The key mechanisms of ReSpawn
are: (1) analyzing the fault tolerance of SNNs; and (2) improving the SNN fault
tolerance through (a) fault-aware mapping (FAM) in memories, and (b)
fault-aware training-and-mapping (FATM). If the training dataset is not fully
available, FAM is employed through efficient bit-shuffling techniques that
place the significant bits on the non-faulty memory cells and the insignificant
bits on the faulty ones, while minimizing the memory access energy. Meanwhile,
if the training dataset is fully available, FATM is employed by considering the
faulty memory cells in the data mapping and training processes. The
experimental results show that, compared to the baseline SNN without
fault-mitigation techniques, ReSpawn with a fault-aware mapping scheme improves
the accuracy by up to 70% for a network with 900 neurons without retraining.
| [
{
"version": "v1",
"created": "Mon, 23 Aug 2021 16:17:33 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Putra",
"Rachmad Vidya Wicaksana",
""
],
[
"Hanif",
"Muhammad Abdullah",
""
],
[
"Shafique",
"Muhammad",
""
]
] | TITLE: ReSpawn: Energy-Efficient Fault-Tolerance for Spiking Neural Networks
considering Unreliable Memories
ABSTRACT: Spiking neural networks (SNNs) have shown a potential for having low energy
with unsupervised learning capabilities due to their biologically-inspired
computation. However, they may suffer from accuracy degradation if their
processing is performed under the presence of hardware-induced faults in
memories, which can come from manufacturing defects or voltage-induced
approximation errors. Since recent works still focus on the fault-modeling and
random fault injection in SNNs, the impact of memory faults in SNN hardware
architectures on accuracy and the respective fault-mitigation techniques are
not thoroughly explored. Toward this, we propose ReSpawn, a novel framework for
mitigating the negative impacts of faults in both the off-chip and on-chip
memories for resilient and energy-efficient SNNs. The key mechanisms of ReSpawn
are: (1) analyzing the fault tolerance of SNNs; and (2) improving the SNN fault
tolerance through (a) fault-aware mapping (FAM) in memories, and (b)
fault-aware training-and-mapping (FATM). If the training dataset is not fully
available, FAM is employed through efficient bit-shuffling techniques that
place the significant bits on the non-faulty memory cells and the insignificant
bits on the faulty ones, while minimizing the memory access energy. Meanwhile,
if the training dataset is fully available, FATM is employed by considering the
faulty memory cells in the data mapping and training processes. The
experimental results show that, compared to the baseline SNN without
fault-mitigation techniques, ReSpawn with a fault-aware mapping scheme improves
the accuracy by up to 70% for a network with 900 neurons without retraining.
|
2202.04202 | Paul Bertin | Paul Bertin, Jarrid Rector-Brooks, Deepak Sharma, Thomas Gaudelet,
Andrew Anighoro, Torsten Gross, Francisco Martinez-Pena, Eileen L. Tang,
Suraj M S, Cristian Regep, Jeremy Hayter, Maksym Korablyov, Nicholas
Valiante, Almer van der Sloot, Mike Tyers, Charles Roberts, Michael M.
Bronstein, Luke L. Lairson, Jake P. Taylor-King, and Yoshua Bengio | RECOVER: sequential model optimization platform for combination drug
repurposing identifies novel synergistic compounds in vitro | null | null | null | null | q-bio.QM cs.LG | http://creativecommons.org/licenses/by/4.0/ | For large libraries of small molecules, exhaustive combinatorial chemical
screens become infeasible to perform when considering a range of disease
models, assay conditions, and dose ranges. Deep learning models have achieved
state of the art results in silico for the prediction of synergy scores.
However, databases of drug combinations are biased towards synergistic agents
and these results do not necessarily generalise out of distribution. We employ
a sequential model optimization search utilising a deep learning model to
quickly discover synergistic drug combinations active against a cancer cell
line, requiring substantially less screening than an exhaustive evaluation. Our
small scale wet lab experiments only account for evaluation of ~5% of the total
search space. After only 3 rounds of ML-guided in vitro experimentation
(including a calibration round), we find that the set of drug pairs queried is
enriched for highly synergistic combinations; two additional rounds of
ML-guided experiments were performed to ensure reproducibility of trends.
Remarkably, we rediscover drug combinations later confirmed to be under study
within clinical trials. Moreover, we find that drug embeddings generated using
only structural information begin to reflect mechanisms of action. Prior in
silico benchmarking suggests we can enrich search queries by a factor of ~5-10x
for highly synergistic drug combinations by using sequential rounds of
evaluation when compared to random selection, or by a factor of >3x when using
a pretrained model selecting all drug combinations at a single time point.
| [
{
"version": "v1",
"created": "Mon, 7 Feb 2022 02:54:29 GMT"
},
{
"version": "v2",
"created": "Sun, 4 Sep 2022 21:19:05 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 21:58:51 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Bertin",
"Paul",
""
],
[
"Rector-Brooks",
"Jarrid",
""
],
[
"Sharma",
"Deepak",
""
],
[
"Gaudelet",
"Thomas",
""
],
[
"Anighoro",
"Andrew",
""
],
[
"Gross",
"Torsten",
""
],
[
"Martinez-Pena",
"Francisco",
""
],
[
"Tang",
"Eileen L.",
""
],
[
"S",
"Suraj M",
""
],
[
"Regep",
"Cristian",
""
],
[
"Hayter",
"Jeremy",
""
],
[
"Korablyov",
"Maksym",
""
],
[
"Valiante",
"Nicholas",
""
],
[
"van der Sloot",
"Almer",
""
],
[
"Tyers",
"Mike",
""
],
[
"Roberts",
"Charles",
""
],
[
"Bronstein",
"Michael M.",
""
],
[
"Lairson",
"Luke L.",
""
],
[
"Taylor-King",
"Jake P.",
""
],
[
"Bengio",
"Yoshua",
""
]
] | TITLE: RECOVER: sequential model optimization platform for combination drug
repurposing identifies novel synergistic compounds in vitro
ABSTRACT: For large libraries of small molecules, exhaustive combinatorial chemical
screens become infeasible to perform when considering a range of disease
models, assay conditions, and dose ranges. Deep learning models have achieved
state of the art results in silico for the prediction of synergy scores.
However, databases of drug combinations are biased towards synergistic agents
and these results do not necessarily generalise out of distribution. We employ
a sequential model optimization search utilising a deep learning model to
quickly discover synergistic drug combinations active against a cancer cell
line, requiring substantially less screening than an exhaustive evaluation. Our
small scale wet lab experiments only account for evaluation of ~5% of the total
search space. After only 3 rounds of ML-guided in vitro experimentation
(including a calibration round), we find that the set of drug pairs queried is
enriched for highly synergistic combinations; two additional rounds of
ML-guided experiments were performed to ensure reproducibility of trends.
Remarkably, we rediscover drug combinations later confirmed to be under study
within clinical trials. Moreover, we find that drug embeddings generated using
only structural information begin to reflect mechanisms of action. Prior in
silico benchmarking suggests we can enrich search queries by a factor of ~5-10x
for highly synergistic drug combinations by using sequential rounds of
evaluation when compared to random selection, or by a factor of >3x when using
a pretrained model selecting all drug combinations at a single time point.
|
2203.03814 | Han Joo Chae | Han Joo Chae, Seunghwan Lee, Hyewon Son, Seungyeob Han, Taebin Lim | Generating 3D Bio-Printable Patches Using Wound Segmentation and
Reconstruction to Treat Diabetic Foot Ulcers | Accepted to CVPR 2022 | null | 10.1109/CVPR52688.2022.00257 | null | eess.IV cs.CV cs.HC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce AiD Regen, a novel system that generates 3D wound models
combining 2D semantic segmentation with 3D reconstruction so that they can be
printed via 3D bio-printers during the surgery to treat diabetic foot ulcers
(DFUs). AiD Regen seamlessly binds the full pipeline, which includes RGB-D
image capturing, semantic segmentation, boundary-guided point-cloud processing,
3D model reconstruction, and 3D printable G-code generation, into a single
system that can be used out of the box. We developed a multi-stage data
preprocessing method to handle small and unbalanced DFU image datasets. AiD
Regen's human-in-the-loop machine learning interface enables clinicians to not
only create 3D regenerative patches with just a few touch interactions but also
customize and confirm wound boundaries. As evidenced by our experiments, our
model outperforms prior wound segmentation models and our reconstruction
algorithm is capable of generating 3D wound models with compelling accuracy. We
further conducted a case study on a real DFU patient and demonstrated the
effectiveness of AiD Regen in treating DFU wounds.
| [
{
"version": "v1",
"created": "Tue, 8 Mar 2022 02:29:32 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chae",
"Han Joo",
""
],
[
"Lee",
"Seunghwan",
""
],
[
"Son",
"Hyewon",
""
],
[
"Han",
"Seungyeob",
""
],
[
"Lim",
"Taebin",
""
]
] | TITLE: Generating 3D Bio-Printable Patches Using Wound Segmentation and
Reconstruction to Treat Diabetic Foot Ulcers
ABSTRACT: We introduce AiD Regen, a novel system that generates 3D wound models
combining 2D semantic segmentation with 3D reconstruction so that they can be
printed via 3D bio-printers during the surgery to treat diabetic foot ulcers
(DFUs). AiD Regen seamlessly binds the full pipeline, which includes RGB-D
image capturing, semantic segmentation, boundary-guided point-cloud processing,
3D model reconstruction, and 3D printable G-code generation, into a single
system that can be used out of the box. We developed a multi-stage data
preprocessing method to handle small and unbalanced DFU image datasets. AiD
Regen's human-in-the-loop machine learning interface enables clinicians to not
only create 3D regenerative patches with just a few touch interactions but also
customize and confirm wound boundaries. As evidenced by our experiments, our
model outperforms prior wound segmentation models and our reconstruction
algorithm is capable of generating 3D wound models with compelling accuracy. We
further conducted a case study on a real DFU patient and demonstrated the
effectiveness of AiD Regen in treating DFU wounds.
|
2203.04571 | Michael Hersche | Michael Hersche, Mustafa Zeqiri, Luca Benini, Abu Sebastian, Abbas
Rahimi | A Neuro-vector-symbolic Architecture for Solving Raven's Progressive
Matrices | Updated version with additional NVSA end-to-end training,
generalization experiments, and PGM experiments | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | Neither deep neural networks nor symbolic AI alone has approached the kind of
intelligence expressed in humans. This is mainly because neural networks are
not able to decompose joint representations to obtain distinct objects (the
so-called binding problem), while symbolic AI suffers from exhaustive rule
searches, among other problems. These two problems are still pronounced in
neuro-symbolic AI which aims to combine the best of the two paradigms. Here, we
show that the two problems can be addressed with our proposed
neuro-vector-symbolic architecture (NVSA) by exploiting its powerful operators
on high-dimensional distributed representations that serve as a common language
between neural networks and symbolic AI. The efficacy of NVSA is demonstrated
by solving the Raven's progressive matrices datasets. Compared to
state-of-the-art deep neural network and neuro-symbolic approaches, end-to-end
training of NVSA achieves a new record of 87.7% average accuracy in RAVEN, and
88.1% in I-RAVEN datasets. Moreover, compared to the symbolic reasoning within
the neuro-symbolic approaches, the probabilistic reasoning of NVSA with less
expensive operations on the distributed representations is two orders of
magnitude faster. Our code is available at
https://github.com/IBM/neuro-vector-symbolic-architectures.
| [
{
"version": "v1",
"created": "Wed, 9 Mar 2022 08:29:21 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 14:09:50 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Hersche",
"Michael",
""
],
[
"Zeqiri",
"Mustafa",
""
],
[
"Benini",
"Luca",
""
],
[
"Sebastian",
"Abu",
""
],
[
"Rahimi",
"Abbas",
""
]
] | TITLE: A Neuro-vector-symbolic Architecture for Solving Raven's Progressive
Matrices
ABSTRACT: Neither deep neural networks nor symbolic AI alone has approached the kind of
intelligence expressed in humans. This is mainly because neural networks are
not able to decompose joint representations to obtain distinct objects (the
so-called binding problem), while symbolic AI suffers from exhaustive rule
searches, among other problems. These two problems are still pronounced in
neuro-symbolic AI which aims to combine the best of the two paradigms. Here, we
show that the two problems can be addressed with our proposed
neuro-vector-symbolic architecture (NVSA) by exploiting its powerful operators
on high-dimensional distributed representations that serve as a common language
between neural networks and symbolic AI. The efficacy of NVSA is demonstrated
by solving the Raven's progressive matrices datasets. Compared to
state-of-the-art deep neural network and neuro-symbolic approaches, end-to-end
training of NVSA achieves a new record of 87.7% average accuracy in RAVEN, and
88.1% in I-RAVEN datasets. Moreover, compared to the symbolic reasoning within
the neuro-symbolic approaches, the probabilistic reasoning of NVSA with less
expensive operations on the distributed representations is two orders of
magnitude faster. Our code is available at
https://github.com/IBM/neuro-vector-symbolic-architectures.
|
2203.14402 | Yue Chen | Yue Chen, Xuan Wang, Xingyu Chen, Qi Zhang, Xiaoyu Li, Yu Guo, Jue
Wang, Fei Wang | UV Volumes for Real-time Rendering of Editable Free-view Human
Performance | Accepted to CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural volume rendering enables photo-realistic renderings of a human
performer in free-view, a critical task in immersive VR/AR applications. But
the practice is severely limited by high computational costs in the rendering
process. To solve this problem, we propose the UV Volumes, a new approach that
can render an editable free-view video of a human performer in real-time. It
separates the high-frequency (i.e., non-smooth) human appearance from the 3D
volume, and encodes them into 2D neural texture stacks (NTS). The smooth UV
volumes allow much smaller and shallower neural networks to obtain densities
and texture coordinates in 3D while capturing detailed appearance in 2D NTS.
For editability, the mapping between the parameterized human model and the
smooth texture coordinates allows us a better generalization on novel poses and
shapes. Furthermore, the use of NTS enables interesting applications, e.g.,
retexturing. Extensive experiments on CMU Panoptic, ZJU Mocap, and H36M
datasets show that our model can render 960 x 540 images in 30FPS on average
with comparable photo-realism to state-of-the-art methods. The project and
supplementary materials are available at https://fanegg.github.io/UV-Volumes.
| [
{
"version": "v1",
"created": "Sun, 27 Mar 2022 21:54:36 GMT"
},
{
"version": "v2",
"created": "Fri, 27 May 2022 08:04:43 GMT"
},
{
"version": "v3",
"created": "Fri, 17 Jun 2022 11:42:17 GMT"
},
{
"version": "v4",
"created": "Fri, 11 Nov 2022 08:43:24 GMT"
},
{
"version": "v5",
"created": "Fri, 3 Mar 2023 14:20:23 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chen",
"Yue",
""
],
[
"Wang",
"Xuan",
""
],
[
"Chen",
"Xingyu",
""
],
[
"Zhang",
"Qi",
""
],
[
"Li",
"Xiaoyu",
""
],
[
"Guo",
"Yu",
""
],
[
"Wang",
"Jue",
""
],
[
"Wang",
"Fei",
""
]
] | TITLE: UV Volumes for Real-time Rendering of Editable Free-view Human
Performance
ABSTRACT: Neural volume rendering enables photo-realistic renderings of a human
performer in free-view, a critical task in immersive VR/AR applications. But
the practice is severely limited by high computational costs in the rendering
process. To solve this problem, we propose the UV Volumes, a new approach that
can render an editable free-view video of a human performer in real-time. It
separates the high-frequency (i.e., non-smooth) human appearance from the 3D
volume, and encodes them into 2D neural texture stacks (NTS). The smooth UV
volumes allow much smaller and shallower neural networks to obtain densities
and texture coordinates in 3D while capturing detailed appearance in 2D NTS.
For editability, the mapping between the parameterized human model and the
smooth texture coordinates allows us a better generalization on novel poses and
shapes. Furthermore, the use of NTS enables interesting applications, e.g.,
retexturing. Extensive experiments on CMU Panoptic, ZJU Mocap, and H36M
datasets show that our model can render 960 x 540 images in 30FPS on average
with comparable photo-realism to state-of-the-art methods. The project and
supplementary materials are available at https://fanegg.github.io/UV-Volumes.
|
2203.15952 | Shaojin Ding | Shaojin Ding, Phoenix Meadowlark, Yanzhang He, Lukasz Lew, Shivani
Agrawal, Oleg Rybakov | 4-bit Conformer with Native Quantization Aware Training for Speech
Recognition | Published at INTERSPEECH 2022 | null | null | null | eess.AS cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reducing the latency and model size has always been a significant research
problem for live Automatic Speech Recognition (ASR) application scenarios.
Along this direction, model quantization has become an increasingly popular
approach to compress neural networks and reduce computation cost. Most of the
existing practical ASR systems apply post-training 8-bit quantization. To
achieve a higher compression rate without introducing additional performance
regression, in this study, we propose to develop 4-bit ASR models with native
quantization aware training, which leverages native integer operations to
effectively optimize both training and inference. We conducted two experiments
on state-of-the-art Conformer-based ASR models to evaluate our proposed
quantization technique. First, we explored the impact of different precisions
for both weight and activation quantization on the LibriSpeech dataset, and
obtained a lossless 4-bit Conformer model with 5.8x size reduction compared to
the float32 model. Following this, we for the first time investigated and
revealed the viability of 4-bit quantization on a practical ASR system that is
trained with large-scale datasets, and produced a lossless Conformer ASR model
with mixed 4-bit and 8-bit weights that has 5x size reduction compared to the
float32 model.
| [
{
"version": "v1",
"created": "Tue, 29 Mar 2022 23:57:15 GMT"
},
{
"version": "v2",
"created": "Fri, 24 Jun 2022 22:17:06 GMT"
},
{
"version": "v3",
"created": "Tue, 25 Oct 2022 22:19:41 GMT"
},
{
"version": "v4",
"created": "Fri, 3 Mar 2023 01:52:28 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Ding",
"Shaojin",
""
],
[
"Meadowlark",
"Phoenix",
""
],
[
"He",
"Yanzhang",
""
],
[
"Lew",
"Lukasz",
""
],
[
"Agrawal",
"Shivani",
""
],
[
"Rybakov",
"Oleg",
""
]
] | TITLE: 4-bit Conformer with Native Quantization Aware Training for Speech
Recognition
ABSTRACT: Reducing the latency and model size has always been a significant research
problem for live Automatic Speech Recognition (ASR) application scenarios.
Along this direction, model quantization has become an increasingly popular
approach to compress neural networks and reduce computation cost. Most of the
existing practical ASR systems apply post-training 8-bit quantization. To
achieve a higher compression rate without introducing additional performance
regression, in this study, we propose to develop 4-bit ASR models with native
quantization aware training, which leverages native integer operations to
effectively optimize both training and inference. We conducted two experiments
on state-of-the-art Conformer-based ASR models to evaluate our proposed
quantization technique. First, we explored the impact of different precisions
for both weight and activation quantization on the LibriSpeech dataset, and
obtained a lossless 4-bit Conformer model with 5.8x size reduction compared to
the float32 model. Following this, we for the first time investigated and
revealed the viability of 4-bit quantization on a practical ASR system that is
trained with large-scale datasets, and produced a lossless Conformer ASR model
with mixed 4-bit and 8-bit weights that has 5x size reduction compared to the
float32 model.
|
2203.16172 | Yiqing Shen | Yiqing Shen, Liwu Xu, Yuzhe Yang, Yaqian Li, Yandong Guo | Self-Distillation from the Last Mini-Batch for Consistency
Regularization | 10 pages | Proceedings of the IEEE/CVF Conference on Computer Vision and
Pattern Recognition (CVPR), 2022, pp. 11943-11952 | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Knowledge distillation (KD) shows a bright promise as a powerful
regularization strategy to boost generalization ability by leveraging learned
sample-level soft targets. Yet, employing a complex pre-trained teacher network
or an ensemble of peer students in existing KD is both time-consuming and
computationally costly. Various self KD methods have been proposed to achieve
higher distillation efficiency. However, they either require extra network
architecture modification or are difficult to parallelize. To cope with these
challenges, we propose an efficient and reliable self-distillation framework,
named Self-Distillation from Last Mini-Batch (DLB). Specifically, we rearrange
the sequential sampling by constraining half of each mini-batch coinciding with
the previous iteration. Meanwhile, the rest half will coincide with the
upcoming iteration. Afterwards, the former half mini-batch distills on-the-fly
soft targets generated in the previous iteration. Our proposed mechanism guides
the training stability and consistency, resulting in robustness to label noise.
Moreover, our method is easy to implement, without taking up extra run-time
memory or requiring model structure modification. Experimental results on three
classification benchmarks illustrate that our approach can consistently
outperform state-of-the-art self-distillation approaches with different network
architectures. Additionally, our method shows strong compatibility with
augmentation strategies by gaining additional performance improvement. The code
is available at https://github.com/Meta-knowledge-Lab/DLB.
| [
{
"version": "v1",
"created": "Wed, 30 Mar 2022 09:50:24 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Shen",
"Yiqing",
""
],
[
"Xu",
"Liwu",
""
],
[
"Yang",
"Yuzhe",
""
],
[
"Li",
"Yaqian",
""
],
[
"Guo",
"Yandong",
""
]
] | TITLE: Self-Distillation from the Last Mini-Batch for Consistency
Regularization
ABSTRACT: Knowledge distillation (KD) shows a bright promise as a powerful
regularization strategy to boost generalization ability by leveraging learned
sample-level soft targets. Yet, employing a complex pre-trained teacher network
or an ensemble of peer students in existing KD is both time-consuming and
computationally costly. Various self KD methods have been proposed to achieve
higher distillation efficiency. However, they either require extra network
architecture modification or are difficult to parallelize. To cope with these
challenges, we propose an efficient and reliable self-distillation framework,
named Self-Distillation from Last Mini-Batch (DLB). Specifically, we rearrange
the sequential sampling by constraining half of each mini-batch coinciding with
the previous iteration. Meanwhile, the rest half will coincide with the
upcoming iteration. Afterwards, the former half mini-batch distills on-the-fly
soft targets generated in the previous iteration. Our proposed mechanism guides
the training stability and consistency, resulting in robustness to label noise.
Moreover, our method is easy to implement, without taking up extra run-time
memory or requiring model structure modification. Experimental results on three
classification benchmarks illustrate that our approach can consistently
outperform state-of-the-art self-distillation approaches with different network
architectures. Additionally, our method shows strong compatibility with
augmentation strategies by gaining additional performance improvement. The code
is available at https://github.com/Meta-knowledge-Lab/DLB.
|
2204.01868 | Leandro Marcomini | Leandro Arab Marcomini, Andr\'e Luiz Cunha | Truck Axle Detection with Convolutional Neural Networks | Code and dataset available for donwload, links provided | null | null | null | cs.CV cs.NE | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Axle count in trucks is important to the classification of vehicles and to
the operation of road systems. It is used in the determination of service fees
and in the impact on the pavement. Although axle count can be achieved with
traditional methods, such as manual labor, it is increasingly possible to count
axles using deep learning and computer vision methods. This paper aims to
compare three deep-learning object detection algorithms, YOLO, Faster R-CNN,
and SSD, for the detection of truck axles. A dataset was built to provide
training and testing examples for the neural networks. The training was done on
different base models, to increase training time efficiency and to compare
results. We evaluated results based on five metrics: precision, recall, mAP,
F1-score, and FPS count. Results indicate that YOLO and SSD have similar
accuracy and performance, with more than 96\% mAP for both models. Datasets and
codes are publicly available for download.
| [
{
"version": "v1",
"created": "Mon, 4 Apr 2022 22:11:49 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 12:41:10 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Marcomini",
"Leandro Arab",
""
],
[
"Cunha",
"André Luiz",
""
]
] | TITLE: Truck Axle Detection with Convolutional Neural Networks
ABSTRACT: Axle count in trucks is important to the classification of vehicles and to
the operation of road systems. It is used in the determination of service fees
and in the impact on the pavement. Although axle count can be achieved with
traditional methods, such as manual labor, it is increasingly possible to count
axles using deep learning and computer vision methods. This paper aims to
compare three deep-learning object detection algorithms, YOLO, Faster R-CNN,
and SSD, for the detection of truck axles. A dataset was built to provide
training and testing examples for the neural networks. The training was done on
different base models, to increase training time efficiency and to compare
results. We evaluated results based on five metrics: precision, recall, mAP,
F1-score, and FPS count. Results indicate that YOLO and SSD have similar
accuracy and performance, with more than 96\% mAP for both models. Datasets and
codes are publicly available for download.
|
2204.02385 | Danilo Comminiello | Eric Guizzo, Tillman Weyde, Simone Scardapane, Danilo Comminiello | Learning Speech Emotion Representations in the Quaternion Domain | Accepted for Publication in IEEE/ACM Transactions on Audio, Speech
and Language Processing | null | 10.1109/TASLP.2023.3250840 | null | eess.AS cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The modeling of human emotion expression in speech signals is an important,
yet challenging task. The high resource demand of speech emotion recognition
models, combined with the the general scarcity of emotion-labelled data are
obstacles to the development and application of effective solutions in this
field. In this paper, we present an approach to jointly circumvent these
difficulties. Our method, named RH-emo, is a novel semi-supervised architecture
aimed at extracting quaternion embeddings from real-valued monoaural
spectrograms, enabling the use of quaternion-valued networks for speech emotion
recognition tasks. RH-emo is a hybrid real/quaternion autoencoder network that
consists of a real-valued encoder in parallel to a real-valued emotion
classifier and a quaternion-valued decoder. On the one hand, the classifier
permits to optimize each latent axis of the embeddings for the classification
of a specific emotion-related characteristic: valence, arousal, dominance and
overall emotion. On the other hand, the quaternion reconstruction enables the
latent dimension to develop intra-channel correlations that are required for an
effective representation as a quaternion entity. We test our approach on speech
emotion recognition tasks using four popular datasets: Iemocap, Ravdess, EmoDb
and Tess, comparing the performance of three well-established real-valued CNN
architectures (AlexNet, ResNet-50, VGG) and their quaternion-valued equivalent
fed with the embeddings created with RH-emo. We obtain a consistent improvement
in the test accuracy for all datasets, while drastically reducing the
resources' demand of models. Moreover, we performed additional experiments and
ablation studies that confirm the effectiveness of our approach. The RH-emo
repository is available at: https://github.com/ispamm/rhemo.
| [
{
"version": "v1",
"created": "Tue, 5 Apr 2022 17:45:09 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 13:16:25 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Guizzo",
"Eric",
""
],
[
"Weyde",
"Tillman",
""
],
[
"Scardapane",
"Simone",
""
],
[
"Comminiello",
"Danilo",
""
]
] | TITLE: Learning Speech Emotion Representations in the Quaternion Domain
ABSTRACT: The modeling of human emotion expression in speech signals is an important,
yet challenging task. The high resource demand of speech emotion recognition
models, combined with the the general scarcity of emotion-labelled data are
obstacles to the development and application of effective solutions in this
field. In this paper, we present an approach to jointly circumvent these
difficulties. Our method, named RH-emo, is a novel semi-supervised architecture
aimed at extracting quaternion embeddings from real-valued monoaural
spectrograms, enabling the use of quaternion-valued networks for speech emotion
recognition tasks. RH-emo is a hybrid real/quaternion autoencoder network that
consists of a real-valued encoder in parallel to a real-valued emotion
classifier and a quaternion-valued decoder. On the one hand, the classifier
permits to optimize each latent axis of the embeddings for the classification
of a specific emotion-related characteristic: valence, arousal, dominance and
overall emotion. On the other hand, the quaternion reconstruction enables the
latent dimension to develop intra-channel correlations that are required for an
effective representation as a quaternion entity. We test our approach on speech
emotion recognition tasks using four popular datasets: Iemocap, Ravdess, EmoDb
and Tess, comparing the performance of three well-established real-valued CNN
architectures (AlexNet, ResNet-50, VGG) and their quaternion-valued equivalent
fed with the embeddings created with RH-emo. We obtain a consistent improvement
in the test accuracy for all datasets, while drastically reducing the
resources' demand of models. Moreover, we performed additional experiments and
ablation studies that confirm the effectiveness of our approach. The RH-emo
repository is available at: https://github.com/ispamm/rhemo.
|
2205.15419 | Gabriel Laberge | Gabriel Laberge, Ulrich A\"ivodji, Satoshi Hara, Mario Marchand. and
Foutse Khomh | Fool SHAP with Stealthily Biased Sampling | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | SHAP explanations aim at identifying which features contribute the most to
the difference in model prediction at a specific input versus a background
distribution. Recent studies have shown that they can be manipulated by
malicious adversaries to produce arbitrary desired explanations. However,
existing attacks focus solely on altering the black-box model itself. In this
paper, we propose a complementary family of attacks that leave the model intact
and manipulate SHAP explanations using stealthily biased sampling of the data
points used to approximate expectations w.r.t the background distribution. In
the context of fairness audit, we show that our attack can reduce the
importance of a sensitive feature when explaining the difference in outcomes
between groups while remaining undetected. More precisely, experiments
performed on real-world datasets showed that our attack could yield up to a
90\% relative decrease in amplitude of the sensitive feature attribution. These
results highlight the manipulability of SHAP explanations and encourage
auditors to treat them with skepticism.
| [
{
"version": "v1",
"created": "Mon, 30 May 2022 20:33:46 GMT"
},
{
"version": "v2",
"created": "Thu, 29 Sep 2022 14:39:33 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 15:10:57 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Laberge",
"Gabriel",
""
],
[
"Aïvodji",
"Ulrich",
""
],
[
"Hara",
"Satoshi",
""
],
[
"Marchand.",
"Mario",
""
],
[
"Khomh",
"Foutse",
""
]
] | TITLE: Fool SHAP with Stealthily Biased Sampling
ABSTRACT: SHAP explanations aim at identifying which features contribute the most to
the difference in model prediction at a specific input versus a background
distribution. Recent studies have shown that they can be manipulated by
malicious adversaries to produce arbitrary desired explanations. However,
existing attacks focus solely on altering the black-box model itself. In this
paper, we propose a complementary family of attacks that leave the model intact
and manipulate SHAP explanations using stealthily biased sampling of the data
points used to approximate expectations w.r.t the background distribution. In
the context of fairness audit, we show that our attack can reduce the
importance of a sensitive feature when explaining the difference in outcomes
between groups while remaining undetected. More precisely, experiments
performed on real-world datasets showed that our attack could yield up to a
90\% relative decrease in amplitude of the sensitive feature attribution. These
results highlight the manipulability of SHAP explanations and encourage
auditors to treat them with skepticism.
|
2207.02862 | Gabriel Loaiza-Ganem | Bradley C.A. Brown, Anthony L. Caterini, Brendan Leigh Ross, Jesse C.
