id
large_string | submitter
large_string | authors
large_string | title
large_string | comments
large_string | journal-ref
large_string | doi
large_string | report-no
large_string | categories
large_string | license
large_string | abstract
large_string | versions
large list | update_date
timestamp[ms] | authors_parsed
large list | prompt
large_string | label
string | prob
float64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2301.04298 | Yalin E. Sagduyu | Yalin E. Sagduyu, Sennur Ulukus, Aylin Yener | Age of Information in Deep Learning-Driven Task-Oriented Communications | null | null | null | null | cs.IT cs.AI cs.LG cs.NI eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper studies the notion of age in task-oriented communications that
aims to execute a task at a receiver utilizing the data at its transmitter. The
transmitter-receiver operations are modeled as an encoder-decoder pair that is
jointly trained while considering channel effects. The encoder converts data
samples into feature vectors of small dimension and transmits them with a small
number of channel uses thereby reducing the number of transmissions and
latency. Instead of reconstructing input samples, the decoder performs a task,
e.g., classification, on the received signals. Applying different deep neural
networks of encoder-decoder pairs on MNIST and CIFAR-10 image datasets, the
classifier accuracy is shown to increase with the number of channel uses at the
expense of longer service time. The peak age of task information (PAoTI) is
introduced to analyze this accuracy-latency tradeoff when the age grows unless
a received signal is classified correctly. By incorporating channel and traffic
effects, design guidelines are obtained for task-oriented communications by
characterizing how the PAoTI first decreases and then increases with the number
of channel uses. A dynamic update mechanism is presented to adapt the number of
channel uses to channel and traffic conditions, and reduce the PAoTI in
task-oriented communications.
| [
{
"version": "v1",
"created": "Wed, 11 Jan 2023 04:15:51 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 16:16:18 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Sagduyu",
"Yalin E.",
""
],
[
"Ulukus",
"Sennur",
""
],
[
"Yener",
"Aylin",
""
]
] | TITLE: Age of Information in Deep Learning-Driven Task-Oriented Communications
ABSTRACT: This paper studies the notion of age in task-oriented communications that
aims to execute a task at a receiver utilizing the data at its transmitter. The
transmitter-receiver operations are modeled as an encoder-decoder pair that is
jointly trained while considering channel effects. The encoder converts data
samples into feature vectors of small dimension and transmits them with a small
number of channel uses thereby reducing the number of transmissions and
latency. Instead of reconstructing input samples, the decoder performs a task,
e.g., classification, on the received signals. Applying different deep neural
networks of encoder-decoder pairs on MNIST and CIFAR-10 image datasets, the
classifier accuracy is shown to increase with the number of channel uses at the
expense of longer service time. The peak age of task information (PAoTI) is
introduced to analyze this accuracy-latency tradeoff when the age grows unless
a received signal is classified correctly. By incorporating channel and traffic
effects, design guidelines are obtained for task-oriented communications by
characterizing how the PAoTI first decreases and then increases with the number
of channel uses. A dynamic update mechanism is presented to adapt the number of
channel uses to channel and traffic conditions, and reduce the PAoTI in
task-oriented communications.
| no_new_dataset | 0.948965 |
2301.10921 | Hao Chen | Hao Chen, Ran Tao, Yue Fan, Yidong Wang, Jindong Wang, Bernt Schiele,
Xing Xie, Bhiksha Raj, Marios Savvides | SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised
Learning | Accepted by ICLR 2023 | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | The critical challenge of Semi-Supervised Learning (SSL) is how to
effectively leverage the limited labeled data and massive unlabeled data to
improve the model's generalization performance. In this paper, we first revisit
the popular pseudo-labeling methods via a unified sample weighting formulation
and demonstrate the inherent quantity-quality trade-off problem of
pseudo-labeling with thresholding, which may prohibit learning. To this end, we
propose SoftMatch to overcome the trade-off by maintaining both high quantity
and high quality of pseudo-labels during training, effectively exploiting the
unlabeled data. We derive a truncated Gaussian function to weight samples based
on their confidence, which can be viewed as a soft version of the confidence
threshold. We further enhance the utilization of weakly-learned classes by
proposing a uniform alignment approach. In experiments, SoftMatch shows
substantial improvements across a wide variety of benchmarks, including image,
text, and imbalanced classification.