Cresswell, Gabriel Loaiza-Ganem | Verifying the Union of Manifolds Hypothesis for Image Data | ICLR 2023 | null | null | null | stat.ML cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Deep learning has had tremendous success at learning low-dimensional
representations of high-dimensional data. This success would be impossible if
there was no hidden low-dimensional structure in data of interest; this
existence is posited by the manifold hypothesis, which states that the data
lies on an unknown manifold of low intrinsic dimension. In this paper, we argue
that this hypothesis does not properly capture the low-dimensional structure
typically present in image data. Assuming that data lies on a single manifold
implies intrinsic dimension is identical across the entire data space, and does
not allow for subregions of this space to have a different number of factors of
variation. To address this deficiency, we consider the union of manifolds
hypothesis, which states that data lies on a disjoint union of manifolds of
varying intrinsic dimensions. We empirically verify this hypothesis on
commonly-used image datasets, finding that indeed, observed data lies on a
disconnected set and that intrinsic dimension is not constant. We also provide
insights into the implications of the union of manifolds hypothesis in deep
learning, both supervised and unsupervised, showing that designing models with
an inductive bias for this structure improves performance across classification
and generative modelling tasks. Our code is available at
https://github.com/layer6ai-labs/UoMH.
| [
{
"version": "v1",
"created": "Wed, 6 Jul 2022 18:00:00 GMT"
},
{
"version": "v2",
"created": "Mon, 28 Nov 2022 19:00:19 GMT"
},
{
"version": "v3",
"created": "Thu, 2 Mar 2023 19:00:00 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Brown",
"Bradley C. A.",
""
],
[
"Caterini",
"Anthony L.",
""
],
[
"Ross",
"Brendan Leigh",
""
],
[
"Cresswell",
"Jesse C.",
""
],
[
"Loaiza-Ganem",
"Gabriel",
""
]
] | TITLE: Verifying the Union of Manifolds Hypothesis for Image Data
ABSTRACT: Deep learning has had tremendous success at learning low-dimensional
representations of high-dimensional data. This success would be impossible if
there was no hidden low-dimensional structure in data of interest; this
existence is posited by the manifold hypothesis, which states that the data
lies on an unknown manifold of low intrinsic dimension. In this paper, we argue
that this hypothesis does not properly capture the low-dimensional structure
typically present in image data. Assuming that data lies on a single manifold
implies intrinsic dimension is identical across the entire data space, and does
not allow for subregions of this space to have a different number of factors of
variation. To address this deficiency, we consider the union of manifolds
hypothesis, which states that data lies on a disjoint union of manifolds of
varying intrinsic dimensions. We empirically verify this hypothesis on
commonly-used image datasets, finding that indeed, observed data lies on a
disconnected set and that intrinsic dimension is not constant. We also provide
insights into the implications of the union of manifolds hypothesis in deep
learning, both supervised and unsupervised, showing that designing models with
an inductive bias for this structure improves performance across classification
and generative modelling tasks. Our code is available at
https://github.com/layer6ai-labs/UoMH.
|
2207.12534 | Huan Wang | Huan Wang and Yun Fu | Trainability Preserving Neural Pruning | ICLR'23 Camera Ready. 21 Pages. Code:
https://github.com/MingSun-Tse/TPP | null | null | null | cs.LG cs.AI cs.CV cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many recent works have shown trainability plays a central role in neural
network pruning -- unattended broken trainability can lead to severe
under-performance and unintentionally amplify the effect of retraining learning
rate, resulting in biased (or even misinterpreted) benchmark results. This
paper introduces trainability preserving pruning (TPP), a scalable method to
preserve network trainability against pruning, aiming for improved pruning
performance and being more robust to retraining hyper-parameters (e.g.,
learning rate). Specifically, we propose to penalize the gram matrix of
convolutional filters to decorrelate the pruned filters from the retained
filters. In addition to the convolutional layers, per the spirit of preserving
the trainability of the whole network, we also propose to regularize the batch
normalization parameters (scale and bias). Empirical studies on linear MLP
networks show that TPP can perform on par with the oracle trainability recovery
scheme. On nonlinear ConvNets (ResNet56/VGG19) on CIFAR10/100, TPP outperforms
the other counterpart approaches by an obvious margin. Moreover, results on
ImageNet-1K with ResNets suggest that TPP consistently performs more favorably
against other top-performing structured pruning approaches. Code:
https://github.com/MingSun-Tse/TPP.
| [
{
"version": "v1",
"created": "Mon, 25 Jul 2022 21:15:47 GMT"
},
{
"version": "v2",
"created": "Fri, 19 Aug 2022 21:13:11 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 05:39:11 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wang",
"Huan",
""
],
[
"Fu",
"Yun",
""
]
] | TITLE: Trainability Preserving Neural Pruning
ABSTRACT: Many recent works have shown trainability plays a central role in neural
network pruning -- unattended broken trainability can lead to severe
under-performance and unintentionally amplify the effect of retraining learning
rate, resulting in biased (or even misinterpreted) benchmark results. This
paper introduces trainability preserving pruning (TPP), a scalable method to
preserve network trainability against pruning, aiming for improved pruning
performance and being more robust to retraining hyper-parameters (e.g.,
learning rate). Specifically, we propose to penalize the gram matrix of
convolutional filters to decorrelate the pruned filters from the retained
filters. In addition to the convolutional layers, per the spirit of preserving
the trainability of the whole network, we also propose to regularize the batch
normalization parameters (scale and bias). Empirical studies on linear MLP
networks show that TPP can perform on par with the oracle trainability recovery
scheme. On nonlinear ConvNets (ResNet56/VGG19) on CIFAR10/100, TPP outperforms
the other counterpart approaches by an obvious margin. Moreover, results on
ImageNet-1K with ResNets suggest that TPP consistently performs more favorably
against other top-performing structured pruning approaches. Code:
https://github.com/MingSun-Tse/TPP.
|
2208.01890 | Qiong Wu | Qiong Wu, Xiaobo Wang, Qiang Fan, Pingyi Fan, Cui Zhang and Zhengquan
Li | High stable and accurate vehicle selection scheme based on federated
edge learning in vehicular networks | This paper has been accepted by China Communications | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated edge learning (FEEL) technology for vehicular networks is
considered as a promising technology to reduce the computation workload while
keeping the privacy of users. In the FEEL system, vehicles upload data to the
edge servers, which train the vehicles' data to update local models and then
return the result to vehicles to avoid sharing the original data. However, the
cache queue in the edge is limited and the channel between edge server and each
vehicle is time-varying. Thus, it is challenging to select a suitable number of
vehicles to ensure that the uploaded data can keep a stable cache queue in edge
server while maximizing the learning accuracy. Moreover, selecting vehicles
with different resource statuses to update data will affect the total amount of
data involved in training, which further affects the model accuracy. In this
paper, we propose a vehicle selection scheme, which maximizes the learning
accuracy while ensuring the stability of the cache queue, where the statuses of
all the vehicles in the coverage of edge server are taken into account. The
performance of this scheme is evaluated through simulation experiments, which
indicates that our proposed scheme can perform better than the known benchmark
scheme.
| [
{
"version": "v1",
"created": "Wed, 3 Aug 2022 07:39:37 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 10:23:49 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wu",
"Qiong",
""
],
[
"Wang",
"Xiaobo",
""
],
[
"Fan",
"Qiang",
""
],
[
"Fan",
"Pingyi",
""
],
[
"Zhang",
"Cui",
""
],
[
"Li",
"Zhengquan",
""
]
] | TITLE: High stable and accurate vehicle selection scheme based on federated
edge learning in vehicular networks
ABSTRACT: Federated edge learning (FEEL) technology for vehicular networks is
considered as a promising technology to reduce the computation workload while
keeping the privacy of users. In the FEEL system, vehicles upload data to the
edge servers, which train the vehicles' data to update local models and then
return the result to vehicles to avoid sharing the original data. However, the
cache queue in the edge is limited and the channel between edge server and each
vehicle is time-varying. Thus, it is challenging to select a suitable number of
vehicles to ensure that the uploaded data can keep a stable cache queue in edge
server while maximizing the learning accuracy. Moreover, selecting vehicles
with different resource statuses to update data will affect the total amount of
data involved in training, which further affects the model accuracy. In this
paper, we propose a vehicle selection scheme, which maximizes the learning
accuracy while ensuring the stability of the cache queue, where the statuses of
all the vehicles in the coverage of edge server are taken into account. The
performance of this scheme is evaluated through simulation experiments, which
indicates that our proposed scheme can perform better than the known benchmark
scheme.
|
2208.04933 | Jimmy Smith | Jimmy T.H. Smith, Andrew Warrington, and Scott W. Linderman | Simplified State Space Layers for Sequence Modeling | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Models using structured state space sequence (S4) layers have achieved
state-of-the-art performance on long-range sequence modeling tasks. An S4 layer
combines linear state space models (SSMs), the HiPPO framework, and deep
learning to achieve high performance. We build on the design of the S4 layer
and introduce a new state space layer, the S5 layer. Whereas an S4 layer uses
many independent single-input, single-output SSMs, the S5 layer uses one
multi-input, multi-output SSM. We establish a connection between S5 and S4, and
use this to develop the initialization and parameterization used by the S5
model. The result is a state space layer that can leverage efficient and widely
implemented parallel scans, allowing S5 to match the computational efficiency
of S4, while also achieving state-of-the-art performance on several long-range
sequence modeling tasks. S5 averages 87.4% on the long range arena benchmark,
and 98.5% on the most difficult Path-X task.
| [
{
"version": "v1",
"created": "Tue, 9 Aug 2022 17:57:43 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Oct 2022 03:42:06 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 18:35:28 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Smith",
"Jimmy T. H.",
""
],
[
"Warrington",
"Andrew",
""
],
[
"Linderman",
"Scott W.",
""
]
] | TITLE: Simplified State Space Layers for Sequence Modeling
ABSTRACT: Models using structured state space sequence (S4) layers have achieved
state-of-the-art performance on long-range sequence modeling tasks. An S4 layer
combines linear state space models (SSMs), the HiPPO framework, and deep
learning to achieve high performance. We build on the design of the S4 layer
and introduce a new state space layer, the S5 layer. Whereas an S4 layer uses
many independent single-input, single-output SSMs, the S5 layer uses one
multi-input, multi-output SSM. We establish a connection between S5 and S4, and
use this to develop the initialization and parameterization used by the S5
model. The result is a state space layer that can leverage efficient and widely
implemented parallel scans, allowing S5 to match the computational efficiency
of S4, while also achieving state-of-the-art performance on several long-range
sequence modeling tasks. S5 averages 87.4% on the long range arena benchmark,
and 98.5% on the most difficult Path-X task.
|
2208.09322 | Jianfei Ma | Jianfei Ma | Entropy Augmented Reinforcement Learning | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep reinforcement learning was instigated with the presence of trust region
methods, being scalable and efficient. However, the pessimism of such
algorithms, among which it forces to constrain in a trust region by all means,
has been proven to suppress the exploration and harm the performance.
Exploratory algorithm such as SAC, while utilizes the entropy to encourage
exploration, implicitly optimizing another objective yet. We first observed
this inconsistency, and therefore put forward an analogous augmentation
technique, which combines well with the on-policy algorithms, when a value
critic is involved. Surprisingly, the proposed method consistently satisfies
the soft policy improvement theorem, while being more extensible. As the
analysis advises, it is crucial to control the temperature coefficient to
balance the exploration and exploitation. Empirical tests on MuJoCo benchmark
tasks show that the agent is heartened towards higher reward regions, and
enjoys a finer performance. Furthermore, we verify the exploration bonus of our
method on a set of custom environments.
| [
{
"version": "v1",
"created": "Fri, 19 Aug 2022 13:09:32 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 12:25:11 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Ma",
"Jianfei",
""
]
] | TITLE: Entropy Augmented Reinforcement Learning
ABSTRACT: Deep reinforcement learning was instigated with the presence of trust region
methods, being scalable and efficient. However, the pessimism of such
algorithms, among which it forces to constrain in a trust region by all means,
has been proven to suppress the exploration and harm the performance.
Exploratory algorithm such as SAC, while utilizes the entropy to encourage
exploration, implicitly optimizing another objective yet. We first observed
this inconsistency, and therefore put forward an analogous augmentation
technique, which combines well with the on-policy algorithms, when a value
critic is involved. Surprisingly, the proposed method consistently satisfies
the soft policy improvement theorem, while being more extensible. As the
analysis advises, it is crucial to control the temperature coefficient to
balance the exploration and exploitation. Empirical tests on MuJoCo benchmark
tasks show that the agent is heartened towards higher reward regions, and
enjoys a finer performance. Furthermore, we verify the exploration bonus of our
method on a set of custom environments.
|
2208.13661 | Chongyang Tao | Kai Zhang, Chongyang Tao, Tao Shen, Can Xu, Xiubo Geng, Binxing Jiao,
Daxin Jiang | LED: Lexicon-Enlightened Dense Retriever for Large-Scale Retrieval | 14 pages, 6 tables, 4 figures. WWW 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Retrieval models based on dense representations in semantic space have become
an indispensable branch for first-stage retrieval. These retrievers benefit
from surging advances in representation learning towards compressive global
sequence-level embeddings. However, they are prone to overlook local salient
phrases and entity mentions in texts, which usually play pivot roles in
first-stage retrieval. To mitigate this weakness, we propose to make a dense
retriever align a well-performing lexicon-aware representation model. The
alignment is achieved by weakened knowledge distillations to enlighten the
retriever via two aspects -- 1) a lexicon-augmented contrastive objective to
challenge the dense encoder and 2) a pair-wise rank-consistent regularization
to make dense model's behavior incline to the other. We evaluate our model on
three public benchmarks, which shows that with a comparable lexicon-aware
retriever as the teacher, our proposed dense one can bring consistent and
significant improvements, and even outdo its teacher. In addition, we found our
improvement on the dense retriever is complementary to the standard ranker
distillation, which can further lift state-of-the-art performance.
| [
{
"version": "v1",
"created": "Mon, 29 Aug 2022 15:09:28 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 03:42:13 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhang",
"Kai",
""
],
[
"Tao",
"Chongyang",
""
],
[
"Shen",
"Tao",
""
],
[
"Xu",
"Can",
""
],
[
"Geng",
"Xiubo",
""
],
[
"Jiao",
"Binxing",
""
],
[
"Jiang",
"Daxin",
""
]
] | TITLE: LED: Lexicon-Enlightened Dense Retriever for Large-Scale Retrieval
ABSTRACT: Retrieval models based on dense representations in semantic space have become
an indispensable branch for first-stage retrieval. These retrievers benefit
from surging advances in representation learning towards compressive global
sequence-level embeddings. However, they are prone to overlook local salient
phrases and entity mentions in texts, which usually play pivot roles in
first-stage retrieval. To mitigate this weakness, we propose to make a dense
retriever align a well-performing lexicon-aware representation model. The
alignment is achieved by weakened knowledge distillations to enlighten the
retriever via two aspects -- 1) a lexicon-augmented contrastive objective to
challenge the dense encoder and 2) a pair-wise rank-consistent regularization
to make dense model's behavior incline to the other. We evaluate our model on
three public benchmarks, which shows that with a comparable lexicon-aware
retriever as the teacher, our proposed dense one can bring consistent and
significant improvements, and even outdo its teacher. In addition, we found our
improvement on the dense retriever is complementary to the standard ranker
distillation, which can further lift state-of-the-art performance.
|
2209.00082 | Pierre Zins | Pierre Zins, Yuanlu Xu, Edmond Boyer, Stefanie Wuhrer, Tony Tung | Multi-View Reconstruction using Signed Ray Distance Functions (SRDF) | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we investigate a new optimization framework for multi-view 3D
shape reconstructions. Recent differentiable rendering approaches have provided
breakthrough performances with implicit shape representations though they can
still lack precision in the estimated geometries. On the other hand multi-view
stereo methods can yield pixel wise geometric accuracy with local depth
predictions along viewing rays. Our approach bridges the gap between the two
strategies with a novel volumetric shape representation that is implicit but
parameterized with pixel depths to better materialize the shape surface with
consistent signed distances along viewing rays. The approach retains
pixel-accuracy while benefiting from volumetric integration in the
optimization. To this aim, depths are optimized by evaluating, at each 3D
location within the volumetric discretization, the agreement between the depth
prediction consistency and the photometric consistency for the corresponding
pixels. The optimization is agnostic to the associated photo-consistency term
which can vary from a median-based baseline to more elaborate criteria learned
functions. Our experiments demonstrate the benefit of the volumetric
integration with depth predictions. They also show that our approach
outperforms existing approaches over standard 3D benchmarks with better
geometry estimations.
| [
{
"version": "v1",
"created": "Wed, 31 Aug 2022 19:32:17 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 17:27:47 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zins",
"Pierre",
""
],
[
"Xu",
"Yuanlu",
""
],
[
"Boyer",
"Edmond",
""
],
[
"Wuhrer",
"Stefanie",
""
],
[
"Tung",
"Tony",
""
]
] | TITLE: Multi-View Reconstruction using Signed Ray Distance Functions (SRDF)
ABSTRACT: In this paper, we investigate a new optimization framework for multi-view 3D
shape reconstructions. Recent differentiable rendering approaches have provided
breakthrough performances with implicit shape representations though they can
still lack precision in the estimated geometries. On the other hand multi-view
stereo methods can yield pixel wise geometric accuracy with local depth
predictions along viewing rays. Our approach bridges the gap between the two
strategies with a novel volumetric shape representation that is implicit but
parameterized with pixel depths to better materialize the shape surface with
consistent signed distances along viewing rays. The approach retains
pixel-accuracy while benefiting from volumetric integration in the
optimization. To this aim, depths are optimized by evaluating, at each 3D
location within the volumetric discretization, the agreement between the depth
prediction consistency and the photometric consistency for the corresponding
pixels. The optimization is agnostic to the associated photo-consistency term
which can vary from a median-based baseline to more elaborate criteria learned
functions. Our experiments demonstrate the benefit of the volumetric
integration with depth predictions. They also show that our approach
outperforms existing approaches over standard 3D benchmarks with better
geometry estimations.
|
2209.00404 | Laurent Colbois | Laurent Colbois, S\'ebastien Marcel | On the detection of morphing attacks generated by GANs | Accepted for publication at BIOSIG 2022 | null | 10.1109/BIOSIG55365.2022.9897046 | null | cs.CV eess.IV | http://creativecommons.org/licenses/by-sa/4.0/ | Recent works have demonstrated the feasibility of GAN-based morphing attacks
that reach similar success rates as more traditional landmark-based methods.
This new type of "deep" morphs might require the development of new adequate
detectors to protect face recognition systems. We explore simple deep morph
detection baselines based on spectral features and LBP histograms features, as
well as on CNN models, both in the intra-dataset and cross-dataset case. We
observe that simple LBP-based systems are already quite accurate in the
intra-dataset setting, but struggle with generalization, a phenomenon that is
partially mitigated by fusing together several of those systems at score-level.
We conclude that a pretrained ResNet effective for GAN image detection is the
most effective overall, reaching close to perfect accuracy. We note however
that LBP-based systems maintain a level of interest : additionally to their
lower computational requirements and increased interpretability with respect to
CNNs, LBP+ResNet fusions sometimes also showcase increased performance versus
ResNet-only, hinting that LBP-based systems can focus on meaningful signal that
is not necessarily picked up by the CNN detector.
| [
{
"version": "v1",
"created": "Thu, 1 Sep 2022 12:28:55 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Colbois",
"Laurent",
""
],
[
"Marcel",
"Sébastien",
""
]
] | TITLE: On the detection of morphing attacks generated by GANs
ABSTRACT: Recent works have demonstrated the feasibility of GAN-based morphing attacks
that reach similar success rates as more traditional landmark-based methods.
This new type of "deep" morphs might require the development of new adequate
detectors to protect face recognition systems. We explore simple deep morph
detection baselines based on spectral features and LBP histograms features, as
well as on CNN models, both in the intra-dataset and cross-dataset case. We
observe that simple LBP-based systems are already quite accurate in the
intra-dataset setting, but struggle with generalization, a phenomenon that is
partially mitigated by fusing together several of those systems at score-level.
We conclude that a pretrained ResNet effective for GAN image detection is the
most effective overall, reaching close to perfect accuracy. We note however
that LBP-based systems maintain a level of interest : additionally to their
lower computational requirements and increased interpretability with respect to
CNNs, LBP+ResNet fusions sometimes also showcase increased performance versus
ResNet-only, hinting that LBP-based systems can focus on meaningful signal that
is not necessarily picked up by the CNN detector.
|
2209.03102 | Yang Jiao | Yang Jiao, Zequn Jie, Shaoxiang Chen, Jingjing Chen, Lin Ma, Yu-Gang
Jiang | MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth
Seeds for 3D Object Detection | Accepted by CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Fusing LiDAR and camera information is essential for achieving accurate and
reliable 3D object detection in autonomous driving systems. This is challenging
due to the difficulty of combining multi-granularity geometric and semantic
features from two drastically different modalities. Recent approaches aim at
exploring the semantic densities of camera features through lifting points in
2D camera images (referred to as seeds) into 3D space, and then incorporate 2D
semantics via cross-modal interaction or fusion techniques. However, depth
information is under-investigated in these approaches when lifting points into
3D space, thus 2D semantics can not be reliably fused with 3D points. Moreover,
their multi-modal fusion strategy, which is implemented as concatenation or
attention, either can not effectively fuse 2D and 3D information or is unable
to perform fine-grained interactions in the voxel space. To this end, we
propose a novel framework with better utilization of the depth information and
fine-grained cross-modal interaction between LiDAR and camera, which consists
of two important components. First, a Multi-Depth Unprojection (MDU) method
with depth-aware designs is used to enhance the depth quality of the lifted
points at each interaction level. Second, a Gated Modality-Aware Convolution
(GMA-Conv) block is applied to modulate voxels involved with the camera
modality in a fine-grained manner and then aggregate multi-modal features into
a unified space. Together they provide the detection head with more
comprehensive features from LiDAR and camera. On the nuScenes test benchmark,
our proposed method, abbreviated as MSMDFusion, achieves state-of-the-art 3D
object detection results with 71.5% mAP and 74.0% NDS, and strong tracking
results with 74.0% AMOTA without using test-time-augmentation and ensemble
techniques. The code is available at https://github.com/SxJyJay/MSMDFusion.
| [
{
"version": "v1",
"created": "Wed, 7 Sep 2022 12:29:29 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Nov 2022 03:06:17 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 08:16:08 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Jiao",
"Yang",
""
],
[
"Jie",
"Zequn",
""
],
[
"Chen",
"Shaoxiang",
""
],
[
"Chen",
"Jingjing",
""
],
[
"Ma",
"Lin",
""
],
[
"Jiang",
"Yu-Gang",
""
]
] | TITLE: MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth
Seeds for 3D Object Detection
ABSTRACT: Fusing LiDAR and camera information is essential for achieving accurate and
reliable 3D object detection in autonomous driving systems. This is challenging
due to the difficulty of combining multi-granularity geometric and semantic
features from two drastically different modalities. Recent approaches aim at
exploring the semantic densities of camera features through lifting points in
2D camera images (referred to as seeds) into 3D space, and then incorporate 2D
semantics via cross-modal interaction or fusion techniques. However, depth
information is under-investigated in these approaches when lifting points into
3D space, thus 2D semantics can not be reliably fused with 3D points. Moreover,
their multi-modal fusion strategy, which is implemented as concatenation or
attention, either can not effectively fuse 2D and 3D information or is unable
to perform fine-grained interactions in the voxel space. To this end, we
propose a novel framework with better utilization of the depth information and
fine-grained cross-modal interaction between LiDAR and camera, which consists
of two important components. First, a Multi-Depth Unprojection (MDU) method
with depth-aware designs is used to enhance the depth quality of the lifted
points at each interaction level. Second, a Gated Modality-Aware Convolution
(GMA-Conv) block is applied to modulate voxels involved with the camera
modality in a fine-grained manner and then aggregate multi-modal features into
a unified space. Together they provide the detection head with more
comprehensive features from LiDAR and camera. On the nuScenes test benchmark,
our proposed method, abbreviated as MSMDFusion, achieves state-of-the-art 3D
object detection results with 71.5% mAP and 74.0% NDS, and strong tracking
results with 74.0% AMOTA without using test-time-augmentation and ensemble
techniques. The code is available at https://github.com/SxJyJay/MSMDFusion.
|
2209.10271 | Dominik Bär | Dominik B\"ar, Fausto Calderon, Michael Lawlor, Sophia Licklederer,
Manuel Totzauer, Stefan Feuerriegel | Analyzing Social Media Activities at Bellingcat | Accepted at the ACM Web Science Conference (WebSci, 2023) | null | 10.1145/3578503.3583604 | null | cs.SI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Open-source journalism emerged as a new phenomenon in the media ecosystem,
which uses crowdsourcing to fact-check and generate investigative reports for
world events using open sources (e.g., social media). A particularly prominent
example is Bellingcat. Bellingcat is known for its investigations on the
illegal use of chemical weapons during the Syrian war, the Russian
responsibility for downing flight MH17, the identification of the perpetrators
in the attempted murder of Alexei Navalny, and war crimes in the Russo-Ukraine
war. Crucial for this is social media in order to disseminate findings and
crowdsource fact-checks. In this work, we characterize the social media
activities at Bellingcat on Twitter. For this, we built a comprehensive dataset
of all N=24,682 tweets posted by Bellingcat on Twitter since its inception in
July 2014. Our analysis is three-fold: (1) We analyze how Bellingcat uses
Twitter to disseminate information and collect information from its follower
base. Here, we find a steady increase in both posts and replies over time,
particularly during the Russo-Ukrainian war, which is in line with the growing
importance of Bellingcat for the traditional media ecosystem. (2) We identify
characteristics of posts that are successful in eliciting user engagement. User
engagement is particularly large for posts embedding additional media items and
with a more negative sentiment. (3) We examine how the follower base has
responded to the Russian invasion of Ukraine. Here, we find that the sentiment
has become more polarized and negative. We attribute this to a ~13-fold
increase in bots interacting with the Bellingcat account. Overall, our findings
provide recommendations for how open-source journalism such as Bellingcat can
successfully operate on social media.
| [
{
"version": "v1",
"created": "Wed, 21 Sep 2022 11:24:04 GMT"
},
{
"version": "v2",
"created": "Wed, 30 Nov 2022 16:37:58 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 11:12:52 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Bär",
"Dominik",
""
],
[
"Calderon",
"Fausto",
""
],
[
"Lawlor",
"Michael",
""
],
[
"Licklederer",
"Sophia",
""
],
[
"Totzauer",
"Manuel",
""
],
[
"Feuerriegel",
"Stefan",
""
]
] | TITLE: Analyzing Social Media Activities at Bellingcat
ABSTRACT: Open-source journalism emerged as a new phenomenon in the media ecosystem,
which uses crowdsourcing to fact-check and generate investigative reports for
world events using open sources (e.g., social media). A particularly prominent
example is Bellingcat. Bellingcat is known for its investigations on the
illegal use of chemical weapons during the Syrian war, the Russian
responsibility for downing flight MH17, the identification of the perpetrators
in the attempted murder of Alexei Navalny, and war crimes in the Russo-Ukraine
war. Crucial for this is social media in order to disseminate findings and
crowdsource fact-checks. In this work, we characterize the social media
activities at Bellingcat on Twitter. For this, we built a comprehensive dataset
of all N=24,682 tweets posted by Bellingcat on Twitter since its inception in
July 2014. Our analysis is three-fold: (1) We analyze how Bellingcat uses
Twitter to disseminate information and collect information from its follower
base. Here, we find a steady increase in both posts and replies over time,
particularly during the Russo-Ukrainian war, which is in line with the growing
importance of Bellingcat for the traditional media ecosystem. (2) We identify
characteristics of posts that are successful in eliciting user engagement. User
engagement is particularly large for posts embedding additional media items and
with a more negative sentiment. (3) We examine how the follower base has
responded to the Russian invasion of Ukraine. Here, we find that the sentiment
has become more polarized and negative. We attribute this to a ~13-fold
increase in bots interacting with the Bellingcat account. Overall, our findings
provide recommendations for how open-source journalism such as Bellingcat can
successfully operate on social media.
|
2210.02083 | Teodora Pandeva | Teodora Pandeva, Patrick Forr\'e | Multi-View Independent Component Analysis with Shared and Individual
Sources | null | null | null | null | cs.LG eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Independent component analysis (ICA) is a blind source separation method for
linear disentanglement of independent latent sources from observed data. We
investigate the special setting of noisy linear ICA where the observations are
split among different views, each receiving a mixture of shared and individual
sources. We prove that the corresponding linear structure is identifiable, and
the source distribution can be recovered. To computationally estimate the
sources, we optimize a constrained form of the joint log-likelihood of the
observed data among all views. We also show empirically that our objective
recovers the sources also in the case when the measurements are corrupted by
noise. Furthermore, we propose a model selection procedure for recovering the
number of shared sources which we verify empirically. Finally, we apply the
proposed model in a challenging real-life application, where the estimated
shared sources from two large transcriptome datasets (observed data) provided
by two different labs (two different views) lead to recovering (shared) sources
utilized for finding a plausible representation of the underlying graph
structure.
| [
{
"version": "v1",
"created": "Wed, 5 Oct 2022 08:23:05 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 15:58:47 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Pandeva",
"Teodora",
""
],
[
"Forré",
"Patrick",
""
]
] | TITLE: Multi-View Independent Component Analysis with Shared and Individual
Sources
ABSTRACT: Independent component analysis (ICA) is a blind source separation method for
linear disentanglement of independent latent sources from observed data. We
investigate the special setting of noisy linear ICA where the observations are
split among different views, each receiving a mixture of shared and individual
sources. We prove that the corresponding linear structure is identifiable, and
the source distribution can be recovered. To computationally estimate the
sources, we optimize a constrained form of the joint log-likelihood of the
observed data among all views. We also show empirically that our objective
recovers the sources also in the case when the measurements are corrupted by
noise. Furthermore, we propose a model selection procedure for recovering the
number of shared sources which we verify empirically. Finally, we apply the
proposed model in a challenging real-life application, where the estimated
shared sources from two large transcriptome datasets (observed data) provided
by two different labs (two different views) lead to recovering (shared) sources
utilized for finding a plausible representation of the underlying graph
structure.
|
2210.03802 | Jihwan Jeong | Jihwan Jeong, Xiaoyu Wang, Michael Gimelfarb, Hyunwoo Kim, Baher
Abdulhai, Scott Sanner | Conservative Bayesian Model-Based Value Expansion for Offline Policy
Optimization | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Offline reinforcement learning (RL) addresses the problem of learning a
performant policy from a fixed batch of data collected by following some
behavior policy. Model-based approaches are particularly appealing in the
offline setting since they can extract more learning signals from the logged
dataset by learning a model of the environment. However, the performance of
existing model-based approaches falls short of model-free counterparts, due to
the compounding of estimation errors in the learned model. Driven by this
observation, we argue that it is critical for a model-based method to
understand when to trust the model and when to rely on model-free estimates,
and how to act conservatively w.r.t. both. To this end, we derive an elegant
and simple methodology called conservative Bayesian model-based value expansion
for offline policy optimization (CBOP), that trades off model-free and
model-based estimates during the policy evaluation step according to their
epistemic uncertainties, and facilitates conservatism by taking a lower bound
on the Bayesian posterior value estimate. On the standard D4RL continuous
control tasks, we find that our method significantly outperforms previous
model-based approaches: e.g., MOPO by $116.4$%, MOReL by $23.2$% and COMBO by
$23.7$%. Further, CBOP achieves state-of-the-art performance on $11$ out of
$18$ benchmark datasets while doing on par on the remaining datasets.
| [
{
"version": "v1",
"created": "Fri, 7 Oct 2022 20:13:50 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 16:36:06 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Jeong",
"Jihwan",
""
],
[
"Wang",
"Xiaoyu",
""
],
[
"Gimelfarb",
"Michael",
""
],
[
"Kim",
"Hyunwoo",
""
],
[
"Abdulhai",
"Baher",
""
],
[
"Sanner",
"Scott",
""
]
] | TITLE: Conservative Bayesian Model-Based Value Expansion for Offline Policy
Optimization
ABSTRACT: Offline reinforcement learning (RL) addresses the problem of learning a
performant policy from a fixed batch of data collected by following some
behavior policy. Model-based approaches are particularly appealing in the
offline setting since they can extract more learning signals from the logged
dataset by learning a model of the environment. However, the performance of
existing model-based approaches falls short of model-free counterparts, due to
the compounding of estimation errors in the learned model. Driven by this
observation, we argue that it is critical for a model-based method to
understand when to trust the model and when to rely on model-free estimates,
and how to act conservatively w.r.t. both. To this end, we derive an elegant
and simple methodology called conservative Bayesian model-based value expansion
for offline policy optimization (CBOP), that trades off model-free and
model-based estimates during the policy evaluation step according to their
epistemic uncertainties, and facilitates conservatism by taking a lower bound
on the Bayesian posterior value estimate. On the standard D4RL continuous
control tasks, we find that our method significantly outperforms previous
model-based approaches: e.g., MOPO by $116.4$%, MOReL by $23.2$% and COMBO by
$23.7$%. Further, CBOP achieves state-of-the-art performance on $11$ out of
$18$ benchmark datasets while doing on par on the remaining datasets.
|
2211.07214 | Yifan Lu | Yifan Lu, Quanhao Li, Baoan Liu, Mehrdad Dianati, Chen Feng, Siheng
Chen, Yanfeng Wang | Robust Collaborative 3D Object Detection in Presence of Pose Errors | Accepted by ICRA2023. This version updates the final results with
data augmentation and correct AP calculations(sort the detections of all
samples by confidence) | null | null | null | cs.CV cs.MA cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Collaborative 3D object detection exploits information exchange among
multiple agents to enhance accuracy of object detection in presence of sensor
impairments such as occlusion. However, in practice, pose estimation errors due
to imperfect localization would cause spatial message misalignment and
significantly reduce the performance of collaboration. To alleviate adverse
impacts of pose errors, we propose CoAlign, a novel hybrid collaboration
framework that is robust to unknown pose errors. The proposed solution relies
on a novel agent-object pose graph modeling to enhance pose consistency among
collaborating agents. Furthermore, we adopt a multi-scale data fusion strategy
to aggregate intermediate features at multiple spatial resolutions. Comparing
with previous works, which require ground-truth pose for training supervision,
our proposed CoAlign is more practical since it doesn't require any
ground-truth pose supervision in the training and makes no specific assumptions
on pose errors. Extensive evaluation of the proposed method is carried out on
multiple datasets, certifying that CoAlign significantly reduce relative
localization error and achieving the state of art detection performance when
pose errors exist. Code are made available for the use of the research
community at https://github.com/yifanlu0227/CoAlign.
| [
{
"version": "v1",
"created": "Mon, 14 Nov 2022 09:11:14 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Nov 2022 08:30:23 GMT"
},
{
"version": "v3",
"created": "Fri, 3 Mar 2023 01:18:39 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Lu",
"Yifan",
""
],
[
"Li",
"Quanhao",
""
],
[
"Liu",
"Baoan",
""
],
[
"Dianati",
"Mehrdad",
""
],
[
"Feng",
"Chen",
""
],
[
"Chen",
"Siheng",
""
],
[
"Wang",
"Yanfeng",
""
]
] | TITLE: Robust Collaborative 3D Object Detection in Presence of Pose Errors
ABSTRACT: Collaborative 3D object detection exploits information exchange among
multiple agents to enhance accuracy of object detection in presence of sensor
impairments such as occlusion. However, in practice, pose estimation errors due
to imperfect localization would cause spatial message misalignment and
significantly reduce the performance of collaboration. To alleviate adverse
impacts of pose errors, we propose CoAlign, a novel hybrid collaboration
framework that is robust to unknown pose errors. The proposed solution relies
on a novel agent-object pose graph modeling to enhance pose consistency among
collaborating agents. Furthermore, we adopt a multi-scale data fusion strategy
to aggregate intermediate features at multiple spatial resolutions. Comparing
with previous works, which require ground-truth pose for training supervision,
our proposed CoAlign is more practical since it doesn't require any
ground-truth pose supervision in the training and makes no specific assumptions
on pose errors. Extensive evaluation of the proposed method is carried out on
multiple datasets, certifying that CoAlign significantly reduce relative
localization error and achieving the state of art detection performance when
pose errors exist. Code are made available for the use of the research
community at https://github.com/yifanlu0227/CoAlign.
|
2211.14924 | Sauradip Nag | Sauradip Nag, Xiatian Zhu, Yi-Zhe Song and Tao Xiang | Post-Processing Temporal Action Detection | CVPR 2023; Code available at https://github.com/sauradip/GAP | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing Temporal Action Detection (TAD) methods typically take a
pre-processing step in converting an input varying-length video into a
fixed-length snippet representation sequence, before temporal boundary
estimation and action classification. This pre-processing step would temporally
downsample the video, reducing the inference resolution and hampering the
detection performance in the original temporal resolution. In essence, this is
due to a temporal quantization error introduced during the resolution
downsampling and recovery. This could negatively impact the TAD performance,
but is largely ignored by existing methods. To address this problem, in this
work we introduce a novel model-agnostic post-processing method without model
redesign and retraining. Specifically, we model the start and end points of
action instances with a Gaussian distribution for enabling temporal boundary
inference at a sub-snippet level. We further introduce an efficient
Taylor-expansion based approximation, dubbed as Gaussian Approximated
Post-processing (GAP). Extensive experiments demonstrate that our GAP can
consistently improve a wide variety of pre-trained off-the-shelf TAD models on
the challenging ActivityNet (+0.2% -0.7% in average mAP) and THUMOS (+0.2%
-0.5% in average mAP) benchmarks. Such performance gains are already
significant and highly comparable to those achieved by novel model designs.
Also, GAP can be integrated with model training for further performance gain.