| [
{
"version": "v1",
"created": "Thu, 26 Jan 2023 03:53:25 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 15:49:07 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Chen",
"Hao",
""
],
[
"Tao",
"Ran",
""
],
[
"Fan",
"Yue",
""
],
[
"Wang",
"Yidong",
""
],
[
"Wang",
"Jindong",
""
],
[
"Schiele",
"Bernt",
""
],
[
"Xie",
"Xing",
""
],
[
"Raj",
"Bhiksha",
""
],
[
"Savvides",
"Marios",
""
]
] | TITLE: SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised
Learning
ABSTRACT: The critical challenge of Semi-Supervised Learning (SSL) is how to
effectively leverage the limited labeled data and massive unlabeled data to
improve the model's generalization performance. In this paper, we first revisit
the popular pseudo-labeling methods via a unified sample weighting formulation
and demonstrate the inherent quantity-quality trade-off problem of
pseudo-labeling with thresholding, which may prohibit learning. To this end, we
propose SoftMatch to overcome the trade-off by maintaining both high quantity
and high quality of pseudo-labels during training, effectively exploiting the
unlabeled data. We derive a truncated Gaussian function to weight samples based
on their confidence, which can be viewed as a soft version of the confidence
threshold. We further enhance the utilization of weakly-learned classes by
proposing a uniform alignment approach. In experiments, SoftMatch shows
substantial improvements across a wide variety of benchmarks, including image,
text, and imbalanced classification.
| no_new_dataset | 0.949716 |
2302.06232 | Ryumei Nakada | Ryumei Nakada, Halil Ibrahim Gulluk, Zhun Deng, Wenlong Ji, James Zou,
Linjun Zhang | Understanding Multimodal Contrastive Learning and Incorporating Unpaired
Data | 42 pages, 3 figures, accepted by AISTATS 2023; a link to GitHub
repository added, style corrected, acknowledgements section added | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Language-supervised vision models have recently attracted great attention in
computer vision. A common approach to build such models is to use contrastive
learning on paired data across the two modalities, as exemplified by
Contrastive Language-Image Pre-Training (CLIP). In this paper, under linear
representation settings, (i) we initiate the investigation of a general class
of nonlinear loss functions for multimodal contrastive learning (MMCL)
including CLIP loss and show its connection to singular value decomposition
(SVD). Namely, we show that each step of loss minimization by gradient descent
can be seen as performing SVD on a contrastive cross-covariance matrix. Based
on this insight, (ii) we analyze the performance of MMCL. We quantitatively
show that the feature learning ability of MMCL can be better than that of
unimodal contrastive learning applied to each modality even under the presence
of wrongly matched pairs. This characterizes the robustness of MMCL to noisy
data. Furthermore, when we have access to additional unpaired data, (iii) we
propose a new MMCL loss that incorporates additional unpaired datasets. We show
that the algorithm can detect the ground-truth pairs and improve performance by
fully exploiting unpaired datasets. The performance of the proposed algorithm
was verified by numerical experiments.
| [
{
"version": "v1",
"created": "Mon, 13 Feb 2023 10:11:05 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Feb 2023 00:56:16 GMT"
},
{
"version": "v3",
"created": "Wed, 15 Mar 2023 01:38:19 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Nakada",
"Ryumei",
""
],
[
"Gulluk",
"Halil Ibrahim",
""
],
[
"Deng",
"Zhun",
""
],
[
"Ji",
"Wenlong",
""
],
[
"Zou",
"James",
""
],
[
"Zhang",
"Linjun",
""
]
] | TITLE: Understanding Multimodal Contrastive Learning and Incorporating Unpaired
Data
ABSTRACT: Language-supervised vision models have recently attracted great attention in
computer vision. A common approach to build such models is to use contrastive
learning on paired data across the two modalities, as exemplified by
Contrastive Language-Image Pre-Training (CLIP). In this paper, under linear
representation settings, (i) we initiate the investigation of a general class
of nonlinear loss functions for multimodal contrastive learning (MMCL)
including CLIP loss and show its connection to singular value decomposition
(SVD). Namely, we show that each step of loss minimization by gradient descent
can be seen as performing SVD on a contrastive cross-covariance matrix. Based
on this insight, (ii) we analyze the performance of MMCL. We quantitatively
show that the feature learning ability of MMCL can be better than that of
unimodal contrastive learning applied to each modality even under the presence
of wrongly matched pairs. This characterizes the robustness of MMCL to noisy
data. Furthermore, when we have access to additional unpaired data, (iii) we
propose a new MMCL loss that incorporates additional unpaired datasets. We show
that the algorithm can detect the ground-truth pairs and improve performance by
fully exploiting unpaired datasets. The performance of the proposed algorithm
was verified by numerical experiments.
| no_new_dataset | 0.942612 |
2302.07322 | Changye Li | Changye Li, Weizhe Xu, Trevor Cohen, Martin Michalowski, and Serguei
Pakhomov | TRESTLE: Toolkit for Reproducible Execution of Speech, Text and Language
Experiments | Accepted at AMIA Informatics Summit | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The evidence is growing that machine and deep learning methods can learn the
subtle differences between the language produced by people with various forms
of cognitive impairment such as dementia and cognitively healthy individuals.
Valuable public data repositories such as TalkBank have made it possible for
researchers in the computational community to join forces and learn from each
other to make significant advances in this area. However, due to variability in
approaches and data selection strategies used by various researchers, results
obtained by different groups have been difficult to compare directly. In this
paper, we present TRESTLE (\textbf{T}oolkit for \textbf{R}eproducible
\textbf{E}xecution of \textbf{S}peech \textbf{T}ext and \textbf{L}anguage
\textbf{E}xperiments), an open source platform that focuses on two datasets
from the TalkBank repository with dementia detection as an illustrative domain.