Importantly, GAP enables lower temporal resolutions for more efficient
inference, facilitating low-resource applications. The code will be available
in https://github.com/sauradip/GAP
| [
{
"version": "v1",
"created": "Sun, 27 Nov 2022 19:50:37 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 07:23:20 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Nag",
"Sauradip",
""
],
[
"Zhu",
"Xiatian",
""
],
[
"Song",
"Yi-Zhe",
""
],
[
"Xiang",
"Tao",
""
]
] | TITLE: Post-Processing Temporal Action Detection
ABSTRACT: Existing Temporal Action Detection (TAD) methods typically take a
pre-processing step in converting an input varying-length video into a
fixed-length snippet representation sequence, before temporal boundary
estimation and action classification. This pre-processing step would temporally
downsample the video, reducing the inference resolution and hampering the
detection performance in the original temporal resolution. In essence, this is
due to a temporal quantization error introduced during the resolution
downsampling and recovery. This could negatively impact the TAD performance,
but is largely ignored by existing methods. To address this problem, in this
work we introduce a novel model-agnostic post-processing method without model
redesign and retraining. Specifically, we model the start and end points of
action instances with a Gaussian distribution for enabling temporal boundary
inference at a sub-snippet level. We further introduce an efficient
Taylor-expansion based approximation, dubbed as Gaussian Approximated
Post-processing (GAP). Extensive experiments demonstrate that our GAP can
consistently improve a wide variety of pre-trained off-the-shelf TAD models on
the challenging ActivityNet (+0.2% -0.7% in average mAP) and THUMOS (+0.2%
-0.5% in average mAP) benchmarks. Such performance gains are already
significant and highly comparable to those achieved by novel model designs.
Also, GAP can be integrated with model training for further performance gain.
Importantly, GAP enables lower temporal resolutions for more efficient
inference, facilitating low-resource applications. The code will be available
in https://github.com/sauradip/GAP
|
2212.09612 | Chiara Calascibetta | Chiara Calascibetta, Luca Biferale, Francesco Borra, Antonio Celani
and Massimo Cencini | Taming Lagrangian Chaos with Multi-Objective Reinforcement Learning | 10 pages, 7 figures | null | 10.1140/epje/s10189-023-00271-0 | null | physics.flu-dyn cs.LG cs.MA nlin.CD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of two active particles in 2D complex flows with the
multi-objective goals of minimizing both the dispersion rate and the energy
consumption of the pair. We approach the problem by means of Multi Objective
Reinforcement Learning (MORL), combining scalarization techniques together with
a Q-learning algorithm, for Lagrangian drifters that have variable swimming
velocity. We show that MORL is able to find a set of trade-off solutions
forming an optimal Pareto frontier. As a benchmark, we show that a set of
heuristic strategies are dominated by the MORL solutions. We consider the
situation in which the agents cannot update their control variables
continuously, but only after a discrete (decision) time, $\tau$. We show that
there is a range of decision times, in between the Lyapunov time and the
continuous updating limit, where Reinforcement Learning finds strategies that
significantly improve over heuristics. In particular, we discuss how large
decision times require enhanced knowledge of the flow, whereas for smaller
$\tau$ all a priori heuristic strategies become Pareto optimal.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 16:50:58 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Calascibetta",
"Chiara",
""
],
[
"Biferale",
"Luca",
""
],
[
"Borra",
"Francesco",
""
],
[
"Celani",
"Antonio",
""
],
[
"Cencini",
"Massimo",
""
]
] | TITLE: Taming Lagrangian Chaos with Multi-Objective Reinforcement Learning
ABSTRACT: We consider the problem of two active particles in 2D complex flows with the
multi-objective goals of minimizing both the dispersion rate and the energy
consumption of the pair. We approach the problem by means of Multi Objective
Reinforcement Learning (MORL), combining scalarization techniques together with
a Q-learning algorithm, for Lagrangian drifters that have variable swimming
velocity. We show that MORL is able to find a set of trade-off solutions
forming an optimal Pareto frontier. As a benchmark, we show that a set of
heuristic strategies are dominated by the MORL solutions. We consider the
situation in which the agents cannot update their control variables
continuously, but only after a discrete (decision) time, $\tau$. We show that
there is a range of decision times, in between the Lyapunov time and the
continuous updating limit, where Reinforcement Learning finds strategies that
significantly improve over heuristics. In particular, we discuss how large
decision times require enhanced knowledge of the flow, whereas for smaller
$\tau$ all a priori heuristic strategies become Pareto optimal.
|
2212.13661 | Yang Zhou | Yang Zhou, Fuzeng Li, Xianfeng Xu, Zhen Zhang, Alexandre Ravey,
Marie-C\'ecile P\'era, Ruiqing Ma | Cost-minimization predictive energy management of a postal-delivery fuel
cell electric vehicle with intelligent battery State-of-Charge Planner | Submission to journal | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | Fuel cell electric vehicles have earned substantial attentions in recent
decades due to their high-efficiency and zero-emission features, while the high
operating costs remain the major barrier towards their large-scale
commercialization. In such context, this paper aims to devise an energy
management strategy for an urban postal-delivery fuel cell electric vehicle for
operating cost mitigation. First, a data-driven dual-loop spatial-domain
battery state-of-charge reference estimator is designed to guide battery energy
depletion, which is trained by real-world driving data collected in postal
delivery missions. Then, a fuzzy C-means clustering enhanced Markov speed
predictor is constructed to project the upcoming velocity. Lastly, combining
the state-of-charge reference and the forecasted speed, a model predictive
control-based cost-optimization energy management strategy is established to
mitigate vehicle operating costs imposed by energy consumption and power-source
degradations. Validation results have shown that 1) the proposed strategy could
mitigate the operating cost by 4.43% and 7.30% in average versus benchmark
strategies, denoting its superiority in term of cost-reduction and 2) the
computation burden per step of the proposed strategy is averaged at 0.123ms,
less than the sampling time interval 1s, proving its potential of real-time
applications.
| [
{
"version": "v1",
"created": "Wed, 28 Dec 2022 01:42:49 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 00:41:15 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhou",
"Yang",
""
],
[
"Li",
"Fuzeng",
""
],
[
"Xu",
"Xianfeng",
""
],
[
"Zhang",
"Zhen",
""
],
[
"Ravey",
"Alexandre",
""
],
[
"Péra",
"Marie-Cécile",
""
],
[
"Ma",
"Ruiqing",
""
]
] | TITLE: Cost-minimization predictive energy management of a postal-delivery fuel
cell electric vehicle with intelligent battery State-of-Charge Planner
ABSTRACT: Fuel cell electric vehicles have earned substantial attentions in recent
decades due to their high-efficiency and zero-emission features, while the high
operating costs remain the major barrier towards their large-scale
commercialization. In such context, this paper aims to devise an energy
management strategy for an urban postal-delivery fuel cell electric vehicle for
operating cost mitigation. First, a data-driven dual-loop spatial-domain
battery state-of-charge reference estimator is designed to guide battery energy
depletion, which is trained by real-world driving data collected in postal
delivery missions. Then, a fuzzy C-means clustering enhanced Markov speed
predictor is constructed to project the upcoming velocity. Lastly, combining
the state-of-charge reference and the forecasted speed, a model predictive
control-based cost-optimization energy management strategy is established to
mitigate vehicle operating costs imposed by energy consumption and power-source
degradations. Validation results have shown that 1) the proposed strategy could
mitigate the operating cost by 4.43% and 7.30% in average versus benchmark
strategies, denoting its superiority in term of cost-reduction and 2) the
computation burden per step of the proposed strategy is averaged at 0.123ms,
less than the sampling time interval 1s, proving its potential of real-time
applications.
|
2301.01214 | Georgia Papacharalampous | Georgia Papacharalampous, Hristos Tyralis, Anastasios Doulamis,
Nikolaos Doulamis | Comparison of tree-based ensemble algorithms for merging satellite and
earth-observed precipitation data at the daily time scale | null | Hydrology 10 (2023) 50 | 10.3390/hydrology10020050 | null | cs.LG stat.AP stat.CO stat.ME | http://creativecommons.org/licenses/by/4.0/ | Merging satellite products and ground-based measurements is often required
for obtaining precipitation datasets that simultaneously cover large regions
with high density and are more accurate than pure satellite precipitation
products. Machine and statistical learning regression algorithms are regularly
utilized in this endeavour. At the same time, tree-based ensemble algorithms
are adopted in various fields for solving regression problems with high
accuracy and low computational cost. Still, information on which tree-based
ensemble algorithm to select for correcting satellite precipitation products
for the contiguous United States (US) at the daily time scale is missing from
the literature. In this study, we worked towards filling this methodological
gap by conducting an extensive comparison between three algorithms of the
category of interest, specifically between random forests, gradient boosting
machines (gbm) and extreme gradient boosting (XGBoost). We used daily data from
the PERSIANN (Precipitation Estimation from Remotely Sensed Information using
Artificial Neural Networks) and the IMERG (Integrated Multi-satellitE
Retrievals for GPM) gridded datasets. We also used earth-observed precipitation
data from the Global Historical Climatology Network daily (GHCNd) database. The
experiments referred to the entire contiguous US and additionally included the
application of the linear regression algorithm for benchmarking purposes. The
results suggest that XGBoost is the best-performing tree-based ensemble
algorithm among those compared...
| [
{
"version": "v1",
"created": "Sat, 31 Dec 2022 11:14:45 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 16:10:26 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Papacharalampous",
"Georgia",
""
],
[
"Tyralis",
"Hristos",
""
],
[
"Doulamis",
"Anastasios",
""
],
[
"Doulamis",
"Nikolaos",
""
]
] | TITLE: Comparison of tree-based ensemble algorithms for merging satellite and
earth-observed precipitation data at the daily time scale
ABSTRACT: Merging satellite products and ground-based measurements is often required
for obtaining precipitation datasets that simultaneously cover large regions
with high density and are more accurate than pure satellite precipitation
products. Machine and statistical learning regression algorithms are regularly
utilized in this endeavour. At the same time, tree-based ensemble algorithms
are adopted in various fields for solving regression problems with high
accuracy and low computational cost. Still, information on which tree-based
ensemble algorithm to select for correcting satellite precipitation products
for the contiguous United States (US) at the daily time scale is missing from
the literature. In this study, we worked towards filling this methodological
gap by conducting an extensive comparison between three algorithms of the
category of interest, specifically between random forests, gradient boosting
machines (gbm) and extreme gradient boosting (XGBoost). We used daily data from
the PERSIANN (Precipitation Estimation from Remotely Sensed Information using
Artificial Neural Networks) and the IMERG (Integrated Multi-satellitE
Retrievals for GPM) gridded datasets. We also used earth-observed precipitation
data from the Global Historical Climatology Network daily (GHCNd) database. The
experiments referred to the entire contiguous US and additionally included the
application of the linear regression algorithm for benchmarking purposes. The
results suggest that XGBoost is the best-performing tree-based ensemble
algorithm among those compared...
|
2301.01252 | Georgia Papacharalampous | Georgia Papacharalampous, Hristos Tyralis, Anastasios Doulamis,
Nikolaos Doulamis | Comparison of machine learning algorithms for merging gridded satellite
and earth-observed precipitation data | null | Water 15 (2023) 634 | 10.3390/w15040634 | null | physics.ao-ph cs.LG stat.AP stat.CO stat.ME | http://creativecommons.org/licenses/by/4.0/ | Gridded satellite precipitation datasets are useful in hydrological
applications as they cover large regions with high density. However, they are
not accurate in the sense that they do not agree with ground-based
measurements. An established means for improving their accuracy is to correct
them by adopting machine learning algorithms. This correction takes the form of
a regression problem, in which the ground-based measurements have the role of
the dependent variable and the satellite data are the predictor variables,
together with topography factors (e.g., elevation). Most studies of this kind
involve a limited number of machine learning algorithms, and are conducted for
a small region and for a limited time period. Thus, the results obtained
through them are of local importance and do not provide more general guidance
and best practices. To provide results that are generalizable and to contribute
to the delivery of best practices, we here compare eight state-of-the-art
machine learning algorithms in correcting satellite precipitation data for the
entire contiguous United States and for a 15-year period. We use monthly data
from the PERSIANN (Precipitation Estimation from Remotely Sensed Information
using Artificial Neural Networks) gridded dataset, together with monthly
earth-observed precipitation data from the Global Historical Climatology
Network monthly database, version 2 (GHCNm). The results suggest that extreme
gradient boosting (XGBoost) and random forests are the most accurate in terms
of the squared error scoring function. The remaining algorithms can be ordered
as follows from the best to the worst: Bayesian regularized feed-forward neural
networks, multivariate adaptive polynomial splines (poly-MARS), gradient
boosting machines (gbm), multivariate adaptive regression splines (MARS),
feed-forward neural networks, linear regression.
| [
{
"version": "v1",
"created": "Sat, 17 Dec 2022 09:39:39 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 16:01:28 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Papacharalampous",
"Georgia",
""
],
[
"Tyralis",
"Hristos",
""
],
[
"Doulamis",
"Anastasios",
""
],
[
"Doulamis",
"Nikolaos",
""
]
] | TITLE: Comparison of machine learning algorithms for merging gridded satellite
and earth-observed precipitation data
ABSTRACT: Gridded satellite precipitation datasets are useful in hydrological
applications as they cover large regions with high density. However, they are
not accurate in the sense that they do not agree with ground-based
measurements. An established means for improving their accuracy is to correct
them by adopting machine learning algorithms. This correction takes the form of
a regression problem, in which the ground-based measurements have the role of
the dependent variable and the satellite data are the predictor variables,
together with topography factors (e.g., elevation). Most studies of this kind
involve a limited number of machine learning algorithms, and are conducted for
a small region and for a limited time period. Thus, the results obtained
through them are of local importance and do not provide more general guidance
and best practices. To provide results that are generalizable and to contribute
to the delivery of best practices, we here compare eight state-of-the-art
machine learning algorithms in correcting satellite precipitation data for the
entire contiguous United States and for a 15-year period. We use monthly data
from the PERSIANN (Precipitation Estimation from Remotely Sensed Information
using Artificial Neural Networks) gridded dataset, together with monthly
earth-observed precipitation data from the Global Historical Climatology
Network monthly database, version 2 (GHCNm). The results suggest that extreme
gradient boosting (XGBoost) and random forests are the most accurate in terms
of the squared error scoring function. The remaining algorithms can be ordered
as follows from the best to the worst: Bayesian regularized feed-forward neural
networks, multivariate adaptive polynomial splines (poly-MARS), gradient
boosting machines (gbm), multivariate adaptive regression splines (MARS),
feed-forward neural networks, linear regression.
|
2301.05403 | Hongrui Xuan | Hongrui Xuan, Yi Liu, Bohan Li, Hongzhi Yin | Knowledge Enhancement for Contrastive Multi-Behavior Recommendation | null | Proceedings of the Sixteenth ACM International Conference on Web
Search and Data Mining. 2023: 195-203 | 10.1145/3539597.3570386 | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A well-designed recommender system can accurately capture the attributes of
users and items, reflecting the unique preferences of individuals. Traditional
recommendation techniques usually focus on modeling the singular type of
behaviors between users and items. However, in many practical recommendation
scenarios (e.g., social media, e-commerce), there exist multi-typed interactive
behaviors in user-item relationships, such as click, tag-as-favorite, and
purchase in online shopping platforms. Thus, how to make full use of
multi-behavior information for recommendation is of great importance to the
existing system, which presents challenges in two aspects that need to be
explored: (1) Utilizing users' personalized preferences to capture
multi-behavioral dependencies; (2) Dealing with the insufficient recommendation
caused by sparse supervision signal for target behavior. In this work, we
propose a Knowledge Enhancement Multi-Behavior Contrastive Learning
Recommendation (KMCLR) framework, including two Contrastive Learning tasks and
three functional modules to tackle the above challenges, respectively. In
particular, we design the multi-behavior learning module to extract users'
personalized behavior information for user-embedding enhancement, and utilize
knowledge graph in the knowledge enhancement module to derive more robust
knowledge-aware representations for items. In addition, in the optimization
stage, we model the coarse-grained commonalities and the fine-grained
differences between multi-behavior of users to further improve the
recommendation effect. Extensive experiments and ablation tests on the three
real-world datasets indicate our KMCLR outperforms various state-of-the-art
recommendation methods and verify the effectiveness of our method.
| [
{
"version": "v1",
"created": "Fri, 13 Jan 2023 06:24:33 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 17:33:39 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Xuan",
"Hongrui",
""
],
[
"Liu",
"Yi",
""
],
[
"Li",
"Bohan",
""
],
[
"Yin",
"Hongzhi",
""
]
] | TITLE: Knowledge Enhancement for Contrastive Multi-Behavior Recommendation
ABSTRACT: A well-designed recommender system can accurately capture the attributes of
users and items, reflecting the unique preferences of individuals. Traditional
recommendation techniques usually focus on modeling the singular type of
behaviors between users and items. However, in many practical recommendation
scenarios (e.g., social media, e-commerce), there exist multi-typed interactive
behaviors in user-item relationships, such as click, tag-as-favorite, and
purchase in online shopping platforms. Thus, how to make full use of
multi-behavior information for recommendation is of great importance to the
existing system, which presents challenges in two aspects that need to be
explored: (1) Utilizing users' personalized preferences to capture
multi-behavioral dependencies; (2) Dealing with the insufficient recommendation
caused by sparse supervision signal for target behavior. In this work, we
propose a Knowledge Enhancement Multi-Behavior Contrastive Learning
Recommendation (KMCLR) framework, including two Contrastive Learning tasks and
three functional modules to tackle the above challenges, respectively. In
particular, we design the multi-behavior learning module to extract users'
personalized behavior information for user-embedding enhancement, and utilize
knowledge graph in the knowledge enhancement module to derive more robust
knowledge-aware representations for items. In addition, in the optimization
stage, we model the coarse-grained commonalities and the fine-grained
differences between multi-behavior of users to further improve the
recommendation effect. Extensive experiments and ablation tests on the three
real-world datasets indicate our KMCLR outperforms various state-of-the-art
recommendation methods and verify the effectiveness of our method.
|
2301.09121 | Jilan Xu | Jilan Xu, Junlin Hou, Yuejie Zhang, Rui Feng, Yi Wang, Yu Qiao, Weidi
Xie | Learning Open-vocabulary Semantic Segmentation Models From Natural
Language Supervision | Accepted by CVPR2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we consider the problem of open-vocabulary semantic
segmentation (OVS), which aims to segment objects of arbitrary classes instead
of pre-defined, closed-set categories. The main contributions are as follows:
First, we propose a transformer-based model for OVS, termed as OVSegmentor,
which only exploits web-crawled image-text pairs for pre-training without using
any mask annotations. OVSegmentor assembles the image pixels into a set of
learnable group tokens via a slot-attention based binding module, and aligns
the group tokens to the corresponding caption embedding. Second, we propose two
proxy tasks for training, namely masked entity completion and cross-image mask
consistency. The former aims to infer all masked entities in the caption given
the group tokens, that enables the model to learn fine-grained alignment
between visual groups and text entities. The latter enforces consistent mask
predictions between images that contain shared entities, which encourages the
model to learn visual invariance. Third, we construct CC4M dataset for
pre-training by filtering CC12M with frequently appeared entities, which
significantly improves training efficiency. Fourth, we perform zero-shot
transfer on three benchmark datasets, PASCAL VOC 2012, PASCAL Context, and COCO
Object. Our model achieves superior segmentation results over the
state-of-the-art method by using only 3\% data (4M vs 134M) for pre-training.
Code and pre-trained models will be released for future research.
| [
{
"version": "v1",
"created": "Sun, 22 Jan 2023 13:10:05 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 04:23:55 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Xu",
"Jilan",
""
],
[
"Hou",
"Junlin",
""
],
[
"Zhang",
"Yuejie",
""
],
[
"Feng",
"Rui",
""
],
[
"Wang",
"Yi",
""
],
[
"Qiao",
"Yu",
""
],
[
"Xie",
"Weidi",
""
]
] | TITLE: Learning Open-vocabulary Semantic Segmentation Models From Natural
Language Supervision
ABSTRACT: In this paper, we consider the problem of open-vocabulary semantic
segmentation (OVS), which aims to segment objects of arbitrary classes instead
of pre-defined, closed-set categories. The main contributions are as follows:
First, we propose a transformer-based model for OVS, termed as OVSegmentor,
which only exploits web-crawled image-text pairs for pre-training without using
any mask annotations. OVSegmentor assembles the image pixels into a set of
learnable group tokens via a slot-attention based binding module, and aligns
the group tokens to the corresponding caption embedding. Second, we propose two
proxy tasks for training, namely masked entity completion and cross-image mask
consistency. The former aims to infer all masked entities in the caption given
the group tokens, that enables the model to learn fine-grained alignment
between visual groups and text entities. The latter enforces consistent mask
predictions between images that contain shared entities, which encourages the
model to learn visual invariance. Third, we construct CC4M dataset for
pre-training by filtering CC12M with frequently appeared entities, which
significantly improves training efficiency. Fourth, we perform zero-shot
transfer on three benchmark datasets, PASCAL VOC 2012, PASCAL Context, and COCO
Object. Our model achieves superior segmentation results over the
state-of-the-art method by using only 3\% data (4M vs 134M) for pre-training.
Code and pre-trained models will be released for future research.
|
2301.09255 | Teru Nagamori | Teru Nagamori and Hitoshi Kiya | Combined Use of Federated Learning and Image Encryption for
Privacy-Preserving Image Classification with Vision Transformer | null | null | null | null | cs.CV cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | In recent years, privacy-preserving methods for deep learning have become an
urgent problem. Accordingly, we propose the combined use of federated learning
(FL) and encrypted images for privacy-preserving image classification under the
use of the vision transformer (ViT). The proposed method allows us not only to
train models over multiple participants without directly sharing their raw data
but to also protect the privacy of test (query) images for the first time. In
addition, it can also maintain the same accuracy as normally trained models. In
an experiment, the proposed method was demonstrated to well work without any
performance degradation on the CIFAR-10 and CIFAR-100 datasets.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2023 03:41:02 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 14:09:15 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Nagamori",
"Teru",
""
],
[
"Kiya",
"Hitoshi",
""
]
] | TITLE: Combined Use of Federated Learning and Image Encryption for
Privacy-Preserving Image Classification with Vision Transformer
ABSTRACT: In recent years, privacy-preserving methods for deep learning have become an
urgent problem. Accordingly, we propose the combined use of federated learning
(FL) and encrypted images for privacy-preserving image classification under the
use of the vision transformer (ViT). The proposed method allows us not only to
train models over multiple participants without directly sharing their raw data
but to also protect the privacy of test (query) images for the first time. In
addition, it can also maintain the same accuracy as normally trained models. In
an experiment, the proposed method was demonstrated to well work without any
performance degradation on the CIFAR-10 and CIFAR-100 datasets.
|
2302.00586 | Shuo Sun | Shuo Sun and Molei Qin and Xinrun Wang and Bo An | PRUDEX-Compass: Towards Systematic Evaluation of Reinforcement Learning
in Financial Markets | null | null | null | null | q-fin.TR cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | The financial markets, which involve more than $90 trillion market capitals,
attract the attention of innumerable investors around the world. Recently,
reinforcement learning in financial markets (FinRL) has emerged as a promising
direction to train agents for making profitable investment decisions. However,
the evaluation of most FinRL methods only focuses on profit-related measures
and ignores many critical axes, which are far from satisfactory for financial
practitioners to deploy these methods into real-world financial markets.
Therefore, we introduce PRUDEX-Compass, which has 6 axes, i.e., Profitability,
Risk-control, Universality, Diversity, rEliability, and eXplainability, with a
total of 17 measures for a systematic evaluation. Specifically, i) we propose
AlphaMix+ as a strong FinRL baseline, which leverages mixture-of-experts (MoE)
and risk-sensitive approaches to make diversified risk-aware investment
decisions, ii) we evaluate 8 FinRL methods in 4 long-term real-world datasets
of influential financial markets to demonstrate the usage of our
PRUDEX-Compass, iii) PRUDEX-Compass together with 4 real-world datasets,
standard implementation of 8 FinRL methods and a portfolio management
environment is released as public resources to facilitate the design and
comparison of new FinRL methods. We hope that PRUDEX-Compass can not only shed
light on future FinRL research to prevent untrustworthy results from stagnating
FinRL into successful industry deployment but also provide a new challenging
algorithm evaluation scenario for the reinforcement learning (RL) community.
| [
{
"version": "v1",
"created": "Sat, 14 Jan 2023 06:39:03 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 20:03:13 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Sun",
"Shuo",
""
],
[
"Qin",
"Molei",
""
],
[
"Wang",
"Xinrun",
""
],
[
"An",
"Bo",
""
]
] | TITLE: PRUDEX-Compass: Towards Systematic Evaluation of Reinforcement Learning
in Financial Markets
ABSTRACT: The financial markets, which involve more than $90 trillion market capitals,
attract the attention of innumerable investors around the world. Recently,
reinforcement learning in financial markets (FinRL) has emerged as a promising
direction to train agents for making profitable investment decisions. However,
the evaluation of most FinRL methods only focuses on profit-related measures
and ignores many critical axes, which are far from satisfactory for financial
practitioners to deploy these methods into real-world financial markets.
Therefore, we introduce PRUDEX-Compass, which has 6 axes, i.e., Profitability,
Risk-control, Universality, Diversity, rEliability, and eXplainability, with a
total of 17 measures for a systematic evaluation. Specifically, i) we propose
AlphaMix+ as a strong FinRL baseline, which leverages mixture-of-experts (MoE)
and risk-sensitive approaches to make diversified risk-aware investment
decisions, ii) we evaluate 8 FinRL methods in 4 long-term real-world datasets
of influential financial markets to demonstrate the usage of our
PRUDEX-Compass, iii) PRUDEX-Compass together with 4 real-world datasets,
standard implementation of 8 FinRL methods and a portfolio management
environment is released as public resources to facilitate the design and
comparison of new FinRL methods. We hope that PRUDEX-Compass can not only shed
light on future FinRL research to prevent untrustworthy results from stagnating
FinRL into successful industry deployment but also provide a new challenging
algorithm evaluation scenario for the reinforcement learning (RL) community.
|
2302.06683 | Mehryar Abbasi | Mehryar Abbasi, Parvaneh Saeedi | Enhancing Multivariate Time Series Classifiers through Self-Attention
and Relative Positioning Infusion | null | null | null | null | cs.CV eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time Series Classification (TSC) is an important and challenging task for
many visual computing applications. Despite the extensive range of methods
developed for TSC, relatively few utilized Deep Neural Networks (DNNs). In this
paper, we propose two novel attention blocks (Global Temporal Attention and
Temporal Pseudo-Gaussian augmented Self-Attention) that can enhance deep
learning-based TSC approaches, even when such approaches are designed and
optimized for a specific dataset or task. We validate this claim by evaluating
multiple state-of-the-art deep learning-based TSC models on the University of
East Anglia (UEA) benchmark, a standardized collection of 30 Multivariate Time
Series Classification (MTSC) datasets. We show that adding the proposed
attention blocks improves base models' average accuracy by up to 3.6%.
Additionally, the proposed TPS block uses a new injection module to include the
relative positional information in transformers. As a standalone unit with less
computational complexity, it enables TPS to perform better than most of the
state-of-the-art DNN-based TSC methods. The source codes for our experimental
setups and proposed attention blocks are made publicly available.
| [
{
"version": "v1",
"created": "Mon, 13 Feb 2023 20:50:34 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 01:34:34 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Abbasi",
"Mehryar",
""
],
[
"Saeedi",
"Parvaneh",
""
]
] | TITLE: Enhancing Multivariate Time Series Classifiers through Self-Attention
and Relative Positioning Infusion
ABSTRACT: Time Series Classification (TSC) is an important and challenging task for
many visual computing applications. Despite the extensive range of methods
developed for TSC, relatively few utilized Deep Neural Networks (DNNs). In this
paper, we propose two novel attention blocks (Global Temporal Attention and
Temporal Pseudo-Gaussian augmented Self-Attention) that can enhance deep
learning-based TSC approaches, even when such approaches are designed and
optimized for a specific dataset or task. We validate this claim by evaluating
multiple state-of-the-art deep learning-based TSC models on the University of
East Anglia (UEA) benchmark, a standardized collection of 30 Multivariate Time
Series Classification (MTSC) datasets. We show that adding the proposed
attention blocks improves base models' average accuracy by up to 3.6%.
Additionally, the proposed TPS block uses a new injection module to include the
relative positional information in transformers. As a standalone unit with less
computational complexity, it enables TPS to perform better than most of the
state-of-the-art DNN-based TSC methods. The source codes for our experimental
setups and proposed attention blocks are made publicly available.
|
2302.07261 | Raghav Singhal | Raghav Singhal, Mark Goldstein, Rajesh Ranganath | Where to Diffuse, How to Diffuse, and How to Get Back: Automated
Learning for Multivariate Diffusions | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Diffusion-based generative models (DBGMs) perturb data to a target noise
distribution and reverse this process to generate samples. The choice of
noising process, or inference diffusion process, affects both likelihoods and
sample quality. For example, extending the inference process with auxiliary
variables leads to improved sample quality. While there are many such
multivariate diffusions to explore, each new one requires significant
model-specific analysis, hindering rapid prototyping and evaluation. In this
work, we study Multivariate Diffusion Models (MDMs). For any number of
auxiliary variables, we provide a recipe for maximizing a lower-bound on the
MDMs likelihood without requiring any model-specific analysis. We then
demonstrate how to parameterize the diffusion for a specified target noise
distribution; these two points together enable optimizing the inference
diffusion process. Optimizing the diffusion expands easy experimentation from
just a few well-known processes to an automatic search over all linear
diffusions. To demonstrate these ideas, we introduce two new specific
diffusions as well as learn a diffusion process on the MNIST, CIFAR10, and
ImageNet32 datasets. We show learned MDMs match or surpass bits-per-dims (BPDs)
relative to fixed choices of diffusions for a given dataset and model
architecture.
| [
{
"version": "v1",
"created": "Tue, 14 Feb 2023 18:57:04 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 05:40:21 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Singhal",
"Raghav",
""
],
[
"Goldstein",
"Mark",
""
],
[
"Ranganath",
"Rajesh",
""
]
] | TITLE: Where to Diffuse, How to Diffuse, and How to Get Back: Automated
Learning for Multivariate Diffusions
ABSTRACT: Diffusion-based generative models (DBGMs) perturb data to a target noise
distribution and reverse this process to generate samples. The choice of
noising process, or inference diffusion process, affects both likelihoods and
sample quality. For example, extending the inference process with auxiliary
variables leads to improved sample quality. While there are many such
multivariate diffusions to explore, each new one requires significant
model-specific analysis, hindering rapid prototyping and evaluation. In this
work, we study Multivariate Diffusion Models (MDMs). For any number of
auxiliary variables, we provide a recipe for maximizing a lower-bound on the
MDMs likelihood without requiring any model-specific analysis. We then
demonstrate how to parameterize the diffusion for a specified target noise
distribution; these two points together enable optimizing the inference
diffusion process. Optimizing the diffusion expands easy experimentation from
just a few well-known processes to an automatic search over all linear
diffusions. To demonstrate these ideas, we introduce two new specific
diffusions as well as learn a diffusion process on the MNIST, CIFAR10, and
ImageNet32 datasets. We show learned MDMs match or surpass bits-per-dims (BPDs)
relative to fixed choices of diffusions for a given dataset and model
architecture.
|
2302.10784 | Xia Chen | Xia Chen, Manav Mahan Singh, Philipp Geyer | Utilizing Domain Knowledge: Robust Machine Learning for Building Energy
Prediction with Small, Inconsistent Datasets | 9 pages, 3 figures | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The demand for a huge amount of data for machine learning (ML) applications
is currently a bottleneck in an empirically dominated field. We propose a
method to combine prior knowledge with data-driven methods to significantly
reduce their data dependency. In this study, component-based machine learning
(CBML) as the knowledge-encoded data-driven method is examined in the context
of energy-efficient building engineering. It encodes the abstraction of
building structural knowledge as semantic information in the model
organization. We design a case experiment to understand the efficacy of
knowledge-encoded ML in sparse data input (1% - 0.0125% sampling rate). The
result reveals its three advanced features compared with pure ML methods: 1.
Significant improvement in the robustness of ML to extremely small-size and
inconsistent datasets; 2. Efficient data utilization from different entities'
record collections; 3. Characteristics of accepting incomplete data with high
interpretability and reduced training time. All these features provide a
promising path to alleviating the deployment bottleneck of data-intensive
methods and contribute to efficient real-world data usage. Moreover, four
necessary prerequisites are summarized in this study that ensures the target
scenario benefits by combining prior knowledge and ML generalization.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2023 08:56:11 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 16:01:49 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chen",
"Xia",
""
],
[
"Singh",
"Manav Mahan",
""
],
[
"Geyer",
"Philipp",
""
]
] | TITLE: Utilizing Domain Knowledge: Robust Machine Learning for Building Energy
Prediction with Small, Inconsistent Datasets
ABSTRACT: The demand for a huge amount of data for machine learning (ML) applications
is currently a bottleneck in an empirically dominated field. We propose a
method to combine prior knowledge with data-driven methods to significantly
reduce their data dependency. In this study, component-based machine learning
(CBML) as the knowledge-encoded data-driven method is examined in the context
of energy-efficient building engineering. It encodes the abstraction of
building structural knowledge as semantic information in the model
organization. We design a case experiment to understand the efficacy of
knowledge-encoded ML in sparse data input (1% - 0.0125% sampling rate). The
result reveals its three advanced features compared with pure ML methods: 1.
Significant improvement in the robustness of ML to extremely small-size and
inconsistent datasets; 2. Efficient data utilization from different entities'
record collections; 3. Characteristics of accepting incomplete data with high
interpretability and reduced training time. All these features provide a
promising path to alleviating the deployment bottleneck of data-intensive
methods and contribute to efficient real-world data usage. Moreover, four
necessary prerequisites are summarized in this study that ensures the target
scenario benefits by combining prior knowledge and ML generalization.
|
2302.11184 | Jin Zhu | Jin Zhu, Guang Yang and Pietro Lio | A residual dense vision transformer for medical image super-resolution
with segmentation-based perceptual loss fine-tuning | Preprint submitted to Medical Image Analysis and under review | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Super-resolution plays an essential role in medical imaging because it
provides an alternative way to achieve high spatial resolutions and image
quality with no extra acquisition costs. In the past few decades, the rapid
development of deep neural networks has promoted super-resolution performance
with novel network architectures, loss functions and evaluation metrics.
Specifically, vision transformers dominate a broad range of computer vision
tasks, but challenges still exist when applying them to low-level medical image
processing tasks. This paper proposes an efficient vision transformer with
residual dense connections and local feature fusion to achieve efficient
single-image super-resolution (SISR) of medical modalities. Moreover, we
implement a general-purpose perceptual loss with manual control for image
quality improvements of desired aspects by incorporating prior knowledge of
medical image segmentation. Compared with state-of-the-art methods on four
public medical image datasets, the proposed method achieves the best PSNR
scores of 6 modalities among seven modalities. It leads to an average
improvement of $+0.09$ dB PSNR with only 38\% parameters of SwinIR. On the
other hand, the segmentation-based perceptual loss increases $+0.14$ dB PSNR on
average for SOTA methods, including CNNs and vision transformers. Additionally,
we conduct comprehensive ablation studies to discuss potential factors for the
superior performance of vision transformers over CNNs and the impacts of
network and loss function components. The code will be released on GitHub with
the paper published.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 07:39:09 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 06:51:27 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhu",
"Jin",
""
],
[
"Yang",
"Guang",
""
],
[
"Lio",
"Pietro",
""
]
] | TITLE: A residual dense vision transformer for medical image super-resolution
with segmentation-based perceptual loss fine-tuning
ABSTRACT: Super-resolution plays an essential role in medical imaging because it
provides an alternative way to achieve high spatial resolutions and image
quality with no extra acquisition costs. In the past few decades, the rapid
development of deep neural networks has promoted super-resolution performance
with novel network architectures, loss functions and evaluation metrics.
Specifically, vision transformers dominate a broad range of computer vision
tasks, but challenges still exist when applying them to low-level medical image
processing tasks. This paper proposes an efficient vision transformer with
residual dense connections and local feature fusion to achieve efficient
single-image super-resolution (SISR) of medical modalities. Moreover, we
implement a general-purpose perceptual loss with manual control for image
quality improvements of desired aspects by incorporating prior knowledge of
medical image segmentation. Compared with state-of-the-art methods on four
public medical image datasets, the proposed method achieves the best PSNR
scores of 6 modalities among seven modalities. It leads to an average
improvement of $+0.09$ dB PSNR with only 38\% parameters of SwinIR. On the
other hand, the segmentation-based perceptual loss increases $+0.14$ dB PSNR on
average for SOTA methods, including CNNs and vision transformers. Additionally,
we conduct comprehensive ablation studies to discuss potential factors for the
superior performance of vision transformers over CNNs and the impacts of
network and loss function components. The code will be released on GitHub with
the paper published.
|
2302.11329 | Qiheng Mao | Qiheng Mao, Zemin Liu, Chenghao Liu, Jianling Sun | HINormer: Representation Learning On Heterogeneous Information Networks
with Graph Transformer | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent studies have highlighted the limitations of message-passing based
graph neural networks (GNNs), e.g., limited model expressiveness,
over-smoothing, over-squashing, etc. To alleviate these issues, Graph
Transformers (GTs) have been proposed which work in the paradigm that allows
message passing to a larger coverage even across the whole graph. Hinging on
the global range attention mechanism, GTs have shown a superpower for
representation learning on homogeneous graphs. However, the investigation of
GTs on heterogeneous information networks (HINs) is still under-exploited. In
particular, on account of the existence of heterogeneity, HINs show distinct
data characteristics and thus require different treatment. To bridge this gap,
in this paper we investigate the representation learning on HINs with Graph
Transformer, and propose a novel model named HINormer, which capitalizes on a
larger-range aggregation mechanism for node representation learning. In
particular, assisted by two major modules, i.e., a local structure encoder and
a heterogeneous relation encoder, HINormer can capture both the structural and
heterogeneous information of nodes on HINs for comprehensive node
representations. We conduct extensive experiments on four HIN benchmark
datasets, which demonstrate that our proposed model can outperform the
state-of-the-art.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 12:25:07 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 05:44:48 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Mao",
"Qiheng",
""
],
[
"Liu",
"Zemin",
""
],
[
"Liu",
"Chenghao",
""
],
[
"Sun",
"Jianling",
""
]
] | TITLE: HINormer: Representation Learning On Heterogeneous Information Networks
with Graph Transformer
ABSTRACT: Recent studies have highlighted the limitations of message-passing based
graph neural networks (GNNs), e.g., limited model expressiveness,
over-smoothing, over-squashing, etc. To alleviate these issues, Graph
Transformers (GTs) have been proposed which work in the paradigm that allows
message passing to a larger coverage even across the whole graph. Hinging on
the global range attention mechanism, GTs have shown a superpower for
representation learning on homogeneous graphs. However, the investigation of
GTs on heterogeneous information networks (HINs) is still under-exploited. In
particular, on account of the existence of heterogeneity, HINs show distinct
data characteristics and thus require different treatment. To bridge this gap,
in this paper we investigate the representation learning on HINs with Graph
Transformer, and propose a novel model named HINormer, which capitalizes on a
larger-range aggregation mechanism for node representation learning. In
particular, assisted by two major modules, i.e., a local structure encoder and
a heterogeneous relation encoder, HINormer can capture both the structural and
heterogeneous information of nodes on HINs for comprehensive node
representations. We conduct extensive experiments on four HIN benchmark
datasets, which demonstrate that our proposed model can outperform the
state-of-the-art.
|
2303.00123 | Roel Van Beeumen | Roel Van Beeumen, Daan Camps, Neil Mehta | QCLAB++: Simulating Quantum Circuits on GPUs | 13 pages, 10 figures | null | null | null | quant-ph cs.MS cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce qclab++, a light-weight, fully-templated C++ package for
GPU-accelerated quantum circuit simulations. The code offers a high degree of
portability as it has no external dependencies and the GPU kernels are
generated through OpenMP offloading. qclab++ is designed for performance and
numerical stability through highly optimized gate simulation algorithms for
1-qubit, controlled 1-qubit, and 2-qubit gates. Furthermore, we also introduce
qclab, a quantum circuit toolbox for Matlab with a syntax that mimics qclab++.
This provides users the flexibility and ease of use of a scripting language
like Matlab for studying their quantum algorithms, while offering
high-performance GPU acceleration when required. As such, the qclab++ library
offers a unique combination of features. We compare the CPU simulator in
qclab++ with the GPU kernels generated by OpenMP and observe a speedup of over
$40\times$. Furthermore, we also compare qclab++ to other circuit simulation
packages, such as cirq-qsim and qibo, in a series of benchmarks conducted on
NERSC's Perlmutter system and illustrate its competitiveness.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2023 22:56:48 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Van Beeumen",
"Roel",
""
],
[
"Camps",
"Daan",
""
],
[
"Mehta",
"Neil",
""
]
] | TITLE: QCLAB++: Simulating Quantum Circuits on GPUs
ABSTRACT: We introduce qclab++, a light-weight, fully-templated C++ package for
GPU-accelerated quantum circuit simulations. The code offers a high degree of
portability as it has no external dependencies and the GPU kernels are
generated through OpenMP offloading. qclab++ is designed for performance and
numerical stability through highly optimized gate simulation algorithms for
1-qubit, controlled 1-qubit, and 2-qubit gates. Furthermore, we also introduce
qclab, a quantum circuit toolbox for Matlab with a syntax that mimics qclab++.
This provides users the flexibility and ease of use of a scripting language
like Matlab for studying their quantum algorithms, while offering
high-performance GPU acceleration when required. As such, the qclab++ library
offers a unique combination of features. We compare the CPU simulator in
qclab++ with the GPU kernels generated by OpenMP and observe a speedup of over
$40\times$. Furthermore, we also compare qclab++ to other circuit simulation
packages, such as cirq-qsim and qibo, in a series of benchmarks conducted on
NERSC's Perlmutter system and illustrate its competitiveness.
|
2303.00164 | Amanda Baughan | Amanda Baughan, Allison Mercurio, Ariel Liu, Xuezhi Wang, Jilin Chen,
Xiao Ma | A Mixed-Methods Approach to Understanding User Trust after Voice
Assistant Failures | 16 pages, 3 figures. To appear in ACM CHI '23. for associated dataset
file, see https://www.kaggle.com/datasets/googleai/voice-assistant-failures.
Replacing the prior version with clean latex files, content remains unchanged | null | 10.1145/3544548.3581152 | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Despite huge gains in performance in natural language understanding via large
language models in recent years, voice assistants still often fail to meet user
expectations. In this study, we conducted a mixed-methods analysis of how voice
assistant failures affect users' trust in their voice assistants. To illustrate
how users have experienced these failures, we contribute a crowdsourced dataset
of 199 voice assistant failures, categorized across 12 failure sources. Relying
on interview and survey data, we find that certain failures, such as those due
to overcapturing users' input, derail user trust more than others. We
additionally examine how failures impact users' willingness to rely on voice
assistants for future tasks. Users often stop using their voice assistants for
specific tasks that result in failures for a short period of time before
resuming similar usage. We demonstrate the importance of low stakes tasks, such
as playing music, towards building trust after failures.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 01:35:16 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 00:04:01 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Baughan",
"Amanda",
""
],
[
"Mercurio",
"Allison",
""
],
[
"Liu",
"Ariel",
""
],
[
"Wang",
"Xuezhi",
""
],
[
"Chen",
"Jilin",
""
],
[
"Ma",
"Xiao",
""
]
] | TITLE: A Mixed-Methods Approach to Understanding User Trust after Voice
Assistant Failures
ABSTRACT: Despite huge gains in performance in natural language understanding via large
language models in recent years, voice assistants still often fail to meet user
expectations. In this study, we conducted a mixed-methods analysis of how voice
assistant failures affect users' trust in their voice assistants. To illustrate
how users have experienced these failures, we contribute a crowdsourced dataset
of 199 voice assistant failures, categorized across 12 failure sources. Relying
on interview and survey data, we find that certain failures, such as those due
to overcapturing users' input, derail user trust more than others. We
additionally examine how failures impact users' willingness to rely on voice
assistants for future tasks. Users often stop using their voice assistants for
specific tasks that result in failures for a short period of time before
resuming similar usage. We demonstrate the importance of low stakes tasks, such
as playing music, towards building trust after failures.
|
2303.00795 | Sadhana Ravikumar | Sadhana Ravikumar, Ranjit Ittyerah, Sydney Lim, Long Xie, Sandhitsu
Das, Pulkit Khandelwal, Laura E.M. Wisse, Madigan L. Bedard, John L.
Robinson, Terry Schuck, Murray Grossman, John Q. Trojanowski, Edward B. Lee,
M. Dylan Tisdall, Karthik Prabhakaran, John A. Detre, David J. Irwin,
Winifred Trotman, Gabor Mizsei, Emilio Artacho-P\'erula, Maria Mercedes
I\~niguez de Onzono Martin, Maria del Mar Arroyo Jim\'enez, Monica Mu\~noz,
Francisco Javier Molina Romero, Maria del Pilar Marcos Rabal, Sandra
Cebada-S\'anchez, Jos\'e Carlos Delgado Gonz\'alez, Carlos de la Rosa-Prieto,
Marta C\'orcoles Parada, David A. Wolk, Ricardo Insausti, Paul A. Yushkevich | Improved Segmentation of Deep Sulci in Cortical Gray Matter Using a Deep
Learning Framework Incorporating Laplace's Equation | Accepted at the 28th biennial international conference on Information
Processing in Medical Imaging (IPMI 2023) | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | When developing tools for automated cortical segmentation, the ability to
produce topologically correct segmentations is important in order to compute
geometrically valid morphometry measures. In practice, accurate cortical
segmentation is challenged by image artifacts and the highly convoluted anatomy
of the cortex itself. To address this, we propose a novel deep learning-based
cortical segmentation method in which prior knowledge about the geometry of the
cortex is incorporated into the network during the training process. We design
a loss function which uses the theory of Laplace's equation applied to the
cortex to locally penalize unresolved boundaries between tightly folded sulci.
Using an ex vivo MRI dataset of human medial temporal lobe specimens, we
demonstrate that our approach outperforms baseline segmentation networks, both
quantitatively and qualitatively.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 19:48:45 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 15:47:01 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Ravikumar",
"Sadhana",
""
],
[
"Ittyerah",
"Ranjit",
""
],
[
"Lim",
"Sydney",
""
],
[
"Xie",
"Long",
""
],
[
"Das",
"Sandhitsu",
""
],
[
"Khandelwal",
"Pulkit",
""
],
[
"Wisse",
"Laura E. M.",
""
],
[
"Bedard",
"Madigan L.",
""
],
[
"Robinson",
"John L.",
""
],
[
"Schuck",
"Terry",
""
],
[
"Grossman",
"Murray",
""
],
[
"Trojanowski",
"John Q.",
""
],
[
"Lee",
"Edward B.",
""
],
[
"Tisdall",
"M. Dylan",
""
],
[
"Prabhakaran",
"Karthik",
""
],
[
"Detre",
"John A.",
""
],
[
"Irwin",
"David J.",
""
],
[
"Trotman",
"Winifred",
""
],
[
"Mizsei",
"Gabor",
""
],
[
"Artacho-Pérula",
"Emilio",
""
],
[
"Martin",
"Maria Mercedes Iñiguez de Onzono",
""
],
[
"Jiménez",
"Maria del Mar Arroyo",
""
],
[
"Muñoz",
"Monica",
""
],
[
"Romero",
"Francisco Javier Molina",
""
],
[
"Rabal",
"Maria del Pilar Marcos",
""
],
[
"Cebada-Sánchez",
"Sandra",
""
],
[
"González",
"José Carlos Delgado",
""
],
[
"de la Rosa-Prieto",
"Carlos",
""
],
[
"Parada",
"Marta Córcoles",
""
],
[
"Wolk",
"David A.",
""
],
[
"Insausti",
"Ricardo",
""
],
[
"Yushkevich",
"Paul A.",
""
]
] | TITLE: Improved Segmentation of Deep Sulci in Cortical Gray Matter Using a Deep
Learning Framework Incorporating Laplace's Equation
ABSTRACT: When developing tools for automated cortical segmentation, the ability to
produce topologically correct segmentations is important in order to compute
geometrically valid morphometry measures. In practice, accurate cortical
segmentation is challenged by image artifacts and the highly convoluted anatomy
of the cortex itself. To address this, we propose a novel deep learning-based
cortical segmentation method in which prior knowledge about the geometry of the
cortex is incorporated into the network during the training process. We design
a loss function which uses the theory of Laplace's equation applied to the
cortex to locally penalize unresolved boundaries between tightly folded sulci.
Using an ex vivo MRI dataset of human medial temporal lobe specimens, we
demonstrate that our approach outperforms baseline segmentation networks, both
quantitatively and qualitatively.
|
2303.00857 | Sun She | She Sun, Li Zhou and Xiaoran Yan | An Improved Christofides Mechanism for Local Differential Privacy
Framework | null | null | null | null | cs.CR math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The development of Internet technology enables an analysis on the whole
population rather than a certain number of samples, and leads to increasing
requirement for privacy protection. Local differential privacy (LDP) is an
effective standard of privacy measurement; however, its large variance of mean
estimation causes challenges in application. To address this problem, this
paper presents a new LDP approach, an improved Christofides mechanism.
It compared four statistical survey methods for conducting surveys on
sensitive topics -- modified Warner, Simmons, Christofides, and the improved
Christofides mechanism. Specifically, Warner, Simmons and Christofides
mechanisms have been modified to draw a sample from the population without
replacement, to decrease variance. Furthermore, by drawing cards without
replacement based on modified Christofides mechanism, we introduce a new
mechanism called the improved Christofides mechanism, which is found to have
the smallest variance under certain assumption when using LDP as a measurement
of privacy leakage. The assumption is do satisfied usually in the real world.
Actually, we decrease the variance to 28.7% of modified Christofides
mechanism's variance in our experiment based on the HCOVANY dataset -- a real
world dataset of IPUMS USA. This means our method gets a more accurate estimate
by using LDP as a measurement of privacy leakage. This is the first time the
improved Christofides mechanism is proposed for LDP framework based on
comparative analysis of four mechanisms using LDP as the same measurement of
privacy leakage.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 23:11:07 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 02:33:05 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Sun",
"She",
""
],
[
"Zhou",
"Li",
""
],
[
"Yan",
"Xiaoran",
""
]
] | TITLE: An Improved Christofides Mechanism for Local Differential Privacy
Framework
ABSTRACT: The development of Internet technology enables an analysis on the whole
population rather than a certain number of samples, and leads to increasing
requirement for privacy protection. Local differential privacy (LDP) is an
effective standard of privacy measurement; however, its large variance of mean
estimation causes challenges in application. To address this problem, this
paper presents a new LDP approach, an improved Christofides mechanism.
It compared four statistical survey methods for conducting surveys on
sensitive topics -- modified Warner, Simmons, Christofides, and the improved
Christofides mechanism. Specifically, Warner, Simmons and Christofides
mechanisms have been modified to draw a sample from the population without
replacement, to decrease variance. Furthermore, by drawing cards without
replacement based on modified Christofides mechanism, we introduce a new
mechanism called the improved Christofides mechanism, which is found to have
the smallest variance under certain assumption when using LDP as a measurement
of privacy leakage. The assumption is do satisfied usually in the real world.
Actually, we decrease the variance to 28.7% of modified Christofides
mechanism's variance in our experiment based on the HCOVANY dataset -- a real
world dataset of IPUMS USA. This means our method gets a more accurate estimate
by using LDP as a measurement of privacy leakage. This is the first time the
improved Christofides mechanism is proposed for LDP framework based on
comparative analysis of four mechanisms using LDP as the same measurement of
privacy leakage.
|
2303.00870 | Mu-Huan Chung | Mu-Huan Chung, Lu Wang, Sharon Li, Yuhong Yang, Calvin Giang, Khilan
Jerath, Abhay Raman, David Lie, Mark Chignell | Implementing Active Learning in Cybersecurity: Detecting Anomalies in
Redacted Emails | null | null | null | null | cs.HC cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Research on email anomaly detection has typically relied on specially
prepared datasets that may not adequately reflect the type of data that occurs
in industry settings. In our research, at a major financial services company,
privacy concerns prevented inspection of the bodies of emails and attachment
details (although subject headings and attachment filenames were available).
This made labeling possible anomalies in the resulting redacted emails more
difficult. Another source of difficulty is the high volume of emails combined
with the scarcity of resources making machine learning (ML) a necessity, but
also creating a need for more efficient human training of ML models. Active
learning (AL) has been proposed as a way to make human training of ML models
more efficient. However, the implementation of Active Learning methods is a
human-centered AI challenge due to potential human analyst uncertainty, and the
labeling task can be further complicated in domains such as the cybersecurity
domain (or healthcare, aviation, etc.) where mistakes in labeling can have
highly adverse consequences. In this paper we present research results
concerning the application of Active Learning to anomaly detection in redacted
emails, comparing the utility of different methods for implementing active
learning in this context. We evaluate different AL strategies and their impact
on resulting model performance. We also examine how ratings of confidence that
experts have in their labels can inform AL. The results obtained are discussed
in terms of their implications for AL methodology and for the role of experts
in model-assisted email anomaly screening.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 23:53:01 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 02:09:06 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chung",
"Mu-Huan",
""
],
[
"Wang",
"Lu",
""
],
[
"Li",
"Sharon",
""
],
[
"Yang",
"Yuhong",
""
],
[
"Giang",
"Calvin",
""
],
[
"Jerath",
"Khilan",
""
],
[
"Raman",
"Abhay",
""
],
[
"Lie",
"David",
""
],
[
"Chignell",
"Mark",
""
]
] | TITLE: Implementing Active Learning in Cybersecurity: Detecting Anomalies in
Redacted Emails
ABSTRACT: Research on email anomaly detection has typically relied on specially
prepared datasets that may not adequately reflect the type of data that occurs
in industry settings. In our research, at a major financial services company,
privacy concerns prevented inspection of the bodies of emails and attachment
details (although subject headings and attachment filenames were available).
This made labeling possible anomalies in the resulting redacted emails more
difficult. Another source of difficulty is the high volume of emails combined
with the scarcity of resources making machine learning (ML) a necessity, but
also creating a need for more efficient human training of ML models. Active
learning (AL) has been proposed as a way to make human training of ML models
more efficient. However, the implementation of Active Learning methods is a
human-centered AI challenge due to potential human analyst uncertainty, and the
labeling task can be further complicated in domains such as the cybersecurity
domain (or healthcare, aviation, etc.) where mistakes in labeling can have
highly adverse consequences. In this paper we present research results
concerning the application of Active Learning to anomaly detection in redacted
emails, comparing the utility of different methods for implementing active
learning in this context. We evaluate different AL strategies and their impact
on resulting model performance. We also examine how ratings of confidence that
experts have in their labels can inform AL. The results obtained are discussed
in terms of their implications for AL methodology and for the role of experts
in model-assisted email anomaly screening.
|
2303.01289 | Rundong Luo | Rundong Luo, Yifei Wang, Yisen Wang | Rethinking the Effect of Data Augmentation in Adversarial Contrastive
Learning | ICLR 2023 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recent works have shown that self-supervised learning can achieve remarkable
robustness when integrated with adversarial training (AT). However, the
robustness gap between supervised AT (sup-AT) and self-supervised AT (self-AT)
remains significant. Motivated by this observation, we revisit existing self-AT
methods and discover an inherent dilemma that affects self-AT robustness:
either strong or weak data augmentations are harmful to self-AT, and a medium
strength is insufficient to bridge the gap. To resolve this dilemma, we propose
a simple remedy named DYNACL (Dynamic Adversarial Contrastive Learning). In
particular, we propose an augmentation schedule that gradually anneals from a
strong augmentation to a weak one to benefit from both extreme cases. Besides,
we adopt a fast post-processing stage for adapting it to downstream tasks.
Through extensive experiments, we show that DYNACL can improve state-of-the-art
self-AT robustness by 8.84% under Auto-Attack on the CIFAR-10 dataset, and can
even outperform vanilla supervised adversarial training for the first time. Our
code is available at \url{https://github.com/PKU-ML/DYNACL}.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 14:11:54 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Mar 2023 02:21:48 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Luo",
"Rundong",
""
],
[
"Wang",
"Yifei",
""
],
[
"Wang",
"Yisen",
""
]
] | TITLE: Rethinking the Effect of Data Augmentation in Adversarial Contrastive
Learning
ABSTRACT: Recent works have shown that self-supervised learning can achieve remarkable
robustness when integrated with adversarial training (AT). However, the
robustness gap between supervised AT (sup-AT) and self-supervised AT (self-AT)
remains significant. Motivated by this observation, we revisit existing self-AT
methods and discover an inherent dilemma that affects self-AT robustness:
either strong or weak data augmentations are harmful to self-AT, and a medium
strength is insufficient to bridge the gap. To resolve this dilemma, we propose
a simple remedy named DYNACL (Dynamic Adversarial Contrastive Learning). In
particular, we propose an augmentation schedule that gradually anneals from a
strong augmentation to a weak one to benefit from both extreme cases. Besides,
we adopt a fast post-processing stage for adapting it to downstream tasks.
Through extensive experiments, we show that DYNACL can improve state-of-the-art
self-AT robustness by 8.84% under Auto-Attack on the CIFAR-10 dataset, and can
even outperform vanilla supervised adversarial training for the first time. Our
code is available at \url{https://github.com/PKU-ML/DYNACL}.
|
2303.01546 | Abhinanda Ranjit Punnakkal | Abhinanda R. Punnakkal, Suyog S Jadhav, Alexander Horsch, Krishna
Agarwal, Dilip K. Prasad | MiShape: 3D Shape Modelling of Mitochondria in Microscopy | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fluorescence microscopy is a quintessential tool for observing cells and
understanding the underlying mechanisms of life-sustaining processes of all
living organisms. The problem of extracting 3D shape of mitochondria from
fluorescence microscopy images remains unsolved due to the complex and varied
shapes expressed by mitochondria and the poor resolving capacity of these
microscopes. We propose an approach to bridge this gap by learning a shape
prior for mitochondria termed as MiShape, by leveraging high-resolution
electron microscopy data. MiShape is a generative model learned using implicit
representations of mitochondrial shapes. It provides a shape distribution that
can be used to generate infinite realistic mitochondrial shapes. We demonstrate
the representation power of MiShape and its utility for 3D shape reconstruction
given a single 2D fluorescence image or a small 3D stack of 2D slices. We also
showcase applications of our method by deriving simulated fluorescence
microscope datasets that have realistic 3D ground truths for the problem of 2D
segmentation and microscope-to-microscope transformation.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 19:21:21 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Punnakkal",
"Abhinanda R.",
""
],
[
"Jadhav",
"Suyog S",
""
],
[
"Horsch",
"Alexander",
""
],
[
"Agarwal",
"Krishna",
""
],
[
"Prasad",
"Dilip K.",
""
]
] | TITLE: MiShape: 3D Shape Modelling of Mitochondria in Microscopy
ABSTRACT: Fluorescence microscopy is a quintessential tool for observing cells and
understanding the underlying mechanisms of life-sustaining processes of all
living organisms. The problem of extracting 3D shape of mitochondria from
fluorescence microscopy images remains unsolved due to the complex and varied
shapes expressed by mitochondria and the poor resolving capacity of these
microscopes. We propose an approach to bridge this gap by learning a shape
prior for mitochondria termed as MiShape, by leveraging high-resolution
electron microscopy data. MiShape is a generative model learned using implicit
representations of mitochondrial shapes. It provides a shape distribution that
can be used to generate infinite realistic mitochondrial shapes. We demonstrate
the representation power of MiShape and its utility for 3D shape reconstruction
given a single 2D fluorescence image or a small 3D stack of 2D slices. We also
showcase applications of our method by deriving simulated fluorescence
microscope datasets that have realistic 3D ground truths for the problem of 2D
segmentation and microscope-to-microscope transformation.
|
2303.01547 | Soumyabrata Dey | Sichao Li, Sean Banerjee, Natasha Kholgade Banerjee, Soumyabrata Dey | Simultaneous prediction of hand gestures, handedness, and hand keypoints
using thermal images | ICDEC 2022 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Hand gesture detection is a well-explored area in computer vision with
applications in various forms of Human-Computer Interactions. In this work, we
propose a technique for simultaneous hand gesture classification, handedness
detection, and hand keypoints localization using thermal data captured by an
infrared camera. Our method uses a novel deep multi-task learning architecture
that includes shared encoderdecoder layers followed by three branches dedicated
for each mentioned task. We performed extensive experimental validation of our
model on an in-house dataset consisting of 24 users data. The results confirm
higher than 98 percent accuracy for gesture classification, handedness
detection, and fingertips localization, and more than 91 percent accuracy for
wrist points localization.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 19:25:40 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Li",
"Sichao",
""
],
[
"Banerjee",
"Sean",
""
],
[
"Banerjee",
"Natasha Kholgade",
""
],
[
"Dey",
"Soumyabrata",
""
]
] | TITLE: Simultaneous prediction of hand gestures, handedness, and hand keypoints
using thermal images
ABSTRACT: Hand gesture detection is a well-explored area in computer vision with
applications in various forms of Human-Computer Interactions. In this work, we
propose a technique for simultaneous hand gesture classification, handedness
detection, and hand keypoints localization using thermal data captured by an
infrared camera. Our method uses a novel deep multi-task learning architecture
that includes shared encoderdecoder layers followed by three branches dedicated
for each mentioned task. We performed extensive experimental validation of our
model on an in-house dataset consisting of 24 users data. The results confirm
higher than 98 percent accuracy for gesture classification, handedness
detection, and fingertips localization, and more than 91 percent accuracy for
wrist points localization.
|
2303.01557 | Foivos Tsimpourlas | Foivos Tsimpourlas, Pavlos Petoumenos, Min Xu, Chris Cummins, Kim
Hazelwood, Ajitha Rajan, Hugh Leather | BenchDirect: A Directed Language Model for Compiler Benchmarks | arXiv admin note: substantial text overlap with arXiv:2208.06555 | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The exponential increase of hardware-software complexity has made it
impossible for compiler engineers to find the right optimization heuristics
manually. Predictive models have been shown to find near optimal heuristics
with little human effort but they are limited by a severe lack of diverse
benchmarks to train on. Generative AI has been used by researchers to
synthesize benchmarks into existing datasets. However, the synthetic programs
are short, exceedingly simple and lacking diversity in their features.
We develop BenchPress, the first ML compiler benchmark generator that can be
directed within source code feature representations. BenchPress synthesizes
executable functions by infilling code that conditions on the program's left
and right context. BenchPress uses active learning to introduce new benchmarks
with unseen features into the dataset of Grewe's et al. CPU vs GPU heuristic,
improving its acquired performance by 50%. BenchPress targets features that has
been impossible for other synthesizers to reach. In 3 feature spaces, we
outperform human-written code from GitHub, CLgen, CLSmith and the SRCIROR
mutator in targeting the features of Rodinia benchmarks.
BenchPress steers generation with beam search over a feature-agnostic
language model. We improve this with BenchDirect which utilizes a directed LM
that infills programs by jointly observing source code context and the compiler
features that are targeted. BenchDirect achieves up to 36% better accuracy in
targeting the features of Rodinia benchmarks, it is 1.8x more likely to give an
exact match and it speeds up execution time by up to 72% compared to
BenchPress. Both our models produce code that is difficult to distinguish from
human-written code. We conduct a Turing test which shows our models' synthetic
benchmarks are labelled as 'human-written' as often as human-written code from
GitHub.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 20:17:24 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Tsimpourlas",
"Foivos",
""
],
[
"Petoumenos",
"Pavlos",
""
],
[
"Xu",
"Min",
""
],
[
"Cummins",
"Chris",
""
],
[
"Hazelwood",
"Kim",
""
],
[
"Rajan",
"Ajitha",
""
],
[
"Leather",
"Hugh",
""
]
] | TITLE: BenchDirect: A Directed Language Model for Compiler Benchmarks
ABSTRACT: The exponential increase of hardware-software complexity has made it
impossible for compiler engineers to find the right optimization heuristics
manually. Predictive models have been shown to find near optimal heuristics
with little human effort but they are limited by a severe lack of diverse
benchmarks to train on. Generative AI has been used by researchers to
synthesize benchmarks into existing datasets. However, the synthetic programs
are short, exceedingly simple and lacking diversity in their features.
We develop BenchPress, the first ML compiler benchmark generator that can be
directed within source code feature representations. BenchPress synthesizes
executable functions by infilling code that conditions on the program's left
and right context. BenchPress uses active learning to introduce new benchmarks
with unseen features into the dataset of Grewe's et al. CPU vs GPU heuristic,
improving its acquired performance by 50%. BenchPress targets features that has
been impossible for other synthesizers to reach. In 3 feature spaces, we
outperform human-written code from GitHub, CLgen, CLSmith and the SRCIROR
mutator in targeting the features of Rodinia benchmarks.
BenchPress steers generation with beam search over a feature-agnostic
language model. We improve this with BenchDirect which utilizes a directed LM
that infills programs by jointly observing source code context and the compiler
features that are targeted. BenchDirect achieves up to 36% better accuracy in
targeting the features of Rodinia benchmarks, it is 1.8x more likely to give an
exact match and it speeds up execution time by up to 72% compared to
BenchPress. Both our models produce code that is difficult to distinguish from
human-written code. We conduct a Turing test which shows our models' synthetic
benchmarks are labelled as 'human-written' as often as human-written code from
GitHub.
|
2303.01575 | Arpit Narechania | Arpit Narechania and Fan Du and Atanu R Sinha and Ryan A. Rossi and
Jane Hoffswell and Shunan Guo and Eunyee Koh and Shamkant B. Navathe and Alex
Endert | DataPilot: Utilizing Quality and Usage Information for Subset Selection
during Visual Data Preparation | 18 pages, 5 figures, 1 table, ACM CHI 2023 | null | 10.1145/3544548.3581509 | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Selecting relevant data subsets from large, unfamiliar datasets can be
difficult. We address this challenge by modeling and visualizing two kinds of
auxiliary information: (1) quality - the validity and appropriateness of data
required to perform certain analytical tasks; and (2) usage - the historical
utilization characteristics of data across multiple users. Through a design
study with 14 data workers, we integrate this information into a visual data
preparation and analysis tool, DataPilot. DataPilot presents visual cues about
"the good, the bad, and the ugly" aspects of data and provides graphical user
interface controls as interaction affordances, guiding users to perform subset
selection. Through a study with 36 participants, we investigate how DataPilot
helps users navigate a large, unfamiliar tabular dataset, prepare a relevant
subset, and build a visualization dashboard. We find that users selected
smaller, effective subsets with higher quality and usage, and with greater
success and confidence.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 21:03:49 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Narechania",
"Arpit",
""
],
[
"Du",
"Fan",
""
],
[
"Sinha",
"Atanu R",
""
],
[
"Rossi",
"Ryan A.",
""
],
[
"Hoffswell",
"Jane",
""
],
[
"Guo",
"Shunan",
""
],
[
"Koh",
"Eunyee",
""
],
[
"Navathe",
"Shamkant B.",
""
],
[
"Endert",
"Alex",
""
]
] | TITLE: DataPilot: Utilizing Quality and Usage Information for Subset Selection
during Visual Data Preparation
ABSTRACT: Selecting relevant data subsets from large, unfamiliar datasets can be
difficult. We address this challenge by modeling and visualizing two kinds of
auxiliary information: (1) quality - the validity and appropriateness of data
required to perform certain analytical tasks; and (2) usage - the historical
utilization characteristics of data across multiple users. Through a design
study with 14 data workers, we integrate this information into a visual data
preparation and analysis tool, DataPilot. DataPilot presents visual cues about
"the good, the bad, and the ugly" aspects of data and provides graphical user
interface controls as interaction affordances, guiding users to perform subset
selection. Through a study with 36 participants, we investigate how DataPilot
helps users navigate a large, unfamiliar tabular dataset, prepare a relevant
subset, and build a visualization dashboard. We find that users selected
smaller, effective subsets with higher quality and usage, and with greater
success and confidence.
|
2303.01582 | Iason Katsamenis | Iason Katsamenis, Eftychios Protopapadakis, Nikolaos Bakalos,
Anastasios Doulamis, Nikolaos Doulamis, Athanasios Voulodimos | A Few-Shot Attention Recurrent Residual U-Net for Crack Segmentation | 5 pages, 4 figures, 2 tables | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Recent studies indicate that deep learning plays a crucial role in the
automated visual inspection of road infrastructures. However, current learning
schemes are static, implying no dynamic adaptation to users' feedback. To
address this drawback, we present a few-shot learning paradigm for the
automated segmentation of road cracks, which is based on a U-Net architecture
with recurrent residual and attention modules (R2AU-Net). The retraining
strategy dynamically fine-tunes the weights of the U-Net as a few new rectified
samples are being fed into the classifier. Extensive experiments show that the
proposed few-shot R2AU-Net framework outperforms other state-of-the-art
networks in terms of Dice and IoU metrics, on a new dataset, named CrackMap,
which is made publicly available at https://github.com/ikatsamenis/CrackMap.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 21:16:17 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Katsamenis",
"Iason",
""
],
[
"Protopapadakis",
"Eftychios",
""
],
[
"Bakalos",
"Nikolaos",
""
],
[
"Doulamis",
"Anastasios",
""
],
[
"Doulamis",
"Nikolaos",
""
],
[
"Voulodimos",
"Athanasios",
""
]
] | TITLE: A Few-Shot Attention Recurrent Residual U-Net for Crack Segmentation
ABSTRACT: Recent studies indicate that deep learning plays a crucial role in the
automated visual inspection of road infrastructures. However, current learning
schemes are static, implying no dynamic adaptation to users' feedback. To
address this drawback, we present a few-shot learning paradigm for the
automated segmentation of road cracks, which is based on a U-Net architecture
with recurrent residual and attention modules (R2AU-Net). The retraining
strategy dynamically fine-tunes the weights of the U-Net as a few new rectified
samples are being fed into the classifier. Extensive experiments show that the
proposed few-shot R2AU-Net framework outperforms other state-of-the-art
networks in terms of Dice and IoU metrics, on a new dataset, named CrackMap,
which is made publicly available at https://github.com/ikatsamenis/CrackMap.
|
2303.01598 | Achin Jain | Achin Jain, Gurumurthy Swaminathan, Paolo Favaro, Hao Yang, Avinash
Ravichandran, Hrayr Harutyunyan, Alessandro Achille, Onkar Dabeer, Bernt
Schiele, Ashwin Swaminathan, Stefano Soatto | A Meta-Learning Approach to Predicting Performance and Data Requirements | CVPR 2023 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an approach to estimate the number of samples required for a model
to reach a target performance. We find that the power law, the de facto
principle to estimate model performance, leads to large error when using a
small dataset (e.g., 5 samples per class) for extrapolation. This is because
the log-performance error against the log-dataset size follows a nonlinear
progression in the few-shot regime followed by a linear progression in the
high-shot regime. We introduce a novel piecewise power law (PPL) that handles
the two data regimes differently. To estimate the parameters of the PPL, we
introduce a random forest regressor trained via meta learning that generalizes
across classification/detection tasks, ResNet/ViT based architectures, and
random/pre-trained initializations. The PPL improves the performance estimation
on average by 37% across 16 classification and 33% across 10 detection
datasets, compared to the power law. We further extend the PPL to provide a
confidence bound and use it to limit the prediction horizon that reduces
over-estimation of data by 76% on classification and 91% on detection datasets.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 21:48:22 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Jain",
"Achin",
""
],
[
"Swaminathan",
"Gurumurthy",
""
],
[
"Favaro",
"Paolo",
""
],
[
"Yang",
"Hao",
""
],
[
"Ravichandran",
"Avinash",
""
],
[
"Harutyunyan",
"Hrayr",
""
],
[
"Achille",
"Alessandro",
""
],
[
"Dabeer",
"Onkar",
""
],
[
"Schiele",
"Bernt",
""
],
[
"Swaminathan",
"Ashwin",
""
],
[
"Soatto",
"Stefano",
""
]
] | TITLE: A Meta-Learning Approach to Predicting Performance and Data Requirements
ABSTRACT: We propose an approach to estimate the number of samples required for a model
to reach a target performance. We find that the power law, the de facto
principle to estimate model performance, leads to large error when using a
small dataset (e.g., 5 samples per class) for extrapolation. This is because
the log-performance error against the log-dataset size follows a nonlinear
progression in the few-shot regime followed by a linear progression in the
high-shot regime. We introduce a novel piecewise power law (PPL) that handles
the two data regimes differently. To estimate the parameters of the PPL, we
introduce a random forest regressor trained via meta learning that generalizes
across classification/detection tasks, ResNet/ViT based architectures, and
random/pre-trained initializations. The PPL improves the performance estimation
on average by 37% across 16 classification and 33% across 10 detection
datasets, compared to the power law. We further extend the PPL to provide a
confidence bound and use it to limit the prediction horizon that reduces
over-estimation of data by 76% on classification and 91% on detection datasets.
|
2303.01605 | Cheng Jiang | Cheng Jiang, Xinhai Hou, Akhil Kondepudi, Asadur Chowdury, Christian
W. Freudiger, Daniel A. Orringer, Honglak Lee, Todd C. Hollon | Hierarchical discriminative learning improves visual representations of
biomedical microscopy | CVPR 2023. Project page: https://hidisc.mlins.org | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning high-quality, self-supervised, visual representations is essential
to advance the role of computer vision in biomedical microscopy and clinical
medicine. Previous work has focused on self-supervised representation learning
(SSL) methods developed for instance discrimination and applied them directly
to image patches, or fields-of-view, sampled from gigapixel whole-slide images
(WSIs) used for cancer diagnosis. However, this strategy is limited because it
(1) assumes patches from the same patient are independent, (2) neglects the
patient-slide-patch hierarchy of clinical biomedical microscopy, and (3)
requires strong data augmentations that can degrade downstream performance.
Importantly, sampled patches from WSIs of a patient's tumor are a diverse set
of image examples that capture the same underlying cancer diagnosis. This
motivated HiDisc, a data-driven method that leverages the inherent
patient-slide-patch hierarchy of clinical biomedical microscopy to define a
hierarchical discriminative learning task that implicitly learns features of
the underlying diagnosis. HiDisc uses a self-supervised contrastive learning
framework in which positive patch pairs are defined based on a common ancestry
in the data hierarchy, and a unified patch, slide, and patient discriminative
learning objective is used for visual SSL. We benchmark HiDisc visual
representations on two vision tasks using two biomedical microscopy datasets,
and demonstrate that (1) HiDisc pretraining outperforms current
state-of-the-art self-supervised pretraining methods for cancer diagnosis and
genetic mutation prediction, and (2) HiDisc learns high-quality visual
representations using natural patch diversity without strong data
augmentations.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 22:04:42 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Jiang",
"Cheng",
""
],
[
"Hou",
"Xinhai",
""
],
[
"Kondepudi",
"Akhil",
""
],
[
"Chowdury",
"Asadur",
""
],
[
"Freudiger",
"Christian W.",
""
],
[
"Orringer",
"Daniel A.",
""
],
[
"Lee",
"Honglak",
""
],
[
"Hollon",
"Todd C.",
""
]
] | TITLE: Hierarchical discriminative learning improves visual representations of
biomedical microscopy
ABSTRACT: Learning high-quality, self-supervised, visual representations is essential
to advance the role of computer vision in biomedical microscopy and clinical
medicine. Previous work has focused on self-supervised representation learning
(SSL) methods developed for instance discrimination and applied them directly
to image patches, or fields-of-view, sampled from gigapixel whole-slide images
(WSIs) used for cancer diagnosis. However, this strategy is limited because it
(1) assumes patches from the same patient are independent, (2) neglects the
patient-slide-patch hierarchy of clinical biomedical microscopy, and (3)
requires strong data augmentations that can degrade downstream performance.
Importantly, sampled patches from WSIs of a patient's tumor are a diverse set
of image examples that capture the same underlying cancer diagnosis. This
motivated HiDisc, a data-driven method that leverages the inherent
patient-slide-patch hierarchy of clinical biomedical microscopy to define a
hierarchical discriminative learning task that implicitly learns features of
the underlying diagnosis. HiDisc uses a self-supervised contrastive learning
framework in which positive patch pairs are defined based on a common ancestry
in the data hierarchy, and a unified patch, slide, and patient discriminative
learning objective is used for visual SSL. We benchmark HiDisc visual
representations on two vision tasks using two biomedical microscopy datasets,
and demonstrate that (1) HiDisc pretraining outperforms current
state-of-the-art self-supervised pretraining methods for cancer diagnosis and
genetic mutation prediction, and (2) HiDisc learns high-quality visual
representations using natural patch diversity without strong data
augmentations.
|
2303.01630 | Chenyan Wu | Chenyan Wu, Yimu Pan, Yandong Li, James Z. Wang | Learning to Adapt to Online Streams with Distribution Shifts | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Test-time adaptation (TTA) is a technique used to reduce distribution gaps
between the training and testing sets by leveraging unlabeled test data during
inference. In this work, we expand TTA to a more practical scenario, where the
test data comes in the form of online streams that experience distribution
shifts over time. Existing approaches face two challenges: reliance on a large
test data batch from the same domain and the absence of explicitly modeling the
continual distribution evolution process. To address both challenges, we
propose a meta-learning approach that teaches the network to adapt to
distribution-shifting online streams during meta-training. As a result, the
trained model can perform continual adaptation to distribution shifts in
testing, regardless of the batch size restriction, as it has learned during
training. We conducted extensive experiments on benchmarking datasets for TTA,
incorporating a broad range of online distribution-shifting settings. Our
results showed consistent improvements over state-of-the-art methods,
indicating the effectiveness of our approach. In addition, we achieved superior
performance in the video segmentation task, highlighting the potential of our
method for real-world applications.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 23:36:10 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wu",
"Chenyan",
""
],
[
"Pan",
"Yimu",
""
],
[
"Li",
"Yandong",
""
],
[
"Wang",
"James Z.",
""
]
] | TITLE: Learning to Adapt to Online Streams with Distribution Shifts
ABSTRACT: Test-time adaptation (TTA) is a technique used to reduce distribution gaps
between the training and testing sets by leveraging unlabeled test data during
inference. In this work, we expand TTA to a more practical scenario, where the
test data comes in the form of online streams that experience distribution
shifts over time. Existing approaches face two challenges: reliance on a large
test data batch from the same domain and the absence of explicitly modeling the
continual distribution evolution process. To address both challenges, we
propose a meta-learning approach that teaches the network to adapt to
distribution-shifting online streams during meta-training. As a result, the
trained model can perform continual adaptation to distribution shifts in
testing, regardless of the batch size restriction, as it has learned during
training. We conducted extensive experiments on benchmarking datasets for TTA,
incorporating a broad range of online distribution-shifting settings. Our
results showed consistent improvements over state-of-the-art methods,
indicating the effectiveness of our approach. In addition, we achieved superior
performance in the video segmentation task, highlighting the potential of our
method for real-world applications.
|
2303.01639 | Jun Rekimoto | Jun Rekimoto | WESPER: Zero-shot and Realtime Whisper to Normal Voice Conversion for
Whisper-based Speech Interactions | ACM CHI 2023 paper | Proceedings of the 2023 CHI Conference on Human Factors in
Computing Systems (CHI '23), April 23--28, 2023 | 10.1145/3544548.3580706 | null | cs.SD cs.HC eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recognizing whispered speech and converting it to normal speech creates many
possibilities for speech interaction. Because the sound pressure of whispered
speech is significantly lower than that of normal speech, it can be used as a
semi-silent speech interaction in public places without being audible to
others. Converting whispers to normal speech also improves the speech quality
for people with speech or hearing impairments. However, conventional speech
conversion techniques do not provide sufficient conversion quality or require
speaker-dependent datasets consisting of pairs of whispered and normal speech
utterances. To address these problems, we propose WESPER, a zero-shot,
real-time whisper-to-normal speech conversion mechanism based on
self-supervised learning. WESPER consists of a speech-to-unit (STU) encoder,
which generates hidden speech units common to both whispered and normal speech,
and a unit-to-speech (UTS) decoder, which reconstructs speech from the encoded
speech units. Unlike the existing methods, this conversion is user-independent
and does not require a paired dataset for whispered and normal speech. The UTS
decoder can reconstruct speech in any target speaker's voice from speech units,
and it requires only an unlabeled target speaker's speech data. We confirmed
that the quality of the speech converted from a whisper was improved while
preserving its natural prosody. Additionally, we confirmed the effectiveness of
the proposed approach to perform speech reconstruction for people with speech
or hearing disabilities. (project page: http://lab.rekimoto.org/projects/wesper
)
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 00:10:25 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Rekimoto",
"Jun",
""
]
] | TITLE: WESPER: Zero-shot and Realtime Whisper to Normal Voice Conversion for
Whisper-based Speech Interactions
ABSTRACT: Recognizing whispered speech and converting it to normal speech creates many
possibilities for speech interaction. Because the sound pressure of whispered
speech is significantly lower than that of normal speech, it can be used as a
semi-silent speech interaction in public places without being audible to
others. Converting whispers to normal speech also improves the speech quality
for people with speech or hearing impairments. However, conventional speech
conversion techniques do not provide sufficient conversion quality or require
speaker-dependent datasets consisting of pairs of whispered and normal speech
utterances. To address these problems, we propose WESPER, a zero-shot,
real-time whisper-to-normal speech conversion mechanism based on
self-supervised learning. WESPER consists of a speech-to-unit (STU) encoder,
which generates hidden speech units common to both whispered and normal speech,
and a unit-to-speech (UTS) decoder, which reconstructs speech from the encoded
speech units. Unlike the existing methods, this conversion is user-independent
and does not require a paired dataset for whispered and normal speech. The UTS
decoder can reconstruct speech in any target speaker's voice from speech units,
and it requires only an unlabeled target speaker's speech data. We confirmed
that the quality of the speech converted from a whisper was improved while
preserving its natural prosody. Additionally, we confirmed the effectiveness of
the proposed approach to perform speech reconstruction for people with speech
or hearing disabilities. (project page: http://lab.rekimoto.org/projects/wesper
)
|
2303.01645 | Ramin Shahbazi | Ramin Shahbazi, Fatemeh Fard | APIContext2Com: Code Comment Generation by Incorporating Pre-Defined API
Documentation | null | null | null | null | cs.SE cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Code comments are significantly helpful in comprehending software programs
and also aid developers to save a great deal of time in software maintenance.
Code comment generation aims to automatically predict comments in natural
language given a code snippet. Several works investigate the effect of
integrating external knowledge on the quality of generated comments. In this
study, we propose a solution, namely APIContext2Com, to improve the
effectiveness of generated comments by incorporating the pre-defined
Application Programming Interface (API) context. The API context includes the
definition and description of the pre-defined APIs that are used within the
code snippets. As the detailed API information expresses the functionality of a
code snippet, it can be helpful in better generating the code summary. We
introduce a seq-2-seq encoder-decoder neural network model with different sets
of multiple encoders to effectively transform distinct inputs into target
comments. A ranking mechanism is also developed to exclude non-informative
APIs, so that we can filter out unrelated APIs. We evaluate our approach using
the Java dataset from CodeSearchNet. The findings reveal that the proposed
model improves the best baseline by 1.88 (8.24 %), 2.16 (17.58 %), 1.38 (18.3
%), 0.73 (14.17 %), 1.58 (14.98 %) and 1.9 (6.92 %) for BLEU1, BLEU2, BLEU3,
BLEU4, METEOR, ROUGE-L respectively. Human evaluation and ablation studies
confirm the quality of the generated comments and the effect of architecture
and ranking APIs.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 00:38:01 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Shahbazi",
"Ramin",
""
],
[
"Fard",
"Fatemeh",
""
]
] | TITLE: APIContext2Com: Code Comment Generation by Incorporating Pre-Defined API
Documentation
ABSTRACT: Code comments are significantly helpful in comprehending software programs
and also aid developers to save a great deal of time in software maintenance.
Code comment generation aims to automatically predict comments in natural
language given a code snippet. Several works investigate the effect of
integrating external knowledge on the quality of generated comments. In this
study, we propose a solution, namely APIContext2Com, to improve the
effectiveness of generated comments by incorporating the pre-defined
Application Programming Interface (API) context. The API context includes the
definition and description of the pre-defined APIs that are used within the
code snippets. As the detailed API information expresses the functionality of a
code snippet, it can be helpful in better generating the code summary. We
introduce a seq-2-seq encoder-decoder neural network model with different sets
of multiple encoders to effectively transform distinct inputs into target
comments. A ranking mechanism is also developed to exclude non-informative
APIs, so that we can filter out unrelated APIs. We evaluate our approach using
the Java dataset from CodeSearchNet. The findings reveal that the proposed
model improves the best baseline by 1.88 (8.24 %), 2.16 (17.58 %), 1.38 (18.3
%), 0.73 (14.17 %), 1.58 (14.98 %) and 1.9 (6.92 %) for BLEU1, BLEU2, BLEU3,
BLEU4, METEOR, ROUGE-L respectively. Human evaluation and ablation studies
confirm the quality of the generated comments and the effect of architecture
and ranking APIs.
|
2303.01665 | Sara Adkins | Sara Adkins, Pedro Sarmento, Mathieu Barthet | LooperGP: A Loopable Sequence Model for Live Coding Performance using
GuitarPro Tablature | The Version of Record of this contribution is published in
Proceedings of EvoMUSART: International Conference on Computational
Intelligence in Music, Sound, Art and Design (Part of EvoStar) 2023 | EvoMUSART: International Conference on Computational Intelligence
in Music, Sound, Art and Design (Part of EvoStar) 2023 | null | null | cs.SD cs.MM eess.AS | http://creativecommons.org/licenses/by/4.0/ | Despite their impressive offline results, deep learning models for symbolic
music generation are not widely used in live performances due to a deficit of
musically meaningful control parameters and a lack of structured musical form
in their outputs. To address these issues we introduce LooperGP, a method for
steering a Transformer-XL model towards generating loopable musical phrases of
a specified number of bars and time signature, enabling a tool for live coding
performances. We show that by training LooperGP on a dataset of 93,681 musical
loops extracted from the DadaGP dataset, we are able to steer its generative
output towards generating 3x as many loopable phrases as our baseline. In a
subjective listening test conducted by 31 participants, LooperGP loops achieved
positive median ratings in originality, musical coherence and loop smoothness,
demonstrating its potential as a performance tool.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 02:00:49 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Adkins",
"Sara",
""
],
[
"Sarmento",
"Pedro",
""
],
[
"Barthet",
"Mathieu",
""
]
] | TITLE: LooperGP: A Loopable Sequence Model for Live Coding Performance using
GuitarPro Tablature
ABSTRACT: Despite their impressive offline results, deep learning models for symbolic
music generation are not widely used in live performances due to a deficit of
musically meaningful control parameters and a lack of structured musical form
in their outputs. To address these issues we introduce LooperGP, a method for
steering a Transformer-XL model towards generating loopable musical phrases of
a specified number of bars and time signature, enabling a tool for live coding
performances. We show that by training LooperGP on a dataset of 93,681 musical
loops extracted from the DadaGP dataset, we are able to steer its generative
output towards generating 3x as many loopable phrases as our baseline. In a
subjective listening test conducted by 31 participants, LooperGP loops achieved
positive median ratings in originality, musical coherence and loop smoothness,
demonstrating its potential as a performance tool.
|
2303.01668 | Chuheng Zhang | Yuanying Cai, Chuheng Zhang, Wei Shen, Xuyun Zhang, Wenjie Ruan,
Longbo Huang | RePreM: Representation Pre-training with Masked Model for Reinforcement
Learning | Accepted by AAAI-23 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Inspired by the recent success of sequence modeling in RL and the use of
masked language model for pre-training, we propose a masked model for
pre-training in RL, RePreM (Representation Pre-training with Masked Model),
which trains the encoder combined with transformer blocks to predict the masked
states or actions in a trajectory. RePreM is simple but effective compared to
existing representation pre-training methods in RL. It avoids algorithmic
sophistication (such as data augmentation or estimating multiple models) with
sequence modeling and generates a representation that captures long-term
dynamics well. Empirically, we demonstrate the effectiveness of RePreM in
various tasks, including dynamic prediction, transfer learning, and
sample-efficient RL with both value-based and actor-critic methods. Moreover,
we show that RePreM scales well with dataset size, dataset quality, and the
scale of the encoder, which indicates its potential towards big RL models.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 02:04:14 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Cai",
"Yuanying",
""
],
[
"Zhang",
"Chuheng",
""
],
[
"Shen",
"Wei",
""
],
[
"Zhang",
"Xuyun",
""
],
[
"Ruan",
"Wenjie",
""
],
[
"Huang",
"Longbo",
""
]
] | TITLE: RePreM: Representation Pre-training with Masked Model for Reinforcement
Learning
ABSTRACT: Inspired by the recent success of sequence modeling in RL and the use of
masked language model for pre-training, we propose a masked model for
pre-training in RL, RePreM (Representation Pre-training with Masked Model),
which trains the encoder combined with transformer blocks to predict the masked
states or actions in a trajectory. RePreM is simple but effective compared to
existing representation pre-training methods in RL. It avoids algorithmic
sophistication (such as data augmentation or estimating multiple models) with
sequence modeling and generates a representation that captures long-term
dynamics well. Empirically, we demonstrate the effectiveness of RePreM in
various tasks, including dynamic prediction, transfer learning, and
sample-efficient RL with both value-based and actor-critic methods. Moreover,
we show that RePreM scales well with dataset size, dataset quality, and the
scale of the encoder, which indicates its potential towards big RL models.
|
2303.01671 | Shuai Xiao | Shuai Xiao, Zaifan Jiang, Shuang Yang | Tile Networks: Learning Optimal Geometric Layout for Whole-page
Recommendation | Published at Proceedings of the 25th International Conference on
Artificial Intelligence and Statistics (AISTATS) 2022 | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Finding optimal configurations in a geometric space is a key challenge in
many technological disciplines. Current approaches either rely heavily on human
domain expertise and are difficult to scale. In this paper we show it is
possible to solve configuration optimization problems for whole-page
recommendation using reinforcement learning. The proposed \textit{Tile
Networks} is a neural architecture that optimizes 2D geometric configurations
by arranging items on proper positions. Empirical results on real dataset
demonstrate its superior performance compared to traditional learning to rank
approaches and recent deep models.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 02:18:55 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Xiao",
"Shuai",
""
],
[
"Jiang",
"Zaifan",
""
],
[
"Yang",
"Shuang",
""
]
] | TITLE: Tile Networks: Learning Optimal Geometric Layout for Whole-page
Recommendation
ABSTRACT: Finding optimal configurations in a geometric space is a key challenge in
many technological disciplines. Current approaches either rely heavily on human
domain expertise and are difficult to scale. In this paper we show it is
possible to solve configuration optimization problems for whole-page
recommendation using reinforcement learning. The proposed \textit{Tile
Networks} is a neural architecture that optimizes 2D geometric configurations
by arranging items on proper positions. Empirical results on real dataset
demonstrate its superior performance compared to traditional learning to rank
approaches and recent deep models.
|
2303.01685 | Kun Hu | Lintao Wang, Kun Hu, Lei Bai, Yu Ding, Wanli Ouyang, Zhiyong Wang | Multi-Scale Control Signal-Aware Transformer for Motion Synthesis
without Phase | null | null | null | null | cs.CV cs.GR | http://creativecommons.org/licenses/by/4.0/ | Synthesizing controllable motion for a character using deep learning has been
a promising approach due to its potential to learn a compact model without
laborious feature engineering. To produce dynamic motion from weak control
signals such as desired paths, existing methods often require auxiliary
information such as phases for alleviating motion ambiguity, which limits their
generalisation capability. As past poses often contain useful auxiliary hints,
in this paper, we propose a task-agnostic deep learning method, namely
Multi-scale Control Signal-aware Transformer (MCS-T), with an attention based
encoder-decoder architecture to discover the auxiliary information implicitly
for synthesizing controllable motion without explicitly requiring auxiliary
information such as phase. Specifically, an encoder is devised to adaptively
formulate the motion patterns of a character's past poses with multi-scale
skeletons, and a decoder driven by control signals to further synthesize and
predict the character's state by paying context-specialised attention to the
encoded past motion patterns. As a result, it helps alleviate the issues of low
responsiveness and slow transition which often happen in conventional methods
not using auxiliary information. Both qualitative and quantitative experimental
results on an existing biped locomotion dataset, which involves diverse types
of motion transitions, demonstrate the effectiveness of our method. In
particular, MCS-T is able to successfully generate motions comparable to those
generated by the methods using auxiliary information.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 02:56:44 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wang",
"Lintao",
""
],
[
"Hu",
"Kun",
""
],
[
"Bai",
"Lei",
""
],
[
"Ding",
"Yu",
""
],
[
"Ouyang",
"Wanli",
""
],
[
"Wang",
"Zhiyong",
""
]
] | TITLE: Multi-Scale Control Signal-Aware Transformer for Motion Synthesis
without Phase
ABSTRACT: Synthesizing controllable motion for a character using deep learning has been
a promising approach due to its potential to learn a compact model without
laborious feature engineering. To produce dynamic motion from weak control
signals such as desired paths, existing methods often require auxiliary
information such as phases for alleviating motion ambiguity, which limits their
generalisation capability. As past poses often contain useful auxiliary hints,
in this paper, we propose a task-agnostic deep learning method, namely
Multi-scale Control Signal-aware Transformer (MCS-T), with an attention based
encoder-decoder architecture to discover the auxiliary information implicitly
for synthesizing controllable motion without explicitly requiring auxiliary
information such as phase. Specifically, an encoder is devised to adaptively
formulate the motion patterns of a character's past poses with multi-scale
skeletons, and a decoder driven by control signals to further synthesize and
predict the character's state by paying context-specialised attention to the
encoded past motion patterns. As a result, it helps alleviate the issues of low
responsiveness and slow transition which often happen in conventional methods
not using auxiliary information. Both qualitative and quantitative experimental
results on an existing biped locomotion dataset, which involves diverse types
of motion transitions, demonstrate the effectiveness of our method. In
particular, MCS-T is able to successfully generate motions comparable to those
generated by the methods using auxiliary information.
|
2303.01686 | Shuo Wang | Shuo Wang, Xinhai Zhao, Hai-Ming Xu, Zehui Chen, Dameng Yu, Jiahao
Chang, Zhen Yang, Feng Zhao | Towards Domain Generalization for Multi-view 3D Object Detection in
Bird-Eye-View | Accepted to CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-view 3D object detection (MV3D-Det) in Bird-Eye-View (BEV) has drawn
extensive attention due to its low cost and high efficiency. Although new
algorithms for camera-only 3D object detection have been continuously proposed,
most of them may risk drastic performance degradation when the domain of input
images differs from that of training. In this paper, we first analyze the
causes of the domain gap for the MV3D-Det task. Based on the covariate shift
assumption, we find that the gap mainly attributes to the feature distribution
of BEV, which is determined by the quality of both depth estimation and 2D
image's feature representation. To acquire a robust depth prediction, we
propose to decouple the depth estimation from the intrinsic parameters of the
camera (i.e. the focal length) through converting the prediction of metric
depth to that of scale-invariant depth and perform dynamic perspective
augmentation to increase the diversity of the extrinsic parameters (i.e. the
camera poses) by utilizing homography. Moreover, we modify the focal length
values to create multiple pseudo-domains and construct an adversarial training
loss to encourage the feature representation to be more domain-agnostic.
Without bells and whistles, our approach, namely DG-BEV, successfully
alleviates the performance drop on the unseen target domain without impairing
the accuracy of the source domain. Extensive experiments on various public
datasets, including Waymo, nuScenes, and Lyft, demonstrate the generalization
and effectiveness of our approach. To the best of our knowledge, this is the
first systematic study to explore a domain generalization method for MV3D-Det.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 02:59:13 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wang",
"Shuo",
""
],
[
"Zhao",
"Xinhai",
""
],
[
"Xu",
"Hai-Ming",
""
],
[
"Chen",
"Zehui",
""
],
[
"Yu",
"Dameng",
""
],
[
"Chang",
"Jiahao",
""
],
[
"Yang",
"Zhen",
""
],
[
"Zhao",
"Feng",
""
]
] | TITLE: Towards Domain Generalization for Multi-view 3D Object Detection in
Bird-Eye-View
ABSTRACT: Multi-view 3D object detection (MV3D-Det) in Bird-Eye-View (BEV) has drawn
extensive attention due to its low cost and high efficiency. Although new
algorithms for camera-only 3D object detection have been continuously proposed,
most of them may risk drastic performance degradation when the domain of input
images differs from that of training. In this paper, we first analyze the
causes of the domain gap for the MV3D-Det task. Based on the covariate shift
assumption, we find that the gap mainly attributes to the feature distribution
of BEV, which is determined by the quality of both depth estimation and 2D
image's feature representation. To acquire a robust depth prediction, we
propose to decouple the depth estimation from the intrinsic parameters of the
camera (i.e. the focal length) through converting the prediction of metric
depth to that of scale-invariant depth and perform dynamic perspective
augmentation to increase the diversity of the extrinsic parameters (i.e. the
camera poses) by utilizing homography. Moreover, we modify the focal length
values to create multiple pseudo-domains and construct an adversarial training
loss to encourage the feature representation to be more domain-agnostic.
Without bells and whistles, our approach, namely DG-BEV, successfully
alleviates the performance drop on the unseen target domain without impairing
the accuracy of the source domain. Extensive experiments on various public
datasets, including Waymo, nuScenes, and Lyft, demonstrate the generalization
and effectiveness of our approach. To the best of our knowledge, this is the
first systematic study to explore a domain generalization method for MV3D-Det.
|
2303.01694 | Shuaiqi Chen | Shuaiqi Chen, Xiaofen Xing, Weibin Zhang, Weidong Chen, Xiangmin Xu | DWFormer: Dynamic Window transFormer for Speech Emotion Recognition | 4 pages, 5 figures, 3 tables, accepted by 2023 International
Conference on Acoustics, Speech, and Signal Processing (ICASSP2023) | null | null | null | cs.SD cs.CL eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech emotion recognition is crucial to human-computer interaction. The
temporal regions that represent different emotions scatter in different parts
of the speech locally. Moreover, the temporal scales of important information
may vary over a large range within and across speech segments. Although
transformer-based models have made progress in this field, the existing models
could not precisely locate important regions at different temporal scales. To
address the issue, we propose Dynamic Window transFormer (DWFormer), a new
architecture that leverages temporal importance by dynamically splitting
samples into windows. Self-attention mechanism is applied within windows for
capturing temporal important information locally in a fine-grained way.
Cross-window information interaction is also taken into account for global
communication. DWFormer is evaluated on both the IEMOCAP and the MELD datasets.
Experimental results show that the proposed model achieves better performance
than the previous state-of-the-art methods.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 03:26:53 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chen",
"Shuaiqi",
""
],
[
"Xing",
"Xiaofen",
""
],
[
"Zhang",
"Weibin",
""
],
[
"Chen",
"Weidong",
""
],
[
"Xu",
"Xiangmin",
""
]
] | TITLE: DWFormer: Dynamic Window transFormer for Speech Emotion Recognition
ABSTRACT: Speech emotion recognition is crucial to human-computer interaction. The
temporal regions that represent different emotions scatter in different parts
of the speech locally. Moreover, the temporal scales of important information
may vary over a large range within and across speech segments. Although
transformer-based models have made progress in this field, the existing models
could not precisely locate important regions at different temporal scales. To
address the issue, we propose Dynamic Window transFormer (DWFormer), a new
architecture that leverages temporal importance by dynamically splitting
samples into windows. Self-attention mechanism is applied within windows for
capturing temporal important information locally in a fine-grained way.
Cross-window information interaction is also taken into account for global
communication. DWFormer is evaluated on both the IEMOCAP and the MELD datasets.
Experimental results show that the proposed model achieves better performance
than the previous state-of-the-art methods.
|
2303.01695 | Kokila Perera | Kokila Perera and Aneta Neumann and Frank Neumann | Evolutionary Multi-Objective Algorithms for the Knapsack Problems with
Stochastic Profits | 14 pages, 1 figure | null | null | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evolutionary multi-objective algorithms have been widely shown to be
successful when utilized for a variety of stochastic combinatorial optimization
problems. Chance constrained optimization plays an important role in complex
real-world scenarios, as it allows decision makers to take into account the
uncertainty of the environment. We consider a version of the knapsack problem
with stochastic profits to guarantee a certain level of confidence in the
profit of the solutions. We introduce the multi-objective formulations of the
profit chance constrained knapsack problem and design three bi-objective
fitness evaluation methods that work independently of the specific confidence
level required. We evaluate our approaches using well-known multi-objective
evolutionary algorithms GSEMO and NSGA-II. In addition, we introduce a
filtering method for GSEMO that improves the quality of the final population by
periodically removing certain solutions from the interim populations based on
their confidence level. We show the effectiveness of our approaches on several
benchmarks for both settings where the knapsack items have fixed uniform
uncertainties and uncertainties that are positively correlated with the
expected profit of an item.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 03:28:51 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Perera",
"Kokila",
""
],
[
"Neumann",
"Aneta",
""
],
[
"Neumann",
"Frank",
""
]
] | TITLE: Evolutionary Multi-Objective Algorithms for the Knapsack Problems with
Stochastic Profits
ABSTRACT: Evolutionary multi-objective algorithms have been widely shown to be
successful when utilized for a variety of stochastic combinatorial optimization
problems. Chance constrained optimization plays an important role in complex
real-world scenarios, as it allows decision makers to take into account the
uncertainty of the environment. We consider a version of the knapsack problem
with stochastic profits to guarantee a certain level of confidence in the
profit of the solutions. We introduce the multi-objective formulations of the
profit chance constrained knapsack problem and design three bi-objective
fitness evaluation methods that work independently of the specific confidence
level required. We evaluate our approaches using well-known multi-objective
evolutionary algorithms GSEMO and NSGA-II. In addition, we introduce a
filtering method for GSEMO that improves the quality of the final population by
periodically removing certain solutions from the interim populations based on
their confidence level. We show the effectiveness of our approaches on several
benchmarks for both settings where the knapsack items have fixed uniform
uncertainties and uncertainties that are positively correlated with the
expected profit of an item.
|
2303.01707 | Wentao Lei | Wentao Lei, Lei Liu, Li Liu | Spatio-Temporal Structure Consistency for Semi-supervised Medical Image
Classification | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Intelligent medical diagnosis has shown remarkable progress based on the
large-scale datasets with precise annotations. However, fewer labeled images
are available due to significantly expensive cost for annotating data by
experts. To fully exploit the easily available unlabeled data, we propose a
novel Spatio-Temporal Structure Consistent (STSC) learning framework.
Specifically, a gram matrix is derived to combine the spatial structure
consistency and temporal structure consistency together. This gram matrix
captures the structural similarity among the representations of different
training samples. At the spatial level, our framework explicitly enforces the
consistency of structural similarity among different samples under
perturbations. At the temporal level, we consider the consistency of the
structural similarity in different training iterations by digging out the
stable sub-structures in a relation graph. Experiments on two medical image
datasets (i.e., ISIC 2018 challenge and ChestX-ray14) show that our method
outperforms state-of-the-art SSL methods. Furthermore, extensive qualitative
analysis on the Gram matrices and heatmaps by Grad-CAM are presented to
validate the effectiveness of our method.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 04:18:09 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Lei",
"Wentao",
""
],
[
"Liu",
"Lei",
""
],
[
"Liu",
"Li",
""
]
] | TITLE: Spatio-Temporal Structure Consistency for Semi-supervised Medical Image
Classification
ABSTRACT: Intelligent medical diagnosis has shown remarkable progress based on the
large-scale datasets with precise annotations. However, fewer labeled images
are available due to significantly expensive cost for annotating data by
experts. To fully exploit the easily available unlabeled data, we propose a
novel Spatio-Temporal Structure Consistent (STSC) learning framework.
Specifically, a gram matrix is derived to combine the spatial structure
consistency and temporal structure consistency together. This gram matrix
captures the structural similarity among the representations of different
training samples. At the spatial level, our framework explicitly enforces the
consistency of structural similarity among different samples under
perturbations. At the temporal level, we consider the consistency of the
structural similarity in different training iterations by digging out the
stable sub-structures in a relation graph. Experiments on two medical image
datasets (i.e., ISIC 2018 challenge and ChestX-ray14) show that our method
outperforms state-of-the-art SSL methods. Furthermore, extensive qualitative
analysis on the Gram matrices and heatmaps by Grad-CAM are presented to
validate the effectiveness of our method.
|
2303.01742 | Shengfang Zhai | Shengfang Zhai, Qingni Shen, Xiaoyi Chen, Weilong Wang, Cong Li,
Yuejian Fang and Zhonghai Wu | NCL: Textual Backdoor Defense Using Noise-augmented Contrastive Learning | 6 pages, 5 figures. To appear in ICASSP 2023 | null | null | null | cs.CR cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | At present, backdoor attacks attract attention as they do great harm to deep
learning models. The adversary poisons the training data making the model being
injected with a backdoor after being trained unconsciously by victims using the
poisoned dataset. In the field of text, however, existing works do not provide
sufficient defense against backdoor attacks. In this paper, we propose a
Noise-augmented Contrastive Learning (NCL) framework to defend against textual
backdoor attacks when training models with untrustworthy data. With the aim of
mitigating the mapping between triggers and the target label, we add
appropriate noise perturbing possible backdoor triggers, augment the training
dataset, and then pull homology samples in the feature space utilizing
contrastive learning objective. Experiments demonstrate the effectiveness of
our method in defending three types of textual backdoor attacks, outperforming
the prior works.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 07:07:04 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhai",
"Shengfang",
""
],
[
"Shen",
"Qingni",
""
],
[
"Chen",
"Xiaoyi",
""
],
[
"Wang",
"Weilong",
""
],
[
"Li",
"Cong",
""
],
[
"Fang",
"Yuejian",
""
],
[
"Wu",
"Zhonghai",
""
]
] | TITLE: NCL: Textual Backdoor Defense Using Noise-augmented Contrastive Learning
ABSTRACT: At present, backdoor attacks attract attention as they do great harm to deep
learning models. The adversary poisons the training data making the model being
injected with a backdoor after being trained unconsciously by victims using the
poisoned dataset. In the field of text, however, existing works do not provide
sufficient defense against backdoor attacks. In this paper, we propose a
Noise-augmented Contrastive Learning (NCL) framework to defend against textual
backdoor attacks when training models with untrustworthy data. With the aim of
mitigating the mapping between triggers and the target label, we add
appropriate noise perturbing possible backdoor triggers, augment the training
dataset, and then pull homology samples in the feature space utilizing
contrastive learning objective. Experiments demonstrate the effectiveness of
our method in defending three types of textual backdoor attacks, outperforming
the prior works.
|
2303.01774 | David Eriksson | Aryan Deshwal, Sebastian Ament, Maximilian Balandat, Eytan Bakshy,
Janardhan Rao Doppa, David Eriksson | Bayesian Optimization over High-Dimensional Combinatorial Spaces via
Dictionary-based Embeddings | Appearing in AISTATS 2023 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of optimizing expensive black-box functions over
high-dimensional combinatorial spaces which arises in many science,
engineering, and ML applications. We use Bayesian Optimization (BO) and propose
a novel surrogate modeling approach for efficiently handling a large number of
binary and categorical parameters. The key idea is to select a number of
discrete structures from the input space (the dictionary) and use them to
define an ordinal embedding for high-dimensional combinatorial structures. This
allows us to use existing Gaussian process models for continuous spaces. We
develop a principled approach based on binary wavelets to construct
dictionaries for binary spaces, and propose a randomized construction method
that generalizes to categorical spaces. We provide theoretical justification to
support the effectiveness of the dictionary-based embeddings. Our experiments
on diverse real-world benchmarks demonstrate the effectiveness of our proposed
surrogate modeling approach over state-of-the-art BO methods.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 08:31:42 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Deshwal",
"Aryan",
""
],
[
"Ament",
"Sebastian",
""
],
[
"Balandat",
"Maximilian",
""
],
[
"Bakshy",
"Eytan",
""
],
[
"Doppa",
"Janardhan Rao",
""
],
[
"Eriksson",
"David",
""
]
] | TITLE: Bayesian Optimization over High-Dimensional Combinatorial Spaces via
Dictionary-based Embeddings
ABSTRACT: We consider the problem of optimizing expensive black-box functions over
high-dimensional combinatorial spaces which arises in many science,
engineering, and ML applications. We use Bayesian Optimization (BO) and propose
a novel surrogate modeling approach for efficiently handling a large number of
binary and categorical parameters. The key idea is to select a number of
discrete structures from the input space (the dictionary) and use them to
define an ordinal embedding for high-dimensional combinatorial structures. This
allows us to use existing Gaussian process models for continuous spaces. We
develop a principled approach based on binary wavelets to construct
dictionaries for binary spaces, and propose a randomized construction method
that generalizes to categorical spaces. We provide theoretical justification to
support the effectiveness of the dictionary-based embeddings. Our experiments
on diverse real-world benchmarks demonstrate the effectiveness of our proposed
surrogate modeling approach over state-of-the-art BO methods.
|
2303.01778 | Zhenheng Tang | Zhenheng Tang, Xiaowen Chu, Ryan Yide Ran, Sunwoo Lee, Shaohuai Shi,
Yonggang Zhang, Yuxin Wang, Alex Qiaozhong Liang, Salman Avestimehr, Chaoyang
He | FedML Parrot: A Scalable Federated Learning System via
Heterogeneity-aware Scheduling on Sequential and Hierarchical Training | null | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated Learning (FL) enables collaborations among clients for train
machine learning models while protecting their data privacy. Existing FL
simulation platforms that are designed from the perspectives of traditional
distributed training, suffer from laborious code migration between simulation
and production, low efficiency, low GPU utility, low scalability with high
hardware requirements and difficulty of simulating stateful clients. In this
work, we firstly demystify the challenges and bottlenecks of simulating FL, and
design a new FL system named as FedML \texttt{Parrot}. It improves the training
efficiency, remarkably relaxes the requirements on the hardware, and supports
efficient large-scale FL experiments with stateful clients by: (1) sequential
training clients on devices; (2) decomposing original aggregation into local
and global aggregation on devices and server respectively; (3) scheduling tasks
to mitigate straggler problems and enhance computing utility; (4) distributed
client state manager to support various FL algorithms. Besides, built upon our
generic APIs and communication interfaces, users can seamlessly transform the
simulation into the real-world deployment without modifying codes. We evaluate
\texttt{Parrot} through extensive experiments for training diverse models on
various FL datasets to demonstrate that \texttt{Parrot} can achieve simulating
over 1000 clients (stateful or stateless) with flexible GPU devices setting ($4
\sim 32$) and high GPU utility, 1.2 $\sim$ 4 times faster than FedScale, and 10
$\sim$ 100 times memory saving than FedML. And we verify that \texttt{Parrot}
works well with homogeneous and heterogeneous devices in three different
clusters. Two FL algorithms with stateful clients and four algorithms with
stateless clients are simulated to verify the wide adaptability of
\texttt{Parrot} to different algorithms.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 08:36:35 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Tang",
"Zhenheng",
""
],
[
"Chu",
"Xiaowen",
""
],
[
"Ran",
"Ryan Yide",
""
],
[
"Lee",
"Sunwoo",
""
],
[
"Shi",
"Shaohuai",
""
],
[
"Zhang",
"Yonggang",
""
],
[
"Wang",
"Yuxin",
""
],
[
"Liang",
"Alex Qiaozhong",
""
],
[
"Avestimehr",
"Salman",
""
],
[
"He",
"Chaoyang",
""
]
] | TITLE: FedML Parrot: A Scalable Federated Learning System via
Heterogeneity-aware Scheduling on Sequential and Hierarchical Training
ABSTRACT: Federated Learning (FL) enables collaborations among clients for train
machine learning models while protecting their data privacy. Existing FL
simulation platforms that are designed from the perspectives of traditional
distributed training, suffer from laborious code migration between simulation
and production, low efficiency, low GPU utility, low scalability with high
hardware requirements and difficulty of simulating stateful clients. In this
work, we firstly demystify the challenges and bottlenecks of simulating FL, and
design a new FL system named as FedML \texttt{Parrot}. It improves the training
efficiency, remarkably relaxes the requirements on the hardware, and supports
efficient large-scale FL experiments with stateful clients by: (1) sequential
training clients on devices; (2) decomposing original aggregation into local
and global aggregation on devices and server respectively; (3) scheduling tasks
to mitigate straggler problems and enhance computing utility; (4) distributed
client state manager to support various FL algorithms. Besides, built upon our
generic APIs and communication interfaces, users can seamlessly transform the
simulation into the real-world deployment without modifying codes. We evaluate
\texttt{Parrot} through extensive experiments for training diverse models on
various FL datasets to demonstrate that \texttt{Parrot} can achieve simulating
over 1000 clients (stateful or stateless) with flexible GPU devices setting ($4
\sim 32$) and high GPU utility, 1.2 $\sim$ 4 times faster than FedScale, and 10
$\sim$ 100 times memory saving than FedML. And we verify that \texttt{Parrot}
works well with homogeneous and heterogeneous devices in three different
clusters. Two FL algorithms with stateful clients and four algorithms with
stateless clients are simulated to verify the wide adaptability of
\texttt{Parrot} to different algorithms.
|
2303.01781 | Muzhaffar Hazman | Muzhaffar Hazman, Susan McKeever, Josephine Griffith | Meme Sentiment Analysis Enhanced with Multimodal Spatial Encoding and
Facial Embedding | Published as chapter in ISBN:978-3-031-26438-2 | In: Longo, L., OReilly, R. (eds) Artificial Intelligence and
Cognitive Science. AICS 2022. Communications in Computer and Information
Science, vol 1662. Springer, Cham | 10.1007/978-3-031-26438-2_25 | null | cs.CV cs.CL | http://creativecommons.org/licenses/by/4.0/ | Internet memes are characterised by the interspersing of text amongst visual
elements. State-of-the-art multimodal meme classifiers do not account for the
relative positions of these elements across the two modalities, despite the
latent meaning associated with where text and visual elements are placed.
Against two meme sentiment classification datasets, we systematically show
performance gains from incorporating the spatial position of visual objects,
faces, and text clusters extracted from memes. In addition, we also present
facial embedding as an impactful enhancement to image representation in a
multimodal meme classifier. Finally, we show that incorporating this spatial
information allows our fully automated approaches to outperform their
corresponding baselines that rely on additional human validation of
OCR-extracted text.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 08:44:20 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Hazman",
"Muzhaffar",
""
],
[
"McKeever",
"Susan",
""
],
[
"Griffith",
"Josephine",
""
]
] | TITLE: Meme Sentiment Analysis Enhanced with Multimodal Spatial Encoding and
Facial Embedding
ABSTRACT: Internet memes are characterised by the interspersing of text amongst visual
elements. State-of-the-art multimodal meme classifiers do not account for the
relative positions of these elements across the two modalities, despite the
latent meaning associated with where text and visual elements are placed.
Against two meme sentiment classification datasets, we systematically show
performance gains from incorporating the spatial position of visual objects,
faces, and text clusters extracted from memes. In addition, we also present
facial embedding as an impactful enhancement to image representation in a
multimodal meme classifier. Finally, we show that incorporating this spatial
information allows our fully automated approaches to outperform their
corresponding baselines that rely on additional human validation of
OCR-extracted text.
|
2303.01786 | Jiawei He | Jiawei He, Chunyun Fu, Xiyang Wang | 3D Multi-Object Tracking Based on Uncertainty-Guided Data Association | 8 pages, 4 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the existing literature, most 3D multi-object tracking algorithms based on
the tracking-by-detection framework employed deterministic tracks and
detections for similarity calculation in the data association stage. Namely,
the inherent uncertainties existing in tracks and detections are overlooked. In
this work, we discard the commonly used deterministic tracks and deterministic
detections for data association, instead, we propose to model tracks and
detections as random vectors in which uncertainties are taken into account.
Then, based on the Jensen-Shannon divergence, the similarity between two
multidimensional distributions, i.e. track and detection, is evaluated for data
association purposes. Lastly, the level of track uncertainty is incorporated in
our cost function design to guide the data association process. Comparative
experiments have been conducted on two typical datasets, KITTI and nuScenes,
and the results indicated that our proposed method outperformed the compared
state-of-the-art 3D tracking algorithms. For the benefit of the community, our
code has been made available at https://github.com/hejiawei2023/UG3DMOT.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 08:53:00 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"He",
"Jiawei",
""
],
[
"Fu",
"Chunyun",
""
],
[
"Wang",
"Xiyang",
""
]
] | TITLE: 3D Multi-Object Tracking Based on Uncertainty-Guided Data Association
ABSTRACT: In the existing literature, most 3D multi-object tracking algorithms based on
the tracking-by-detection framework employed deterministic tracks and
detections for similarity calculation in the data association stage. Namely,
the inherent uncertainties existing in tracks and detections are overlooked. In
this work, we discard the commonly used deterministic tracks and deterministic
detections for data association, instead, we propose to model tracks and
detections as random vectors in which uncertainties are taken into account.
Then, based on the Jensen-Shannon divergence, the similarity between two
multidimensional distributions, i.e. track and detection, is evaluated for data
association purposes. Lastly, the level of track uncertainty is incorporated in
our cost function design to guide the data association process. Comparative
experiments have been conducted on two typical datasets, KITTI and nuScenes,
and the results indicated that our proposed method outperformed the compared
state-of-the-art 3D tracking algorithms. For the benefit of the community, our
code has been made available at https://github.com/hejiawei2023/UG3DMOT.
|
2303.01788 | Xiwen Liang | Xiwen Liang, Minzhe Niu, Jianhua Han, Hang Xu, Chunjing Xu, Xiaodan
Liang | Visual Exemplar Driven Task-Prompting for Unified Perception in
Autonomous Driving | Accepted at CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-task learning has emerged as a powerful paradigm to solve a range of
tasks simultaneously with good efficiency in both computation resources and
inference time. However, these algorithms are designed for different tasks
mostly not within the scope of autonomous driving, thus making it hard to
compare multi-task methods in autonomous driving. Aiming to enable the
comprehensive evaluation of present multi-task learning methods in autonomous
driving, we extensively investigate the performance of popular multi-task
methods on the large-scale driving dataset, which covers four common perception
tasks, i.e., object detection, semantic segmentation, drivable area
segmentation, and lane detection. We provide an in-depth analysis of current
multi-task learning methods under different common settings and find out that
the existing methods make progress but there is still a large performance gap
compared with single-task baselines. To alleviate this dilemma in autonomous
driving, we present an effective multi-task framework, VE-Prompt, which
introduces visual exemplars via task-specific prompting to guide the model
toward learning high-quality task-specific representations. Specifically, we
generate visual exemplars based on bounding boxes and color-based markers,
which provide accurate visual appearances of target categories and further
mitigate the performance gap. Furthermore, we bridge transformer-based encoders
and convolutional layers for efficient and accurate unified perception in
autonomous driving. Comprehensive experimental results on the diverse
self-driving dataset BDD100K show that the VE-Prompt improves the multi-task
baseline and further surpasses single-task models.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 08:54:06 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Liang",
"Xiwen",
""
],
[
"Niu",
"Minzhe",
""
],
[
"Han",
"Jianhua",
""
],
[
"Xu",
"Hang",
""
],
[
"Xu",
"Chunjing",
""
],
[
"Liang",
"Xiaodan",
""
]
] | TITLE: Visual Exemplar Driven Task-Prompting for Unified Perception in
Autonomous Driving
ABSTRACT: Multi-task learning has emerged as a powerful paradigm to solve a range of
tasks simultaneously with good efficiency in both computation resources and
inference time. However, these algorithms are designed for different tasks
mostly not within the scope of autonomous driving, thus making it hard to
compare multi-task methods in autonomous driving. Aiming to enable the
comprehensive evaluation of present multi-task learning methods in autonomous
driving, we extensively investigate the performance of popular multi-task
methods on the large-scale driving dataset, which covers four common perception
tasks, i.e., object detection, semantic segmentation, drivable area
segmentation, and lane detection. We provide an in-depth analysis of current
multi-task learning methods under different common settings and find out that
the existing methods make progress but there is still a large performance gap
compared with single-task baselines. To alleviate this dilemma in autonomous
driving, we present an effective multi-task framework, VE-Prompt, which
introduces visual exemplars via task-specific prompting to guide the model
toward learning high-quality task-specific representations. Specifically, we
generate visual exemplars based on bounding boxes and color-based markers,
which provide accurate visual appearances of target categories and further
mitigate the performance gap. Furthermore, we bridge transformer-based encoders
and convolutional layers for efficient and accurate unified perception in
autonomous driving. Comprehensive experimental results on the diverse
self-driving dataset BDD100K show that the VE-Prompt improves the multi-task
baseline and further surpasses single-task models.
|
2303.01792 | Neta Rabin | Shir Friedman, Gonen Singer, Neta Rabin | Graph-based Extreme Feature Selection for Multi-class Classification
Tasks | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | When processing high-dimensional datasets, a common pre-processing step is
feature selection. Filter-based feature selection algorithms are not tailored
to a specific classification method, but rather rank the relevance of each
feature with respect to the target and the task. This work focuses on a
graph-based, filter feature selection method that is suited for multi-class
classifications tasks. We aim to drastically reduce the number of selected
features, in order to create a sketch of the original data that codes valuable
information for the classification task. The proposed graph-based algorithm is
constructed by combing the Jeffries-Matusita distance with a non-linear
dimension reduction method, diffusion maps. Feature elimination is performed
based on the distribution of the features in the low-dimensional space. Then, a
very small number of feature that have complementary separation strengths, are
selected. Moreover, the low-dimensional embedding allows to visualize the
feature space. Experimental results are provided for public datasets and
compared with known filter-based feature selection techniques.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 09:06:35 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Friedman",
"Shir",
""
],
[
"Singer",
"Gonen",
""
],
[
"Rabin",
"Neta",
""
]
] | TITLE: Graph-based Extreme Feature Selection for Multi-class Classification
Tasks
ABSTRACT: When processing high-dimensional datasets, a common pre-processing step is
feature selection. Filter-based feature selection algorithms are not tailored
to a specific classification method, but rather rank the relevance of each
feature with respect to the target and the task. This work focuses on a
graph-based, filter feature selection method that is suited for multi-class
classifications tasks. We aim to drastically reduce the number of selected
features, in order to create a sketch of the original data that codes valuable
information for the classification task. The proposed graph-based algorithm is
constructed by combing the Jeffries-Matusita distance with a non-linear
dimension reduction method, diffusion maps. Feature elimination is performed
based on the distribution of the features in the low-dimensional space. Then, a
very small number of feature that have complementary separation strengths, are
selected. Moreover, the low-dimensional embedding allows to visualize the
feature space. Experimental results are provided for public datasets and
compared with known filter-based feature selection techniques.
|
2303.01803 | Huixin Sun | Huixin Sun, Baochang Zhang, Yanjing Li, Xianbin Cao | Confidence-driven Bounding Box Localization for Small Object Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Despite advancements in generic object detection, there remains a performance
gap in detecting small objects compared to normal-scale objects. We for the
first time observe that existing bounding box regression methods tend to
produce distorted gradients for small objects and result in less accurate
localization. To address this issue, we present a novel Confidence-driven
Bounding Box Localization (C-BBL) method to rectify the gradients. C-BBL
quantizes continuous labels into grids and formulates two-hot ground truth
labels. In prediction, the bounding box head generates a confidence
distribution over the grids. Unlike the bounding box regression paradigms in
conventional detectors, we introduce a classification-based localization
objective through cross entropy between ground truth and predicted confidence
distribution, generating confidence-driven gradients. Additionally, C-BBL
describes a uncertainty loss based on distribution entropy in labels and
predictions to further reduce the uncertainty in small object localization. The
method is evaluated on multiple detectors using three object detection
benchmarks and consistently improves baseline detectors, achieving
state-of-the-art performance. We also demonstrate the generalizability of C-BBL
to different label systems and effectiveness for high resolution detection,
which validates its prospect as a general solution.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 09:19:08 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Sun",
"Huixin",
""
],
[
"Zhang",
"Baochang",
""
],
[
"Li",
"Yanjing",
""
],
[
"Cao",
"Xianbin",
""
]
] | TITLE: Confidence-driven Bounding Box Localization for Small Object Detection
ABSTRACT: Despite advancements in generic object detection, there remains a performance
gap in detecting small objects compared to normal-scale objects. We for the
first time observe that existing bounding box regression methods tend to
produce distorted gradients for small objects and result in less accurate
localization. To address this issue, we present a novel Confidence-driven
Bounding Box Localization (C-BBL) method to rectify the gradients. C-BBL
quantizes continuous labels into grids and formulates two-hot ground truth
labels. In prediction, the bounding box head generates a confidence
distribution over the grids. Unlike the bounding box regression paradigms in
conventional detectors, we introduce a classification-based localization
objective through cross entropy between ground truth and predicted confidence
distribution, generating confidence-driven gradients. Additionally, C-BBL
describes a uncertainty loss based on distribution entropy in labels and
predictions to further reduce the uncertainty in small object localization. The
method is evaluated on multiple detectors using three object detection
benchmarks and consistently improves baseline detectors, achieving
state-of-the-art performance. We also demonstrate the generalizability of C-BBL
to different label systems and effectiveness for high resolution detection,
which validates its prospect as a general solution.
|
2303.01804 | Jieqi Shi | Jieqi Shi, Peiliang Li, Xiaozhi Chen and Shaojie Shen | Are All Point Clouds Suitable for Completion? Weakly Supervised Quality
Evaluation Network for Point Cloud Completion | ICRA 2023 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the practical application of point cloud completion tasks, real data
quality is usually much worse than the CAD datasets used for training. A small
amount of noisy data will usually significantly impact the overall system's
accuracy. In this paper, we propose a quality evaluation network to score the
point clouds and help judge the quality of the point cloud before applying the
completion model. We believe our scoring method can help researchers select
more appropriate point clouds for subsequent completion and reconstruction and
avoid manual parameter adjustment. Moreover, our evaluation model is fast and
straightforward and can be directly inserted into any model's training or use
process to facilitate the automatic selection and post-processing of point
clouds. We propose a complete dataset construction and model evaluation method
based on ShapeNet. We verify our network using detection and flow estimation
tasks on KITTI, a real-world dataset for autonomous driving. The experimental
results show that our model can effectively distinguish the quality of point
clouds and help in practical tasks.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 09:24:29 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Shi",
"Jieqi",
""
],
[
"Li",
"Peiliang",
""
],
[
"Chen",
"Xiaozhi",
""
],
[
"Shen",
"Shaojie",
""
]
] | TITLE: Are All Point Clouds Suitable for Completion? Weakly Supervised Quality
Evaluation Network for Point Cloud Completion
ABSTRACT: In the practical application of point cloud completion tasks, real data
quality is usually much worse than the CAD datasets used for training. A small
amount of noisy data will usually significantly impact the overall system's
accuracy. In this paper, we propose a quality evaluation network to score the
point clouds and help judge the quality of the point cloud before applying the
completion model. We believe our scoring method can help researchers select
more appropriate point clouds for subsequent completion and reconstruction and
avoid manual parameter adjustment. Moreover, our evaluation model is fast and
straightforward and can be directly inserted into any model's training or use
process to facilitate the automatic selection and post-processing of point
clouds. We propose a complete dataset construction and model evaluation method
based on ShapeNet. We verify our network using detection and flow estimation
tasks on KITTI, a real-world dataset for autonomous driving. The experimental
results show that our model can effectively distinguish the quality of point
clouds and help in practical tasks.
|
2303.01812 | Heinrich Dinkel | Heinrich Dinkel, Yongqing Wang, Zhiyong Yan, Junbo Zhang and Yujun
Wang | Unified Keyword Spotting and Audio Tagging on Mobile Devices with
Transformers | ICASSP 2023 | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Keyword spotting (KWS) is a core human-machine-interaction front-end task for
most modern intelligent assistants. Recently, a unified (UniKW-AT) framework
has been proposed that adds additional capabilities in the form of audio
tagging (AT) to a KWS model. However, previous work did not consider the
real-world deployment of a UniKW-AT model, where factors such as model size and
inference speed are more important than performance alone. This work introduces
three mobile-device deployable models named Unified Transformers (UiT). Our
best model achieves an mAP of 34.09 on Audioset, and an accuracy of 97.76 on
the public Google Speech Commands V1 dataset. Further, we benchmark our
proposed approaches on four mobile platforms, revealing that the proposed UiT
models can achieve a speedup of 2 - 6 times against a competitive MobileNetV2.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 09:38:53 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Dinkel",
"Heinrich",
""
],
[
"Wang",
"Yongqing",
""
],
[
"Yan",
"Zhiyong",
""
],
[
"Zhang",
"Junbo",
""
],
[
"Wang",
"Yujun",
""
]
] | TITLE: Unified Keyword Spotting and Audio Tagging on Mobile Devices with
Transformers
ABSTRACT: Keyword spotting (KWS) is a core human-machine-interaction front-end task for
most modern intelligent assistants. Recently, a unified (UniKW-AT) framework
has been proposed that adds additional capabilities in the form of audio
tagging (AT) to a KWS model. However, previous work did not consider the
real-world deployment of a UniKW-AT model, where factors such as model size and
inference speed are more important than performance alone. This work introduces
three mobile-device deployable models named Unified Transformers (UiT). Our
best model achieves an mAP of 34.09 on Audioset, and an accuracy of 97.76 on
the public Google Speech Commands V1 dataset. Further, we benchmark our
proposed approaches on four mobile platforms, revealing that the proposed UiT
models can achieve a speedup of 2 - 6 times against a competitive MobileNetV2.
|
2303.01844 | Caglar Demir | Caglar Demir, Axel-Cyrille Ngonga Ngomo | Learning Permutation-Invariant Embeddings for Description Logic Concepts | Accepted at IDA 2023 | null | null | null | cs.LO cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Concept learning deals with learning description logic concepts from a
background knowledge and input examples. The goal is to learn a concept that
covers all positive examples, while not covering any negative examples. This
non-trivial task is often formulated as a search problem within an infinite
quasi-ordered concept space. Although state-of-the-art models have been
successfully applied to tackle this problem, their large-scale applications
have been severely hindered due to their excessive exploration incurring
impractical runtimes. Here, we propose a remedy for this limitation. We
reformulate the learning problem as a multi-label classification problem and
propose a neural embedding model (NERO) that learns permutation-invariant
embeddings for sets of examples tailored towards predicting $F_1$ scores of
pre-selected description logic concepts. By ranking such concepts in descending
order of predicted scores, a possible goal concept can be detected within few
retrieval operations, i.e., no excessive exploration. Importantly, top-ranked
concepts can be used to start the search procedure of state-of-the-art symbolic
models in multiple advantageous regions of a concept space, rather than
starting it in the most general concept $\top$. Our experiments on 5 benchmark
datasets with 770 learning problems firmly suggest that NERO significantly
(p-value <1%) outperforms the state-of-the-art models in terms of $F_1$ score,
the number of explored concepts, and the total runtime. We provide an
open-source implementation of our approach.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 10:57:04 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Demir",
"Caglar",
""
],
[
"Ngomo",
"Axel-Cyrille Ngonga",
""
]
] | TITLE: Learning Permutation-Invariant Embeddings for Description Logic Concepts
ABSTRACT: Concept learning deals with learning description logic concepts from a
background knowledge and input examples. The goal is to learn a concept that
covers all positive examples, while not covering any negative examples. This
non-trivial task is often formulated as a search problem within an infinite
quasi-ordered concept space. Although state-of-the-art models have been
successfully applied to tackle this problem, their large-scale applications
have been severely hindered due to their excessive exploration incurring
impractical runtimes. Here, we propose a remedy for this limitation. We
reformulate the learning problem as a multi-label classification problem and
propose a neural embedding model (NERO) that learns permutation-invariant
embeddings for sets of examples tailored towards predicting $F_1$ scores of
pre-selected description logic concepts. By ranking such concepts in descending
order of predicted scores, a possible goal concept can be detected within few
retrieval operations, i.e., no excessive exploration. Importantly, top-ranked
concepts can be used to start the search procedure of state-of-the-art symbolic
models in multiple advantageous regions of a concept space, rather than
starting it in the most general concept $\top$. Our experiments on 5 benchmark
datasets with 770 learning problems firmly suggest that NERO significantly
(p-value <1%) outperforms the state-of-the-art models in terms of $F_1$ score,
the number of explored concepts, and the total runtime. We provide an
open-source implementation of our approach.
|
2303.01845 | Aydin Buluc | Oguz Selvitopi, Saliya Ekanayake, Giulia Guidi, Muaaz G. Awan,
Georgios A. Pavlopoulos, Ariful Azad, Nikos Kyrpides, Leonid Oliker,
Katherine Yelick, Ayd{\i}n Bulu\c{c} | Extreme-scale many-against-many protein similarity search | 2022 ACM Gordon Bell Prize Finalist | SC'22: Proceedings of the International Conference on High
Performance Computing, Networking, Storage and Analysis, 2022 | 10.5555/3571885.3571887 | null | cs.DC cs.PF q-bio.GN | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Similarity search is one of the most fundamental computations that are
regularly performed on ever-increasing protein datasets. Scalability is of
paramount importance for uncovering novel phenomena that occur at very large
scales. We unleash the power of over 20,000 GPUs on the Summit system to
perform all-vs-all protein similarity search on one of the largest publicly
available datasets with 405 million proteins, in less than 3.5 hours, cutting
the time-to-solution for many use cases from weeks. The variability of protein
sequence lengths, as well as the sparsity of the space of pairwise comparisons,
make this a challenging problem in distributed memory. Due to the need to
construct and maintain a data structure holding indices to all other sequences,
this application has a huge memory footprint that makes it hard to scale the
problem sizes. We overcome this memory limitation by innovative matrix-based
blocking techniques, without introducing additional load imbalance.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 10:57:09 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Selvitopi",
"Oguz",
""
],
[
"Ekanayake",
"Saliya",
""
],
[
"Guidi",
"Giulia",
""
],
[
"Awan",
"Muaaz G.",
""
],
[
"Pavlopoulos",
"Georgios A.",
""
],
[
"Azad",
"Ariful",
""
],
[
"Kyrpides",
"Nikos",
""
],
[
"Oliker",
"Leonid",
""
],
[
"Yelick",
"Katherine",
""
],
[
"Buluç",
"Aydın",
""
]
] | TITLE: Extreme-scale many-against-many protein similarity search
ABSTRACT: Similarity search is one of the most fundamental computations that are
regularly performed on ever-increasing protein datasets. Scalability is of
paramount importance for uncovering novel phenomena that occur at very large
scales. We unleash the power of over 20,000 GPUs on the Summit system to
perform all-vs-all protein similarity search on one of the largest publicly
available datasets with 405 million proteins, in less than 3.5 hours, cutting
the time-to-solution for many use cases from weeks. The variability of protein
sequence lengths, as well as the sparsity of the space of pairwise comparisons,
make this a challenging problem in distributed memory. Due to the need to
construct and maintain a data structure holding indices to all other sequences,
this application has a huge memory footprint that makes it hard to scale the
problem sizes. We overcome this memory limitation by innovative matrix-based
blocking techniques, without introducing additional load imbalance.
|
2303.01850 | Malihe Alavi | Malihe Alavi, Farnoush Manavi, Amirhossein Ansari, Ali Hamzeh | LBCIM: Loyalty Based Competitive Influence Maximization with
epsilon-greedy MCTS strategy | 13 pages, 10 figures, 2 pseudo code | null | null | null | cs.SI cs.AI | http://creativecommons.org/licenses/by/4.0/ | Competitive influence maximization has been studied for several years, and
various frameworks have been proposed to model different aspects of information
diffusion under the competitive environment. This work presents a new gameboard
for two competing parties with some new features representing loyalty in social
networks and reflecting the attitude of not completely being loyal to a party
when the opponent offers better suggestions. This behavior can be observed in
most political occasions where each party tries to attract people by making
better suggestions than the opponent and even seeks to impress the fans of the
opposition party to change their minds. In order to identify the best move in
each step of the game framework, an improved Monte Carlo tree search is
developed, which uses some predefined heuristics to apply them on the
simulation step of the algorithm and takes advantage of them to search among
child nodes of the current state and pick the best one using an epsilon-greedy
way instead of choosing them at random. Experimental results on synthetic and
real datasets indicate the outperforming of the proposed strategy against some
well-known and benchmark strategies like general MCTS, minimax algorithm with
alpha-beta pruning, random nodes, nodes with maximum threshold and nodes with
minimum threshold.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 11:11:53 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Alavi",
"Malihe",
""
],
[
"Manavi",
"Farnoush",
""
],
[
"Ansari",
"Amirhossein",
""
],
[
"Hamzeh",
"Ali",
""
]
] | TITLE: LBCIM: Loyalty Based Competitive Influence Maximization with
epsilon-greedy MCTS strategy
ABSTRACT: Competitive influence maximization has been studied for several years, and
various frameworks have been proposed to model different aspects of information
diffusion under the competitive environment. This work presents a new gameboard
for two competing parties with some new features representing loyalty in social
networks and reflecting the attitude of not completely being loyal to a party
when the opponent offers better suggestions. This behavior can be observed in
most political occasions where each party tries to attract people by making
better suggestions than the opponent and even seeks to impress the fans of the
opposition party to change their minds. In order to identify the best move in
each step of the game framework, an improved Monte Carlo tree search is
developed, which uses some predefined heuristics to apply them on the
simulation step of the algorithm and takes advantage of them to search among
child nodes of the current state and pick the best one using an epsilon-greedy
way instead of choosing them at random. Experimental results on synthetic and
real datasets indicate the outperforming of the proposed strategy against some
well-known and benchmark strategies like general MCTS, minimax algorithm with
alpha-beta pruning, random nodes, nodes with maximum threshold and nodes with
minimum threshold.
|
2303.01859 | Steven Morad | Steven Morad, Ryan Kortvelesy, Matteo Bettini, Stephan Liwicki, and
Amanda Prorok | POPGym: Benchmarking Partially Observable Reinforcement Learning | null | null | null | null | cs.LG cs.AI cs.RO | http://creativecommons.org/licenses/by/4.0/ | Real world applications of Reinforcement Learning (RL) are often partially
observable, thus requiring memory. Despite this, partial observability is still
largely ignored by contemporary RL benchmarks and libraries. We introduce
Partially Observable Process Gym (POPGym), a two-part library containing (1) a
diverse collection of 15 partially observable environments, each with multiple
difficulties and (2) implementations of 13 memory model baselines -- the most
in a single RL library. Existing partially observable benchmarks tend to fixate
on 3D visual navigation, which is computationally expensive and only one type
of POMDP. In contrast, POPGym environments are diverse, produce smaller
observations, use less memory, and often converge within two hours of training
on a consumer-grade GPU. We implement our high-level memory API and memory
baselines on top of the popular RLlib framework, providing plug-and-play
compatibility with various training algorithms, exploration strategies, and
distributed training paradigms. Using POPGym, we execute the largest comparison
across RL memory models to date. POPGym is available at
https://github.com/proroklab/popgym.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 11:25:33 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Morad",
"Steven",
""
],
[
"Kortvelesy",
"Ryan",
""
],
[
"Bettini",
"Matteo",
""
],
[
"Liwicki",
"Stephan",
""
],
[
"Prorok",
"Amanda",
""
]
] | TITLE: POPGym: Benchmarking Partially Observable Reinforcement Learning
ABSTRACT: Real world applications of Reinforcement Learning (RL) are often partially
observable, thus requiring memory. Despite this, partial observability is still
largely ignored by contemporary RL benchmarks and libraries. We introduce
Partially Observable Process Gym (POPGym), a two-part library containing (1) a
diverse collection of 15 partially observable environments, each with multiple
difficulties and (2) implementations of 13 memory model baselines -- the most
in a single RL library. Existing partially observable benchmarks tend to fixate
on 3D visual navigation, which is computationally expensive and only one type
of POMDP. In contrast, POPGym environments are diverse, produce smaller
observations, use less memory, and often converge within two hours of training
on a consumer-grade GPU. We implement our high-level memory API and memory
baselines on top of the popular RLlib framework, providing plug-and-play
compatibility with various training algorithms, exploration strategies, and
distributed training paradigms. Using POPGym, we execute the largest comparison
across RL memory models to date. POPGym is available at
https://github.com/proroklab/popgym.
|
2303.01876 | Hyungtae Lim | Hyungtae Lim, Kawon Han, Gunhee Shin, Giseop Kim, Songcheol Hong, Hyun
Myung | ORORA: Outlier-Robust Radar Odometry | null | null | null | null | cs.RO | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Radar sensors are emerging as solutions for perceiving surroundings and
estimating ego-motion in extreme weather conditions. Unfortunately, radar
measurements are noisy and suffer from mutual interference, which degrades the
performance of feature extraction and matching, triggering imprecise matching
pairs, which are referred to as outliers. To tackle the effect of outliers on
radar odometry, a novel outlier-robust method called \textit{ORORA} is
proposed, which is an abbreviation of \textit{Outlier-RObust RAdar odometry}.
To this end, a novel decoupling-based method is proposed, which consists of
graduated non-convexity~(GNC)-based rotation estimation and anisotropic
component-wise translation estimation~(A-COTE). Furthermore, our method
leverages the anisotropic characteristics of radar measurements, each of whose
uncertainty along the azimuthal direction is somewhat larger than that along
the radial direction. As verified in the public dataset, it was demonstrated
that our proposed method yields robust ego-motion estimation performance
compared with other state-of-the-art methods. Our code is available at
https://github.com/url-kaist/outlier-robust-radar-odometry.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 12:14:03 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Lim",
"Hyungtae",
""
],
[
"Han",
"Kawon",
""
],
[
"Shin",
"Gunhee",
""
],
[
"Kim",
"Giseop",
""
],
[
"Hong",
"Songcheol",
""
],
[
"Myung",
"Hyun",
""
]
] | TITLE: ORORA: Outlier-Robust Radar Odometry
ABSTRACT: Radar sensors are emerging as solutions for perceiving surroundings and
estimating ego-motion in extreme weather conditions. Unfortunately, radar
measurements are noisy and suffer from mutual interference, which degrades the
performance of feature extraction and matching, triggering imprecise matching
pairs, which are referred to as outliers. To tackle the effect of outliers on
radar odometry, a novel outlier-robust method called \textit{ORORA} is
proposed, which is an abbreviation of \textit{Outlier-RObust RAdar odometry}.
To this end, a novel decoupling-based method is proposed, which consists of
graduated non-convexity~(GNC)-based rotation estimation and anisotropic
component-wise translation estimation~(A-COTE). Furthermore, our method
leverages the anisotropic characteristics of radar measurements, each of whose
uncertainty along the azimuthal direction is somewhat larger than that along
the radial direction. As verified in the public dataset, it was demonstrated
that our proposed method yields robust ego-motion estimation performance
compared with other state-of-the-art methods. Our code is available at
https://github.com/url-kaist/outlier-robust-radar-odometry.
|
2303.01884 | Sen Pei | Sen Pei, Jingya Yu, Qi Chen, Wozhou He | AutoMatch: A Large-scale Audio Beat Matching Benchmark for Boosting Deep
Learning Assistant Video Editing | 11 pages, 5 figures | null | null | null | cs.SD cs.CV cs.MM eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The explosion of short videos has dramatically reshaped the manners people
socialize, yielding a new trend for daily sharing and access to the latest
information. These rich video resources, on the one hand, benefited from the
popularization of portable devices with cameras, but on the other, they can not
be independent of the valuable editing work contributed by numerous video
creators. In this paper, we investigate a novel and practical problem, namely
audio beat matching (ABM), which aims to recommend the proper transition time
stamps based on the background music. This technique helps to ease the
labor-intensive work during video editing, saving energy for creators so that
they can focus more on the creativity of video content. We formally define the
ABM problem and its evaluation protocol. Meanwhile, a large-scale audio
dataset, i.e., the AutoMatch with over 87k finely annotated background music,
is presented to facilitate this newly opened research direction. To further lay
solid foundations for the following study, we also propose a novel model termed
BeatX to tackle this challenging task. Alongside, we creatively present the
concept of label scope, which eliminates the data imbalance issues and assigns
adaptive weights for the ground truth during the training procedure in one
stop. Though plentiful short video platforms have flourished for a long time,
the relevant research concerning this scenario is not sufficient, and to the
best of our knowledge, AutoMatch is the first large-scale dataset to tackle the
audio beat matching problem. We hope the released dataset and our competitive
baseline can encourage more attention to this line of research. The dataset and
codes will be made publicly available.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 12:30:09 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Pei",
"Sen",
""
],
[
"Yu",
"Jingya",
""
],
[
"Chen",
"Qi",
""
],
[
"He",
"Wozhou",
""
]
] | TITLE: AutoMatch: A Large-scale Audio Beat Matching Benchmark for Boosting Deep
Learning Assistant Video Editing
ABSTRACT: The explosion of short videos has dramatically reshaped the manners people
socialize, yielding a new trend for daily sharing and access to the latest
information. These rich video resources, on the one hand, benefited from the
popularization of portable devices with cameras, but on the other, they can not
be independent of the valuable editing work contributed by numerous video
creators. In this paper, we investigate a novel and practical problem, namely
audio beat matching (ABM), which aims to recommend the proper transition time
stamps based on the background music. This technique helps to ease the
labor-intensive work during video editing, saving energy for creators so that
they can focus more on the creativity of video content. We formally define the
ABM problem and its evaluation protocol. Meanwhile, a large-scale audio
dataset, i.e., the AutoMatch with over 87k finely annotated background music,
is presented to facilitate this newly opened research direction. To further lay
solid foundations for the following study, we also propose a novel model termed
BeatX to tackle this challenging task. Alongside, we creatively present the
concept of label scope, which eliminates the data imbalance issues and assigns
adaptive weights for the ground truth during the training procedure in one
stop. Though plentiful short video platforms have flourished for a long time,
the relevant research concerning this scenario is not sufficient, and to the
best of our knowledge, AutoMatch is the first large-scale dataset to tackle the
audio beat matching problem. We hope the released dataset and our competitive
baseline can encourage more attention to this line of research. The dataset and
codes will be made publicly available.
|
2303.01899 | Sebastian Huch | Sebastian Huch, Luca Scalerandi, Esteban Rivera, Markus Lienkamp | Quantifying the LiDAR Sim-to-Real Domain Shift: A Detailed Investigation
Using Object Detectors and Analyzing Point Clouds at Target-Level | null | IEEE TRANSACTIONS ON INTELLIGENT VEHICLES, 2023 | 10.1109/TIV.2023.3251650 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | LiDAR object detection algorithms based on neural networks for autonomous
driving require large amounts of data for training, validation, and testing. As
real-world data collection and labeling are time-consuming and expensive,
simulation-based synthetic data generation is a viable alternative. However,
using simulated data for the training of neural networks leads to a domain
shift of training and testing data due to differences in scenes, scenarios, and
distributions. In this work, we quantify the sim-to-real domain shift by means
of LiDAR object detectors trained with a new scenario-identical real-world and
simulated dataset. In addition, we answer the questions of how well the
simulated data resembles the real-world data and how well object detectors
trained on simulated data perform on real-world data. Further, we analyze point
clouds at the target-level by comparing real-world and simulated point clouds
within the 3D bounding boxes of the targets. Our experiments show that a
significant sim-to-real domain shift exists even for our scenario-identical
datasets. This domain shift amounts to an average precision reduction of around
14 % for object detectors trained with simulated data. Additional experiments
reveal that this domain shift can be lowered by introducing a simple noise
model in simulation. We further show that a simple downsampling method to model
real-world physics does not influence the performance of the object detectors.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 12:52:01 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Huch",
"Sebastian",
""
],
[
"Scalerandi",
"Luca",
""
],
[
"Rivera",
"Esteban",
""
],
[
"Lienkamp",
"Markus",
""
]
] | TITLE: Quantifying the LiDAR Sim-to-Real Domain Shift: A Detailed Investigation
Using Object Detectors and Analyzing Point Clouds at Target-Level
ABSTRACT: LiDAR object detection algorithms based on neural networks for autonomous
driving require large amounts of data for training, validation, and testing. As
real-world data collection and labeling are time-consuming and expensive,
simulation-based synthetic data generation is a viable alternative. However,
using simulated data for the training of neural networks leads to a domain
shift of training and testing data due to differences in scenes, scenarios, and
distributions. In this work, we quantify the sim-to-real domain shift by means
of LiDAR object detectors trained with a new scenario-identical real-world and
simulated dataset. In addition, we answer the questions of how well the
simulated data resembles the real-world data and how well object detectors
trained on simulated data perform on real-world data. Further, we analyze point
clouds at the target-level by comparing real-world and simulated point clouds
within the 3D bounding boxes of the targets. Our experiments show that a
significant sim-to-real domain shift exists even for our scenario-identical
datasets. This domain shift amounts to an average precision reduction of around
14 % for object detectors trained with simulated data. Additional experiments
reveal that this domain shift can be lowered by introducing a simple noise
model in simulation. We further show that a simple downsampling method to model
real-world physics does not influence the performance of the object detectors.
|
2303.01906 | Liwei Yang | Liwei Yang, Xiang Gu, Jian Sun | Generalized Semantic Segmentation by Self-Supervised Source Domain
Projection and Multi-Level Contrastive Learning | 13 pages, 9 figures, accepted at AAAI 2023 (Oral Presentation) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep networks trained on the source domain show degraded performance when
tested on unseen target domain data. To enhance the model's generalization
ability, most existing domain generalization methods learn domain invariant
features by suppressing domain sensitive features. Different from them, we
propose a Domain Projection and Contrastive Learning (DPCL) approach for
generalized semantic segmentation, which includes two modules: Self-supervised
Source Domain Projection (SSDP) and Multi-level Contrastive Learning (MLCL).
SSDP aims to reduce domain gap by projecting data to the source domain, while
MLCL is a learning scheme to learn discriminative and generalizable features on
the projected data. During test time, we first project the target data by SSDP
to mitigate domain shift, then generate the segmentation results by the learned
segmentation network based on MLCL. At test time, we can update the projected
data by minimizing our proposed pixel-to-pixel contrastive loss to obtain
better results. Extensive experiments for semantic segmentation demonstrate the
favorable generalization capability of our method on benchmark datasets.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 13:07:14 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Yang",
"Liwei",
""
],
[
"Gu",
"Xiang",
""
],
[
"Sun",
"Jian",
""
]
] | TITLE: Generalized Semantic Segmentation by Self-Supervised Source Domain
Projection and Multi-Level Contrastive Learning
ABSTRACT: Deep networks trained on the source domain show degraded performance when
tested on unseen target domain data. To enhance the model's generalization
ability, most existing domain generalization methods learn domain invariant
features by suppressing domain sensitive features. Different from them, we
propose a Domain Projection and Contrastive Learning (DPCL) approach for
generalized semantic segmentation, which includes two modules: Self-supervised
Source Domain Projection (SSDP) and Multi-level Contrastive Learning (MLCL).
SSDP aims to reduce domain gap by projecting data to the source domain, while
MLCL is a learning scheme to learn discriminative and generalizable features on
the projected data. During test time, we first project the target data by SSDP
to mitigate domain shift, then generate the segmentation results by the learned
segmentation network based on MLCL. At test time, we can update the projected
data by minimizing our proposed pixel-to-pixel contrastive loss to obtain
better results. Extensive experiments for semantic segmentation demonstrate the
favorable generalization capability of our method on benchmark datasets.
|
2303.01920 | Philip M\"uller | Felix Meissen, Philip M\"uller, Georgios Kaissis, Daniel Rueckert | Robust Detection Outcome: A Metric for Pathology Detection in Medical
Images | Accepted at MIDL 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Detection of pathologies is a fundamental task in medical imaging and the
evaluation of algorithms that can perform this task automatically is crucial.
However, current object detection metrics for natural images do not reflect the
specific clinical requirements in pathology detection sufficiently. To tackle
this problem, we propose Robust Detection Outcome (RoDeO); a novel metric for
evaluating algorithms for pathology detection in medical images, especially in
chest X-rays. RoDeO evaluates different errors directly and individually, and
reflects clinical needs better than current metrics. Extensive evaluation on
the ChestX-ray8 dataset shows the superiority of our metrics compared to
existing ones. We released the code at https://github.com/FeliMe/RoDeO and
published RoDeO as pip package (rodeometric).
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 13:45:13 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Meissen",
"Felix",
""
],
[
"Müller",
"Philip",
""
],
[
"Kaissis",
"Georgios",
""
],
[
"Rueckert",
"Daniel",
""
]
] | TITLE: Robust Detection Outcome: A Metric for Pathology Detection in Medical
Images
ABSTRACT: Detection of pathologies is a fundamental task in medical imaging and the
evaluation of algorithms that can perform this task automatically is crucial.
However, current object detection metrics for natural images do not reflect the
specific clinical requirements in pathology detection sufficiently. To tackle
this problem, we propose Robust Detection Outcome (RoDeO); a novel metric for
evaluating algorithms for pathology detection in medical images, especially in
chest X-rays. RoDeO evaluates different errors directly and individually, and
reflects clinical needs better than current metrics. Extensive evaluation on
the ChestX-ray8 dataset shows the superiority of our metrics compared to
existing ones. We released the code at https://github.com/FeliMe/RoDeO and
published RoDeO as pip package (rodeometric).
|
2303.01939 | Alnur Alimanov | Alnur Alimanov and Md Baharul Islam | Retinal Image Restoration using Transformer and Cycle-Consistent
Generative Adversarial Network | 4 pages, 3 figures, The International Symposium on Intelligent Signal
Processing and Communication Systems (ISPACS2022) | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Medical imaging plays a significant role in detecting and treating various
diseases. However, these images often happen to be of too poor quality, leading
to decreased efficiency, extra expenses, and even incorrect diagnoses.
Therefore, we propose a retinal image enhancement method using a vision
transformer and convolutional neural network. It builds a cycle-consistent
generative adversarial network that relies on unpaired datasets. It consists of
two generators that translate images from one domain to another (e.g., low- to
high-quality and vice versa), playing an adversarial game with two
discriminators. Generators produce indistinguishable images for discriminators
that predict the original images from generated ones. Generators are a
combination of vision transformer (ViT) encoder and convolutional neural
network (CNN) decoder. Discriminators include traditional CNN encoders. The
resulting improved images have been tested quantitatively using such evaluation
metrics as peak signal-to-noise ratio (PSNR), structural similarity index
measure (SSIM), and qualitatively, i.e., vessel segmentation. The proposed
method successfully reduces the adverse effects of blurring, noise,
illumination disturbances, and color distortions while significantly preserving
structural and color information. Experimental results show the superiority of
the proposed method. Our testing PSNR is 31.138 dB for the first and 27.798 dB
for the second dataset. Testing SSIM is 0.919 and 0.904, respectively.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 14:10:47 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Alimanov",
"Alnur",
""
],
[
"Islam",
"Md Baharul",
""
]
] | TITLE: Retinal Image Restoration using Transformer and Cycle-Consistent
Generative Adversarial Network
ABSTRACT: Medical imaging plays a significant role in detecting and treating various
diseases. However, these images often happen to be of too poor quality, leading
to decreased efficiency, extra expenses, and even incorrect diagnoses.
Therefore, we propose a retinal image enhancement method using a vision
transformer and convolutional neural network. It builds a cycle-consistent
generative adversarial network that relies on unpaired datasets. It consists of
two generators that translate images from one domain to another (e.g., low- to
high-quality and vice versa), playing an adversarial game with two
discriminators. Generators produce indistinguishable images for discriminators
that predict the original images from generated ones. Generators are a
combination of vision transformer (ViT) encoder and convolutional neural
network (CNN) decoder. Discriminators include traditional CNN encoders. The
resulting improved images have been tested quantitatively using such evaluation
metrics as peak signal-to-noise ratio (PSNR), structural similarity index
measure (SSIM), and qualitatively, i.e., vessel segmentation. The proposed
method successfully reduces the adverse effects of blurring, noise,
illumination disturbances, and color distortions while significantly preserving
structural and color information. Experimental results show the superiority of
the proposed method. Our testing PSNR is 31.138 dB for the first and 27.798 dB
for the second dataset. Testing SSIM is 0.919 and 0.904, respectively.
|
2303.01943 | Lukas Mehl | Lukas Mehl, Jenny Schmalfuss, Azin Jahedi, Yaroslava Nalivayko,
Andr\'es Bruhn | Spring: A High-Resolution High-Detail Dataset and Benchmark for Scene
Flow, Optical Flow and Stereo | CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While recent methods for motion and stereo estimation recover an
unprecedented amount of details, such highly detailed structures are neither
adequately reflected in the data of existing benchmarks nor their evaluation
methodology. Hence, we introduce Spring $-$ a large, high-resolution,
high-detail, computer-generated benchmark for scene flow, optical flow, and
stereo. Based on rendered scenes from the open-source Blender movie "Spring",
it provides photo-realistic HD datasets with state-of-the-art visual effects
and ground truth training data. Furthermore, we provide a website to upload,
analyze and compare results. Using a novel evaluation methodology based on a
super-resolved UHD ground truth, our Spring benchmark can assess the quality of
fine structures and provides further detailed performance statistics on
different image regions. Regarding the number of ground truth frames, Spring is
60$\times$ larger than the only scene flow benchmark, KITTI 2015, and
15$\times$ larger than the well-established MPI Sintel optical flow benchmark.
Initial results for recent methods on our benchmark show that estimating fine
details is indeed challenging, as their accuracy leaves significant room for
improvement. The Spring benchmark and the corresponding datasets are available
at http://spring-benchmark.org.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 14:15:48 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Mehl",
"Lukas",
""
],
[
"Schmalfuss",
"Jenny",
""
],
[
"Jahedi",
"Azin",
""
],
[
"Nalivayko",
"Yaroslava",
""
],
[
"Bruhn",
"Andrés",
""
]
] | TITLE: Spring: A High-Resolution High-Detail Dataset and Benchmark for Scene
Flow, Optical Flow and Stereo
ABSTRACT: While recent methods for motion and stereo estimation recover an
unprecedented amount of details, such highly detailed structures are neither
adequately reflected in the data of existing benchmarks nor their evaluation
methodology. Hence, we introduce Spring $-$ a large, high-resolution,
high-detail, computer-generated benchmark for scene flow, optical flow, and
stereo. Based on rendered scenes from the open-source Blender movie "Spring",
it provides photo-realistic HD datasets with state-of-the-art visual effects
and ground truth training data. Furthermore, we provide a website to upload,
analyze and compare results. Using a novel evaluation methodology based on a
super-resolved UHD ground truth, our Spring benchmark can assess the quality of
fine structures and provides further detailed performance statistics on
different image regions. Regarding the number of ground truth frames, Spring is
60$\times$ larger than the only scene flow benchmark, KITTI 2015, and
15$\times$ larger than the well-established MPI Sintel optical flow benchmark.
Initial results for recent methods on our benchmark show that estimating fine
details is indeed challenging, as their accuracy leaves significant room for
improvement. The Spring benchmark and the corresponding datasets are available
at http://spring-benchmark.org.
|
2303.01949 | Magda Bucholc | Magda Bucholc, Charlotte James, Ahmad Al Khleifat, AmanPreet Badhwar,
Natasha Clarke, Amir Dehsarvi, Christopher R. Madan, Sarah J. Marzi, Cameron
Shand, Brian M. Schilder, Stefano Tamburin, Hanz M. Tantiangco, Ilianna
Lourida, David J. Llewellyn, Janice M. Ranson | Artificial Intelligence for Dementia Research Methods Optimization | Magda Bucholc and Charlotte James joint first authors | null | null | null | cs.LG q-bio.QM stat.AP stat.ME | http://creativecommons.org/licenses/by/4.0/ | Introduction: Machine learning (ML) has been extremely successful in
identifying key features from high-dimensional datasets and executing
complicated tasks with human expert levels of accuracy or greater. Methods: We
summarize and critically evaluate current applications of ML in dementia
research and highlight directions for future research. Results: We present an
overview of ML algorithms most frequently used in dementia research and
highlight future opportunities for the use of ML in clinical practice,
experimental medicine, and clinical trials. We discuss issues of
reproducibility, replicability and interpretability and how these impact the
clinical applicability of dementia research. Finally, we give examples of how
state-of-the-art methods, such as transfer learning, multi-task learning, and
reinforcement learning, may be applied to overcome these issues and aid the
translation of research to clinical practice in the future. Discussion:
ML-based models hold great promise to advance our understanding of the
underlying causes and pathological mechanisms of dementia.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 08:50:25 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Bucholc",
"Magda",
""
],
[
"James",
"Charlotte",
""
],
[
"Khleifat",
"Ahmad Al",
""
],
[
"Badhwar",
"AmanPreet",
""
],
[
"Clarke",
"Natasha",
""
],
[
"Dehsarvi",
"Amir",
""
],
[
"Madan",
"Christopher R.",
""
],
[
"Marzi",
"Sarah J.",
""
],
[
"Shand",
"Cameron",
""
],
[
"Schilder",
"Brian M.",
""
],
[
"Tamburin",
"Stefano",
""
],
[
"Tantiangco",
"Hanz M.",
""
],
[
"Lourida",
"Ilianna",
""
],
[
"Llewellyn",
"David J.",
""
],
[
"Ranson",
"Janice M.",
""
]
] | TITLE: Artificial Intelligence for Dementia Research Methods Optimization
ABSTRACT: Introduction: Machine learning (ML) has been extremely successful in
identifying key features from high-dimensional datasets and executing
complicated tasks with human expert levels of accuracy or greater. Methods: We
summarize and critically evaluate current applications of ML in dementia
research and highlight directions for future research. Results: We present an
overview of ML algorithms most frequently used in dementia research and
highlight future opportunities for the use of ML in clinical practice,
experimental medicine, and clinical trials. We discuss issues of
reproducibility, replicability and interpretability and how these impact the
clinical applicability of dementia research. Finally, we give examples of how
state-of-the-art methods, such as transfer learning, multi-task learning, and
reinforcement learning, may be applied to overcome these issues and aid the
translation of research to clinical practice in the future. Discussion:
ML-based models hold great promise to advance our understanding of the
underlying causes and pathological mechanisms of dementia.
|
2303.01950 | Dasa Kusnirakova | Dasa Kusnirakova, Mouzhi Ge, Leonard Walletzky, Barbora Buhnova | Interoperability-oriented Quality Assessment for Czech Open Data | In Proceedings of the 11th International Conference on Data Science,
Technology and Applications - DATA, 446-453, 2022 , Lisbon, Portugal | Proceedings of the 11th International Conference on Data Science,
Technology and Applications - DATA 2022 | 10.5220/0011291900003269 | null | cs.DB | http://creativecommons.org/licenses/by-nc-nd/4.0/ | With the rapid increase of published open datasets, it is crucial to support
the open data progress in smart cities while considering the open data quality.
In the Czech Republic, and its National Open Data Catalogue (NODC), the open
datasets are usually evaluated based on their metadata only, while leaving the
content and the adherence to the recommended data structure to the sole
responsibility of the data providers. The interoperability of open datasets
remains unknown. This paper therefore aims to propose a novel content-aware
quality evaluation framework that assesses the quality of open datasets based
on five data quality dimensions. With the proposed framework, we provide a
fundamental view on the interoperability-oriented data quality of Czech open
datasets, which are published in NODC. Our evaluations find that
domain-specific open data quality assessments are able to detect data quality
issues beyond traditional heuristics used for determining Czech open data
quality, increase their interoperability, and thus increase their potential to
bring value for the society. The findings of this research are beneficial not
only for the case of the Czech Republic, but also can be applied in other
countries that intend to enhance their open data quality evaluation processes.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 14:20:58 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Kusnirakova",
"Dasa",
""
],
[
"Ge",
"Mouzhi",
""
],
[
"Walletzky",
"Leonard",
""
],
[
"Buhnova",
"Barbora",
""
]
] | TITLE: Interoperability-oriented Quality Assessment for Czech Open Data
ABSTRACT: With the rapid increase of published open datasets, it is crucial to support
the open data progress in smart cities while considering the open data quality.
In the Czech Republic, and its National Open Data Catalogue (NODC), the open
datasets are usually evaluated based on their metadata only, while leaving the
content and the adherence to the recommended data structure to the sole
responsibility of the data providers. The interoperability of open datasets
remains unknown. This paper therefore aims to propose a novel content-aware
quality evaluation framework that assesses the quality of open datasets based
on five data quality dimensions. With the proposed framework, we provide a
fundamental view on the interoperability-oriented data quality of Czech open
datasets, which are published in NODC. Our evaluations find that
domain-specific open data quality assessments are able to detect data quality
issues beyond traditional heuristics used for determining Czech open data
quality, increase their interoperability, and thus increase their potential to
bring value for the society. The findings of this research are beneficial not
only for the case of the Czech Republic, but also can be applied in other
countries that intend to enhance their open data quality evaluation processes.
|
2303.01959 | Jinghuai Zhang | Jinghuai Zhang and Jinyuan Jia and Hongbin Liu and Neil Zhenqiang Gong | PointCert: Point Cloud Classification with Deterministic Certified
Robustness Guarantees | CVPR 2023 | null | null | null | cs.CR cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Point cloud classification is an essential component in many
security-critical applications such as autonomous driving and augmented
reality. However, point cloud classifiers are vulnerable to adversarially
perturbed point clouds. Existing certified defenses against adversarial point
clouds suffer from a key limitation: their certified robustness guarantees are
probabilistic, i.e., they produce an incorrect certified robustness guarantee
with some probability. In this work, we propose a general framework, namely
PointCert, that can transform an arbitrary point cloud classifier to be
certifiably robust against adversarial point clouds with deterministic
guarantees. PointCert certifiably predicts the same label for a point cloud
when the number of arbitrarily added, deleted, and/or modified points is less
than a threshold. Moreover, we propose multiple methods to optimize the
certified robustness guarantees of PointCert in three application scenarios. We
systematically evaluate PointCert on ModelNet and ScanObjectNN benchmark
datasets. Our results show that PointCert substantially outperforms
state-of-the-art certified defenses even though their robustness guarantees are
probabilistic.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 14:32:48 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhang",
"Jinghuai",
""
],
[
"Jia",
"Jinyuan",
""
],
[
"Liu",
"Hongbin",
""
],
[
"Gong",
"Neil Zhenqiang",
""
]
] | TITLE: PointCert: Point Cloud Classification with Deterministic Certified
Robustness Guarantees
ABSTRACT: Point cloud classification is an essential component in many
security-critical applications such as autonomous driving and augmented
reality. However, point cloud classifiers are vulnerable to adversarially
perturbed point clouds. Existing certified defenses against adversarial point
clouds suffer from a key limitation: their certified robustness guarantees are
probabilistic, i.e., they produce an incorrect certified robustness guarantee
with some probability. In this work, we propose a general framework, namely
PointCert, that can transform an arbitrary point cloud classifier to be
certifiably robust against adversarial point clouds with deterministic
guarantees. PointCert certifiably predicts the same label for a point cloud
when the number of arbitrarily added, deleted, and/or modified points is less
than a threshold. Moreover, we propose multiple methods to optimize the
certified robustness guarantees of PointCert in three application scenarios. We
systematically evaluate PointCert on ModelNet and ScanObjectNN benchmark
datasets. Our results show that PointCert substantially outperforms
state-of-the-art certified defenses even though their robustness guarantees are
probabilistic.
|
2303.01983 | Xinhang Wan | Xinhang Wan, Xinwang Liu, Jiyuan Liu, Siwei Wang, Yi Wen, Weixuan
Liang, En Zhu, Zhe Liu, Lu Zhou | Auto-weighted Multi-view Clustering for Large-scale Data | null | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-view clustering has gained broad attention owing to its capacity to
exploit complementary information across multiple data views. Although existing
methods demonstrate delightful clustering performance, most of them are of high
time complexity and cannot handle large-scale data. Matrix factorization-based
models are a representative of solving this problem. However, they assume that
the views share a dimension-fixed consensus coefficient matrix and
view-specific base matrices, limiting their representability. Moreover, a
series of large-scale algorithms that bear one or more hyperparameters are
impractical in real-world applications. To address the two issues, we propose
an auto-weighted multi-view clustering (AWMVC) algorithm. Specifically, AWMVC
first learns coefficient matrices from corresponding base matrices of different
dimensions, then fuses them to obtain an optimal consensus matrix. By mapping
original features into distinctive low-dimensional spaces, we can attain more
comprehensive knowledge, thus obtaining better clustering results. Moreover, we
design a six-step alternative optimization algorithm proven to be convergent
theoretically. Also, AWMVC shows excellent performance on various benchmark
datasets compared with existing ones. The code of AWMVC is publicly available
at https://github.com/wanxinhang/AAAI-2023-AWMVC.
| [
{
"version": "v1",
"created": "Sat, 21 Jan 2023 02:17:12 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Wan",
"Xinhang",
""
],
[
"Liu",
"Xinwang",
""
],
[
"Liu",
"Jiyuan",
""
],
[
"Wang",
"Siwei",
""
],
[
"Wen",
"Yi",
""
],
[
"Liang",
"Weixuan",
""
],
[
"Zhu",
"En",
""
],
[
"Liu",
"Zhe",
""
],
[
"Zhou",
"Lu",
""
]
] | TITLE: Auto-weighted Multi-view Clustering for Large-scale Data
ABSTRACT: Multi-view clustering has gained broad attention owing to its capacity to
exploit complementary information across multiple data views. Although existing
methods demonstrate delightful clustering performance, most of them are of high
time complexity and cannot handle large-scale data. Matrix factorization-based
models are a representative of solving this problem. However, they assume that
the views share a dimension-fixed consensus coefficient matrix and
view-specific base matrices, limiting their representability. Moreover, a
series of large-scale algorithms that bear one or more hyperparameters are
impractical in real-world applications. To address the two issues, we propose
an auto-weighted multi-view clustering (AWMVC) algorithm. Specifically, AWMVC
first learns coefficient matrices from corresponding base matrices of different
dimensions, then fuses them to obtain an optimal consensus matrix. By mapping
original features into distinctive low-dimensional spaces, we can attain more
comprehensive knowledge, thus obtaining better clustering results. Moreover, we
design a six-step alternative optimization algorithm proven to be convergent
theoretically. Also, AWMVC shows excellent performance on various benchmark
datasets compared with existing ones. The code of AWMVC is publicly available
at https://github.com/wanxinhang/AAAI-2023-AWMVC.
|
2303.02000 | You Shen | You Shen, Yunzhou Zhang, Yanmin Wu, Zhenyu Wang, Linghao Yang, Sonya
Coleman, Dermot Kerr | BSH-Det3D: Improving 3D Object Detection with BEV Shape Heatmap | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The progress of LiDAR-based 3D object detection has significantly enhanced
developments in autonomous driving and robotics. However, due to the
limitations of LiDAR sensors, object shapes suffer from deterioration in
occluded and distant areas, which creates a fundamental challenge to 3D
perception. Existing methods estimate specific 3D shapes and achieve remarkable
performance. However, these methods rely on extensive computation and memory,
causing imbalances between accuracy and real-time performance. To tackle this
challenge, we propose a novel LiDAR-based 3D object detection model named
BSH-Det3D, which applies an effective way to enhance spatial features by
estimating complete shapes from a bird's eye view (BEV). Specifically, we
design the Pillar-based Shape Completion (PSC) module to predict the
probability of occupancy whether a pillar contains object shapes. The PSC
module generates a BEV shape heatmap for each scene. After integrating with
heatmaps, BSH-Det3D can provide additional information in shape deterioration
areas and generate high-quality 3D proposals. We also design an attention-based
densification fusion module (ADF) to adaptively associate the sparse features
with heatmaps and raw points. The ADF module integrates the advantages of
points and shapes knowledge with negligible overheads. Extensive experiments on
the KITTI benchmark achieve state-of-the-art (SOTA) performance in terms of
accuracy and speed, demonstrating the efficiency and flexibility of BSH-Det3D.
The source code is available on https://github.com/mystorm16/BSH-Det3D.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 15:13:11 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Shen",
"You",
""
],
[
"Zhang",
"Yunzhou",
""
],
[
"Wu",
"Yanmin",
""
],
[
"Wang",
"Zhenyu",
""
],
[
"Yang",
"Linghao",
""
],
[
"Coleman",
"Sonya",
""
],
[
"Kerr",
"Dermot",
""
]
] | TITLE: BSH-Det3D: Improving 3D Object Detection with BEV Shape Heatmap
ABSTRACT: The progress of LiDAR-based 3D object detection has significantly enhanced
developments in autonomous driving and robotics. However, due to the
limitations of LiDAR sensors, object shapes suffer from deterioration in
occluded and distant areas, which creates a fundamental challenge to 3D
perception. Existing methods estimate specific 3D shapes and achieve remarkable
performance. However, these methods rely on extensive computation and memory,
causing imbalances between accuracy and real-time performance. To tackle this
challenge, we propose a novel LiDAR-based 3D object detection model named
BSH-Det3D, which applies an effective way to enhance spatial features by
estimating complete shapes from a bird's eye view (BEV). Specifically, we
design the Pillar-based Shape Completion (PSC) module to predict the
probability of occupancy whether a pillar contains object shapes. The PSC
module generates a BEV shape heatmap for each scene. After integrating with
heatmaps, BSH-Det3D can provide additional information in shape deterioration
areas and generate high-quality 3D proposals. We also design an attention-based
densification fusion module (ADF) to adaptively associate the sparse features
with heatmaps and raw points. The ADF module integrates the advantages of
points and shapes knowledge with negligible overheads. Extensive experiments on
the KITTI benchmark achieve state-of-the-art (SOTA) performance in terms of
accuracy and speed, demonstrating the efficiency and flexibility of BSH-Det3D.
The source code is available on https://github.com/mystorm16/BSH-Det3D.
|
2303.02028 | Vyacheslav Yukalov | T. Kovalenko, S. Vincent, V.I. Yukalov, and D. Sornette | Calibration of Quantum Decision Theory: Aversion to Large Losses and
Predictability of Probabilistic Choices | Latex file, 51 pages, 19 figures | J. Phys. Complex. 4 (2023) 015009 | null | null | cs.AI quant-ph | http://creativecommons.org/licenses/by/4.0/ | We present the first calibration of quantum decision theory (QDT) to a
dataset of binary risky choice. We quantitatively account for the fraction of
choice reversals between two repetitions of the experiment, using a
probabilistic choice formulation in the simplest form without model assumption
or adjustable parameters. The prediction of choice reversal is then refined by
introducing heterogeneity between decision makers through their differentiation
into two groups: ``majoritarian'' and ``contrarian'' (in proportion 3:1). This
supports the first fundamental tenet of QDT, which models choice as an inherent
probabilistic process, where the probability of a prospect can be expressed as
the sum of its utility and attraction factors. We propose to parameterise the
utility factor with a stochastic version of cumulative prospect theory
(logit-CPT), and the attraction factor with a constant absolute risk aversion
(CARA) function. For this dataset, and penalising the larger number of QDT
parameters via the Wilks test of nested hypotheses, the QDT model is found to
perform significantly better than logit-CPT at both the aggregate and
individual levels, and for all considered fit criteria for the first experiment
iteration and for predictions (second ``out-of-sample'' iteration). The
distinctive QDT effect captured by the attraction factor is mostly appreciable
(i.e., most relevant and strongest in amplitude) for prospects with big losses.
Our quantitative analysis of the experimental results supports the existence of
an intrinsic limit of predictability, which is associated with the inherent
probabilistic nature of choice. The results of the paper can find applications
both in the prediction of choice of human decision makers as well as for
organizing the operation of artificial intelligence.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 15:46:26 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Kovalenko",
"T.",
""
],
[
"Vincent",
"S.",
""
],
[
"Yukalov",
"V. I.",
""
],
[
"Sornette",
"D.",
""
]
] | TITLE: Calibration of Quantum Decision Theory: Aversion to Large Losses and
Predictability of Probabilistic Choices
ABSTRACT: We present the first calibration of quantum decision theory (QDT) to a
dataset of binary risky choice. We quantitatively account for the fraction of
choice reversals between two repetitions of the experiment, using a
probabilistic choice formulation in the simplest form without model assumption
or adjustable parameters. The prediction of choice reversal is then refined by
introducing heterogeneity between decision makers through their differentiation
into two groups: ``majoritarian'' and ``contrarian'' (in proportion 3:1). This
supports the first fundamental tenet of QDT, which models choice as an inherent
probabilistic process, where the probability of a prospect can be expressed as
the sum of its utility and attraction factors. We propose to parameterise the
utility factor with a stochastic version of cumulative prospect theory
(logit-CPT), and the attraction factor with a constant absolute risk aversion
(CARA) function. For this dataset, and penalising the larger number of QDT
parameters via the Wilks test of nested hypotheses, the QDT model is found to
perform significantly better than logit-CPT at both the aggregate and
individual levels, and for all considered fit criteria for the first experiment
iteration and for predictions (second ``out-of-sample'' iteration). The
distinctive QDT effect captured by the attraction factor is mostly appreciable
(i.e., most relevant and strongest in amplitude) for prospects with big losses.
Our quantitative analysis of the experimental results supports the existence of
an intrinsic limit of predictability, which is associated with the inherent
probabilistic nature of choice. The results of the paper can find applications
both in the prediction of choice of human decision makers as well as for
organizing the operation of artificial intelligence.
|
2303.02032 | Basheer Qolomany | Aos Mulahuwaish, Matthew Loucks, Basheer Qolomany, and Ala Al-Fuqaha | Topic Modeling Based on Two-Step Flow Theory: Application to Tweets
about Bitcoin | null | null | null | null | cs.SI cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | Digital cryptocurrencies such as Bitcoin have exploded in recent years in
both popularity and value. By their novelty, cryptocurrencies tend to be both
volatile and highly speculative. The capricious nature of these coins is helped
facilitated by social media networks such as Twitter. However, not everyone's
opinion matters equally, with most posts garnering little to no attention.
Additionally, the majority of tweets are retweeted from popular posts. We must
determine whose opinion matters and the difference between influential and
non-influential users. This study separates these two groups and analyzes the
differences between them. It uses Hypertext-induced Topic Selection (HITS)
algorithm, which segregates the dataset based on influence. Topic modeling is
then employed to uncover differences in each group's speech types and what
group may best represent the entire community. We found differences in language
and interest between these two groups regarding Bitcoin and that the opinion
leaders of Twitter are not aligned with the majority of users. There were 2559
opinion leaders (0.72% of users) who accounted for 80% of the authority and the
majority (99.28%) users for the remaining 20% out of a total of 355,139 users.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 15:51:05 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Mulahuwaish",
"Aos",
""
],
[
"Loucks",
"Matthew",
""
],
[
"Qolomany",
"Basheer",
""
],
[
"Al-Fuqaha",
"Ala",
""
]
] | TITLE: Topic Modeling Based on Two-Step Flow Theory: Application to Tweets
about Bitcoin
ABSTRACT: Digital cryptocurrencies such as Bitcoin have exploded in recent years in
both popularity and value. By their novelty, cryptocurrencies tend to be both
volatile and highly speculative. The capricious nature of these coins is helped
facilitated by social media networks such as Twitter. However, not everyone's
opinion matters equally, with most posts garnering little to no attention.
Additionally, the majority of tweets are retweeted from popular posts. We must
determine whose opinion matters and the difference between influential and
non-influential users. This study separates these two groups and analyzes the
differences between them. It uses Hypertext-induced Topic Selection (HITS)
algorithm, which segregates the dataset based on influence. Topic modeling is
then employed to uncover differences in each group's speech types and what
group may best represent the entire community. We found differences in language
and interest between these two groups regarding Bitcoin and that the opinion
leaders of Twitter are not aligned with the majority of users. There were 2559
opinion leaders (0.72% of users) who accounted for 80% of the authority and the
majority (99.28%) users for the remaining 20% out of a total of 355,139 users.
|
2303.02033 | Chen Yiwei | Yiwei Chen, Chen Jiang and Yu Pan | Single-photon Image Super-resolution via Self-supervised Learning | Accepted by ICASSP 2023 | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Single-Photon Image Super-Resolution (SPISR) aims to recover a
high-resolution volumetric photon counting cube from a noisy low-resolution one
by computational imaging algorithms. In real-world scenarios, pairs of training
samples are often expensive or impossible to obtain. By extending Equivariant
Imaging (EI) to volumetric single-photon data, we propose a self-supervised
learning framework for the SPISR task. Particularly, using the Poisson unbiased
Kullback-Leibler risk estimator and equivariance, our method is able to learn
from noisy measurements without ground truths. Comprehensive experiments on
simulated and real-world dataset demonstrate that the proposed method achieves
comparable performance with supervised learning and outperforms
interpolation-based methods.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 15:52:01 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Chen",
"Yiwei",
""
],
[
"Jiang",
"Chen",
""
],
[
"Pan",
"Yu",
""
]
] | TITLE: Single-photon Image Super-resolution via Self-supervised Learning
ABSTRACT: Single-Photon Image Super-Resolution (SPISR) aims to recover a
high-resolution volumetric photon counting cube from a noisy low-resolution one
by computational imaging algorithms. In real-world scenarios, pairs of training
samples are often expensive or impossible to obtain. By extending Equivariant
Imaging (EI) to volumetric single-photon data, we propose a self-supervised
learning framework for the SPISR task. Particularly, using the Poisson unbiased
Kullback-Leibler risk estimator and equivariance, our method is able to learn
from noisy measurements without ground truths. Comprehensive experiments on
simulated and real-world dataset demonstrate that the proposed method achieves
comparable performance with supervised learning and outperforms
interpolation-based methods.
|
2303.02048 | Minh-Toan Nguyen | Minh-Toan Nguyen, Romain Couillet | Asymptotic Bayes risk of semi-supervised multitask learning on Gaussian
mixture | AISTATS 2023 | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | The article considers semi-supervised multitask learning on a Gaussian
mixture model (GMM). Using methods from statistical physics, we compute the
asymptotic Bayes risk of each task in the regime of large datasets in high
dimension, from which we analyze the role of task similarity in learning and
evaluate the performance gain when tasks are learned together rather than
separately. In the supervised case, we derive a simple algorithm that attains
the Bayes optimal performance.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 16:17:30 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Nguyen",
"Minh-Toan",
""
],
[
"Couillet",
"Romain",
""
]
] | TITLE: Asymptotic Bayes risk of semi-supervised multitask learning on Gaussian
mixture
ABSTRACT: The article considers semi-supervised multitask learning on a Gaussian
mixture model (GMM). Using methods from statistical physics, we compute the
asymptotic Bayes risk of each task in the regime of large datasets in high
dimension, from which we analyze the role of task similarity in learning and
evaluate the performance gain when tasks are learned together rather than
separately. In the supervised case, we derive a simple algorithm that attains
the Bayes optimal performance.
|
2303.02052 | Michael (Micky) Fire | Shmuel Horowitz, Dima Kagan, Galit Fuhrmann Alpert, and Michael Fire | Interruptions detection in video conferences | null | null | null | null | cs.CV cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, video conferencing (VC) popularity has skyrocketed for a
wide range of activities. As a result, the number of VC users surged sharply.
The sharp increase in VC usage has been accompanied by various newly emerging
privacy and security challenges. VC meetings became a target for various
security attacks, such as Zoombombing. Other VC-related challenges also
emerged. For example, during COVID lockdowns, educators had to teach in online
environments struggling with keeping students engaged for extended periods. In
parallel, the amount of available VC videos has grown exponentially. Thus,
users and companies are limited in finding abnormal segments in VC meetings
within the converging volumes of data. Such abnormal events that affect most
meeting participants may be indicators of interesting points in time, including
security attacks or other changes in meeting climate, like someone joining a
meeting or sharing a dramatic content. Here, we present a novel algorithm for
detecting abnormal events in VC data. We curated VC publicly available
recordings, including meetings with interruptions. We analyzed the videos using
our algorithm, extracting time windows where abnormal occurrences were
detected. Our algorithm is a pipeline that combines multiple methods in several
steps to detect users' faces in each video frame, track face locations during
the meeting and generate vector representations of a facial expression for each
face in each frame. Vector representations are used to monitor changes in
facial expressions throughout the meeting for each participant. The overall
change in meeting climate is quantified using those parameters across all
participants, and translating them into event anomaly detection. This is the
first open pipeline for automatically detecting anomaly events in VC meetings.
Our model detects abnormal events with 92.3% precision over the collected
dataset.
| [
{
"version": "v1",
"created": "Sat, 25 Feb 2023 17:43:05 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Horowitz",
"Shmuel",
""
],
[
"Kagan",
"Dima",
""
],
[
"Alpert",
"Galit Fuhrmann",
""
],
[
"Fire",
"Michael",
""
]
] | TITLE: Interruptions detection in video conferences
ABSTRACT: In recent years, video conferencing (VC) popularity has skyrocketed for a
wide range of activities. As a result, the number of VC users surged sharply.
The sharp increase in VC usage has been accompanied by various newly emerging
privacy and security challenges. VC meetings became a target for various
security attacks, such as Zoombombing. Other VC-related challenges also
emerged. For example, during COVID lockdowns, educators had to teach in online
environments struggling with keeping students engaged for extended periods. In
parallel, the amount of available VC videos has grown exponentially. Thus,
users and companies are limited in finding abnormal segments in VC meetings
within the converging volumes of data. Such abnormal events that affect most
meeting participants may be indicators of interesting points in time, including
security attacks or other changes in meeting climate, like someone joining a
meeting or sharing a dramatic content. Here, we present a novel algorithm for
detecting abnormal events in VC data. We curated VC publicly available
recordings, including meetings with interruptions. We analyzed the videos using
our algorithm, extracting time windows where abnormal occurrences were
detected. Our algorithm is a pipeline that combines multiple methods in several
steps to detect users' faces in each video frame, track face locations during
the meeting and generate vector representations of a facial expression for each
face in each frame. Vector representations are used to monitor changes in
facial expressions throughout the meeting for each participant. The overall
change in meeting climate is quantified using those parameters across all
participants, and translating them into event anomaly detection. This is the
first open pipeline for automatically detecting anomaly events in VC meetings.
Our model detects abnormal events with 92.3% precision over the collected
dataset.
|
2303.02073 | Yang Yu | Xu-Hui Liu, Feng Xu, Xinyu Zhang, Tianyuan Liu, Shengyi Jiang, Ruifeng
Chen, Zongzhang Zhang, Yang Yu | How To Guide Your Learner: Imitation Learning with Active Adaptive
Expert Involvement | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Imitation learning aims to mimic the behavior of experts without explicit
reward signals. Passive imitation learning methods which use static expert
datasets typically suffer from compounding error, low sample efficiency, and
high hyper-parameter sensitivity. In contrast, active imitation learning
methods solicit expert interventions to address the limitations. However,
recent active imitation learning methods are designed based on human intuitions
or empirical experience without theoretical guarantee. In this paper, we
propose a novel active imitation learning framework based on a teacher-student
interaction model, in which the teacher's goal is to identify the best teaching
behavior and actively affect the student's learning process. By solving the
optimization objective of this framework, we propose a practical
implementation, naming it AdapMen. Theoretical analysis shows that AdapMen can
improve the error bound and avoid compounding error under mild conditions.
Experiments on the MetaDrive benchmark and Atari 2600 games validate our
theoretical analysis and show that our method achieves near-expert performance
with much less expert involvement and total sampling steps than previous
methods. The code is available at https://github.com/liuxhym/AdapMen.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 16:44:33 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Liu",
"Xu-Hui",
""
],
[
"Xu",
"Feng",
""
],
[
"Zhang",
"Xinyu",
""
],
[
"Liu",
"Tianyuan",
""
],
[
"Jiang",
"Shengyi",
""
],
[
"Chen",
"Ruifeng",
""
],
[
"Zhang",
"Zongzhang",
""
],
[
"Yu",
"Yang",
""
]
] | TITLE: How To Guide Your Learner: Imitation Learning with Active Adaptive
Expert Involvement
ABSTRACT: Imitation learning aims to mimic the behavior of experts without explicit
reward signals. Passive imitation learning methods which use static expert
datasets typically suffer from compounding error, low sample efficiency, and
high hyper-parameter sensitivity. In contrast, active imitation learning
methods solicit expert interventions to address the limitations. However,
recent active imitation learning methods are designed based on human intuitions
or empirical experience without theoretical guarantee. In this paper, we
propose a novel active imitation learning framework based on a teacher-student
interaction model, in which the teacher's goal is to identify the best teaching
behavior and actively affect the student's learning process. By solving the
optimization objective of this framework, we propose a practical
implementation, naming it AdapMen. Theoretical analysis shows that AdapMen can
improve the error bound and avoid compounding error under mild conditions.
Experiments on the MetaDrive benchmark and Atari 2600 games validate our
theoretical analysis and show that our method achieves near-expert performance
with much less expert involvement and total sampling steps than previous
methods. The code is available at https://github.com/liuxhym/AdapMen.
|
2303.02128 | Mahdi Gilany | Mahdi Gilany, Paul Wilson, Andrea Perera-Ortega, Amoon Jamzad, Minh
Nguyen Nhat To, Fahimeh Fooladgar, Brian Wodlinger, Purang Abolmaesumi,
Parvin Mousavi | TRUSformer: Improving Prostate Cancer Detection from Micro-Ultrasound
Using Attention and Self-Supervision | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A large body of previous machine learning methods for ultrasound-based
prostate cancer detection classify small regions of interest (ROIs) of
ultrasound signals that lie within a larger needle trace corresponding to a
prostate tissue biopsy (called biopsy core). These ROI-scale models suffer from
weak labeling as histopathology results available for biopsy cores only
approximate the distribution of cancer in the ROIs. ROI-scale models do not
take advantage of contextual information that are normally considered by
pathologists, i.e. they do not consider information about surrounding tissue
and larger-scale trends when identifying cancer. We aim to improve cancer
detection by taking a multi-scale, i.e. ROI-scale and biopsy core-scale,
approach. Methods: Our multi-scale approach combines (i) an "ROI-scale" model
trained using self-supervised learning to extract features from small ROIs and
(ii) a "core-scale" transformer model that processes a collection of extracted
features from multiple ROIs in the needle trace region to predict the tissue
type of the corresponding core. Attention maps, as a byproduct, allow us to
localize cancer at the ROI scale. We analyze this method using a dataset of
micro-ultrasound acquired from 578 patients who underwent prostate biopsy, and
compare our model to baseline models and other large-scale studies in the
literature. Results and Conclusions: Our model shows consistent and substantial
performance improvements compared to ROI-scale-only models. It achieves 80.3%
AUROC, a statistically significant improvement over ROI-scale classification.
We also compare our method to large studies on prostate cancer detection, using
other imaging modalities. Our code is publicly available at
www.github.com/med-i-lab/TRUSFormer
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 18:12:46 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Gilany",
"Mahdi",
""
],
[
"Wilson",
"Paul",
""
],
[
"Perera-Ortega",
"Andrea",
""
],
[
"Jamzad",
"Amoon",
""
],
[
"To",
"Minh Nguyen Nhat",
""
],
[
"Fooladgar",
"Fahimeh",
""
],
[
"Wodlinger",
"Brian",
""
],
[
"Abolmaesumi",
"Purang",
""
],
[
"Mousavi",
"Parvin",
""
]
] | TITLE: TRUSformer: Improving Prostate Cancer Detection from Micro-Ultrasound
Using Attention and Self-Supervision
ABSTRACT: A large body of previous machine learning methods for ultrasound-based
prostate cancer detection classify small regions of interest (ROIs) of
ultrasound signals that lie within a larger needle trace corresponding to a
prostate tissue biopsy (called biopsy core). These ROI-scale models suffer from
weak labeling as histopathology results available for biopsy cores only
approximate the distribution of cancer in the ROIs. ROI-scale models do not
take advantage of contextual information that are normally considered by
pathologists, i.e. they do not consider information about surrounding tissue
and larger-scale trends when identifying cancer. We aim to improve cancer
detection by taking a multi-scale, i.e. ROI-scale and biopsy core-scale,
approach. Methods: Our multi-scale approach combines (i) an "ROI-scale" model
trained using self-supervised learning to extract features from small ROIs and
(ii) a "core-scale" transformer model that processes a collection of extracted
features from multiple ROIs in the needle trace region to predict the tissue
type of the corresponding core. Attention maps, as a byproduct, allow us to
localize cancer at the ROI scale. We analyze this method using a dataset of
micro-ultrasound acquired from 578 patients who underwent prostate biopsy, and
compare our model to baseline models and other large-scale studies in the
literature. Results and Conclusions: Our model shows consistent and substantial
performance improvements compared to ROI-scale-only models. It achieves 80.3%
AUROC, a statistically significant improvement over ROI-scale classification.
We also compare our method to large studies on prostate cancer detection, using
other imaging modalities. Our code is publicly available at
www.github.com/med-i-lab/TRUSFormer
|
2303.02141 | Shiwei Liu | Shiwei Liu, Tianlong Chen, Zhenyu Zhang, Xuxi Chen, Tianjin Huang,
Ajay Jaiswal, Zhangyang Wang | Sparsity May Cry: Let Us Fail (Current) Sparse Neural Networks Together! | We open-source SMC-Bench to assist researchers in building
next-generation sparse algorithms that scale and generalize:
https://github.com/VITA-Group/SMC-Bench | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Sparse Neural Networks (SNNs) have received voluminous attention
predominantly due to growing computational and memory footprints of
consistently exploding parameter count in large-scale models. Similar to their
dense counterparts, recent SNNs generalize just as well and are equipped with
numerous favorable benefits (e.g., low complexity, high scalability, and
robustness), sometimes even better than the original dense networks. As
research effort is focused on developing increasingly sophisticated sparse
algorithms, it is startling that a comprehensive benchmark to evaluate the
effectiveness of these algorithms has been highly overlooked. In absence of a
carefully crafted evaluation benchmark, most if not all, sparse algorithms are
evaluated against fairly simple and naive tasks (eg. CIFAR, ImageNet, GLUE,
etc.), which can potentially camouflage many advantages as well unexpected
predicaments of SNNs. In pursuit of a more general evaluation and unveiling the
true potential of sparse algorithms, we introduce "Sparsity May Cry" Benchmark
(SMC-Bench), a collection of carefully-curated 4 diverse tasks with 10
datasets, that accounts for capturing a wide range of domain-specific and
sophisticated knowledge. Our systemic evaluation of the most representative
sparse algorithms reveals an important obscured observation: the
state-of-the-art magnitude- and/or gradient-based sparse algorithms seemingly
fail to perform on SMC-Bench when applied out-of-the-box, sometimes at
significantly trivial sparsity as low as 5%. By incorporating these
well-thought and diverse tasks, SMC-Bench is designed to favor and encourage
the development of more scalable and generalizable sparse algorithms.
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 18:47:21 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Liu",
"Shiwei",
""
],
[
"Chen",
"Tianlong",
""
],
[
"Zhang",
"Zhenyu",
""
],
[
"Chen",
"Xuxi",
""
],
[
"Huang",
"Tianjin",
""
],
[
"Jaiswal",
"Ajay",
""
],
[
"Wang",
"Zhangyang",
""
]
] | TITLE: Sparsity May Cry: Let Us Fail (Current) Sparse Neural Networks Together!
ABSTRACT: Sparse Neural Networks (SNNs) have received voluminous attention
predominantly due to growing computational and memory footprints of
consistently exploding parameter count in large-scale models. Similar to their
dense counterparts, recent SNNs generalize just as well and are equipped with
numerous favorable benefits (e.g., low complexity, high scalability, and
robustness), sometimes even better than the original dense networks. As
research effort is focused on developing increasingly sophisticated sparse
algorithms, it is startling that a comprehensive benchmark to evaluate the
effectiveness of these algorithms has been highly overlooked. In absence of a
carefully crafted evaluation benchmark, most if not all, sparse algorithms are
evaluated against fairly simple and naive tasks (eg. CIFAR, ImageNet, GLUE,
etc.), which can potentially camouflage many advantages as well unexpected
predicaments of SNNs. In pursuit of a more general evaluation and unveiling the
true potential of sparse algorithms, we introduce "Sparsity May Cry" Benchmark
(SMC-Bench), a collection of carefully-curated 4 diverse tasks with 10
datasets, that accounts for capturing a wide range of domain-specific and
sophisticated knowledge. Our systemic evaluation of the most representative
sparse algorithms reveals an important obscured observation: the
state-of-the-art magnitude- and/or gradient-based sparse algorithms seemingly
fail to perform on SMC-Bench when applied out-of-the-box, sometimes at
significantly trivial sparsity as low as 5%. By incorporating these
well-thought and diverse tasks, SMC-Bench is designed to favor and encourage
the development of more scalable and generalizable sparse algorithms.
|
2303.02153 | Wenliang Zhao | Wenliang Zhao, Yongming Rao, Zuyan Liu, Benlin Liu, Jie Zhou, Jiwen Lu | Unleashing Text-to-Image Diffusion Models for Visual Perception | project page: https://vpd.ivg-research.xyz | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models (DMs) have become the new trend of generative models and
have demonstrated a powerful ability of conditional synthesis. Among those,
text-to-image diffusion models pre-trained on large-scale image-text pairs are
highly controllable by customizable prompts. Unlike the unconditional
generative models that focus on low-level attributes and details, text-to-image
diffusion models contain more high-level knowledge thanks to the
vision-language pre-training. In this paper, we propose VPD (Visual Perception
with a pre-trained Diffusion model), a new framework that exploits the semantic
information of a pre-trained text-to-image diffusion model in visual perception
tasks. Instead of using the pre-trained denoising autoencoder in a
diffusion-based pipeline, we simply use it as a backbone and aim to study how
to take full advantage of the learned knowledge. Specifically, we prompt the
denoising decoder with proper textual inputs and refine the text features with
an adapter, leading to a better alignment to the pre-trained stage and making
the visual contents interact with the text prompts. We also propose to utilize
the cross-attention maps between the visual features and the text features to
provide explicit guidance. Compared with other pre-training methods, we show
that vision-language pre-trained diffusion models can be faster adapted to
downstream visual perception tasks using the proposed VPD. Extensive
experiments on semantic segmentation, referring image segmentation and depth
estimation demonstrates the effectiveness of our method. Notably, VPD attains
0.254 RMSE on NYUv2 depth estimation and 73.3% oIoU on RefCOCO-val referring
image segmentation, establishing new records on these two benchmarks. Code is
available at https://github.com/wl-zhao/VPD
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 18:59:47 GMT"
}
] | 2023-03-06T00:00:00 | [
[
"Zhao",
"Wenliang",
""
],
[
"Rao",
"Yongming",
""
],
[
"Liu",
"Zuyan",
""
],
[
"Liu",
"Benlin",
""
],
[
"Zhou",
"Jie",
""
],
[
"Lu",
"Jiwen",
""
]
] | TITLE: Unleashing Text-to-Image Diffusion Models for Visual Perception
ABSTRACT: Diffusion models (DMs) have become the new trend of generative models and
have demonstrated a powerful ability of conditional synthesis. Among those,
text-to-image diffusion models pre-trained on large-scale image-text pairs are
highly controllable by customizable prompts. Unlike the unconditional
generative models that focus on low-level attributes and details, text-to-image
diffusion models contain more high-level knowledge thanks to the
vision-language pre-training. In this paper, we propose VPD (Visual Perception
with a pre-trained Diffusion model), a new framework that exploits the semantic
information of a pre-trained text-to-image diffusion model in visual perception
tasks. Instead of using the pre-trained denoising autoencoder in a
diffusion-based pipeline, we simply use it as a backbone and aim to study how
to take full advantage of the learned knowledge. Specifically, we prompt the
denoising decoder with proper textual inputs and refine the text features with
an adapter, leading to a better alignment to the pre-trained stage and making
the visual contents interact with the text prompts. We also propose to utilize
the cross-attention maps between the visual features and the text features to
provide explicit guidance. Compared with other pre-training methods, we show
that vision-language pre-trained diffusion models can be faster adapted to
downstream visual perception tasks using the proposed VPD. Extensive
experiments on semantic segmentation, referring image segmentation and depth
estimation demonstrates the effectiveness of our method. Notably, VPD attains
0.254 RMSE on NYUv2 depth estimation and 73.3% oIoU on RefCOCO-val referring
image segmentation, establishing new records on these two benchmarks. Code is
available at https://github.com/wl-zhao/VPD
|
1908.01866 | Chloe He | Chloe He, Gerard Glowacki, Alexis Gkantiragas | Unsupervised Representations of Pollen in Bright-Field Microscopy | Accepted at the Workshop on Computational Biology at the
International Conference on Machine Learning (ICML) in Long Beach, CA, USA on
June 14, 2019 | null | null | null | cs.CV cs.LG cs.NE q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present the first unsupervised deep learning method for pollen analysis
using bright-field microscopy. Using a modest dataset of 650 images of pollen
grains collected from honey, we achieve family level identification of pollen.
We embed images of pollen grains into a low-dimensional latent space and
compare Euclidean and Riemannian metrics on these spaces for clustering. We
propose this system for automated analysis of pollen and other microscopic
biological structures which have only small or unlabelled datasets available.
| [
{
"version": "v1",
"created": "Mon, 5 Aug 2019 21:29:51 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"He",
"Chloe",
""
],
[
"Glowacki",
"Gerard",
""
],
[
"Gkantiragas",
"Alexis",
""
]
] | TITLE: Unsupervised Representations of Pollen in Bright-Field Microscopy
ABSTRACT: We present the first unsupervised deep learning method for pollen analysis
using bright-field microscopy. Using a modest dataset of 650 images of pollen
grains collected from honey, we achieve family level identification of pollen.
We embed images of pollen grains into a low-dimensional latent space and
compare Euclidean and Riemannian metrics on these spaces for clustering. We
propose this system for automated analysis of pollen and other microscopic
biological structures which have only small or unlabelled datasets available.
|
2002.03328 | Yufeng Zhang | Yufeng Zhang, Jialu Pan, Wanwei Liu, Zhenbang Chen, Ji Wang, Zhiming
Liu, Kenli Li, Hongmei Wei | Kullback-Leibler Divergence-Based Out-of-Distribution Detection with
Flow-Based Generative Models | null | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent research has revealed that deep generative models including flow-based
models and Variational Autoencoders may assign higher likelihoods to
out-of-distribution (OOD) data than in-distribution (ID) data. However, we
cannot sample OOD data from the model. This counterintuitive phenomenon has not
been satisfactorily explained and brings obstacles to OOD detection with
flow-based models. In this paper, we prove theorems to investigate the
Kullback-Leibler divergence in flow-based model and give two explanations for
the above phenomenon. Based on our theoretical analysis, we propose a new
method \PADmethod\ to leverage KL divergence and local pixel dependence of
representations to perform anomaly detection. Experimental results on prevalent
benchmarks demonstrate the effectiveness and robustness of our method. For
group anomaly detection, our method achieves 98.1\% AUROC on average with a
small batch size of 5. On the contrary, the baseline typicality test-based
method only achieves 64.6\% AUROC on average due to its failure on challenging
problems. Our method also outperforms the state-of-the-art method by 9.1\%
AUROC. For point-wise anomaly detection, our method achieves 90.7\% AUROC on
average and outperforms the baseline by 5.2\% AUROC. Besides, our method has
the least notable failures and is the most robust one.
| [
{
"version": "v1",
"created": "Sun, 9 Feb 2020 09:54:12 GMT"
},
{
"version": "v2",
"created": "Sun, 12 Jul 2020 11:56:54 GMT"
},
{
"version": "v3",
"created": "Fri, 5 Feb 2021 13:56:04 GMT"
},
{
"version": "v4",
"created": "Thu, 16 Sep 2021 05:59:28 GMT"
},
{
"version": "v5",
"created": "Thu, 2 Mar 2023 06:56:26 GMT"
}
] | 2023-03-03T00:00:00 | [
[
"Zhang",
"Yufeng",
""
],
[
"Pan",
"Jialu",
""
],
[
"Liu",
"Wanwei",
""
],
[
"Chen",
"Zhenbang",
""
],
[
"Wang",
"Ji",
""
],
[
"Liu",
"Zhiming",
""
],
[
"Li",
"Kenli",
""
],
[
"Wei",
"Hongmei",
""
]
] | TITLE: Kullback-Leibler Divergence-Based Out-of-Distribution Detection with
Flow-Based Generative Models
ABSTRACT: Recent research has revealed that deep generative models including flow-based
models and Variational Autoencoders may assign higher likelihoods to
out-of-distribution (OOD) data than in-distribution (ID) data. However, we
cannot sample OOD data from the model. This counterintuitive phenomenon has not
been satisfactorily explained and brings obstacles to OOD detection with
flow-based models. In this paper, we prove theorems to investigate the
Kullback-Leibler divergence in flow-based model and give two explanations for
the above phenomenon. Based on our theoretical analysis, we propose a new
method \PADmethod\ to leverage KL divergence and local pixel dependence of
representations to perform anomaly detection. Experimental results on prevalent
benchmarks demonstrate the effectiveness and robustness of our method. For
group anomaly detection, our method achieves 98.1\% AUROC on average with a
small batch size of 5. On the contrary, the baseline typicality test-based
method only achieves 64.6\% AUROC on average due to its failure on challenging
problems. Our method also outperforms the state-of-the-art method by 9.1\%
AUROC. For point-wise anomaly detection, our method achieves 90.7\% AUROC on
average and outperforms the baseline by 5.2\% AUROC. Besides, our method has
the least notable failures and is the most robust one.
|
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.