Successfully deployed in the hackallenge (Hackathon/Challenge) of the
International Workshop on Health Intelligence at AAAI 2022, TRESTLE provides a
precise digital blueprint of the data pre-processing and selection strategies
that can be reused via TRESTLE by other researchers seeking comparable results
with their peers and current state-of-the-art (SOTA) approaches.
| [
{
"version": "v1",
"created": "Tue, 14 Feb 2023 20:07:31 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 19:35:05 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Li",
"Changye",
""
],
[
"Xu",
"Weizhe",
""
],
[
"Cohen",
"Trevor",
""
],
[
"Michalowski",
"Martin",
""
],
[
"Pakhomov",
"Serguei",
""
]
] | TITLE: TRESTLE: Toolkit for Reproducible Execution of Speech, Text and Language
Experiments
ABSTRACT: The evidence is growing that machine and deep learning methods can learn the
subtle differences between the language produced by people with various forms
of cognitive impairment such as dementia and cognitively healthy individuals.
Valuable public data repositories such as TalkBank have made it possible for
researchers in the computational community to join forces and learn from each
other to make significant advances in this area. However, due to variability in
approaches and data selection strategies used by various researchers, results
obtained by different groups have been difficult to compare directly. In this
paper, we present TRESTLE (\textbf{T}oolkit for \textbf{R}eproducible
\textbf{E}xecution of \textbf{S}peech \textbf{T}ext and \textbf{L}anguage
\textbf{E}xperiments), an open source platform that focuses on two datasets
from the TalkBank repository with dementia detection as an illustrative domain.
Successfully deployed in the hackallenge (Hackathon/Challenge) of the
International Workshop on Health Intelligence at AAAI 2022, TRESTLE provides a
precise digital blueprint of the data pre-processing and selection strategies
that can be reused via TRESTLE by other researchers seeking comparable results
with their peers and current state-of-the-art (SOTA) approaches.
| no_new_dataset | 0.945096 |
2302.10390 | Ke Yu | Ke Yu, Li Sun, Junxiang Chen, Max Reynolds, Tigmanshu Chaudhary,
Kayhan Batmanghelich | DrasCLR: A Self-supervised Framework of Learning Disease-related and
Anatomy-specific Representation for 3D Medical Images | Added some recent references | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large-scale volumetric medical images with annotation are rare, costly, and
time prohibitive to acquire. Self-supervised learning (SSL) offers a promising
pre-training and feature extraction solution for many downstream tasks, as it
only uses unlabeled data. Recently, SSL methods based on instance
discrimination have gained popularity in the medical imaging domain. However,
SSL pre-trained encoders may use many clues in the image to discriminate an
instance that are not necessarily disease-related. Moreover, pathological
patterns are often subtle and heterogeneous, requiring the ability of the
desired method to represent anatomy-specific features that are sensitive to
abnormal changes in different body parts. In this work, we present a novel SSL
framework, named DrasCLR, for 3D medical imaging to overcome these challenges.
We propose two domain-specific contrastive learning strategies: one aims to
capture subtle disease patterns inside a local anatomical region, and the other
aims to represent severe disease patterns that span larger regions. We
formulate the encoder using conditional hyper-parameterized network, in which
the parameters are dependant on the anatomical location, to extract
anatomically sensitive features. Extensive experiments on large-scale computer
tomography (CT) datasets of lung images show that our method improves the
performance of many downstream prediction and segmentation tasks. The
patient-level representation improves the performance of the patient survival
prediction task. We show how our method can detect emphysema subtypes via dense
prediction. We demonstrate that fine-tuning the pre-trained model can
significantly reduce annotation efforts without sacrificing emphysema detection
accuracy. Our ablation study highlights the importance of incorporating
anatomical context into the SSL framework.
| [
{
"version": "v1",
"created": "Tue, 21 Feb 2023 01:32:27 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 15:44:37 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Yu",
"Ke",
""
],
[
"Sun",
"Li",
""
],
[
"Chen",
"Junxiang",
""
],
[
"Reynolds",
"Max",
""
],
[
"Chaudhary",
"Tigmanshu",
""
],
[
"Batmanghelich",
"Kayhan",
""
]
] | TITLE: DrasCLR: A Self-supervised Framework of Learning Disease-related and
Anatomy-specific Representation for 3D Medical Images
ABSTRACT: Large-scale volumetric medical images with annotation are rare, costly, and
time prohibitive to acquire. Self-supervised learning (SSL) offers a promising
pre-training and feature extraction solution for many downstream tasks, as it
only uses unlabeled data. Recently, SSL methods based on instance
discrimination have gained popularity in the medical imaging domain. However,
SSL pre-trained encoders may use many clues in the image to discriminate an
instance that are not necessarily disease-related. Moreover, pathological
patterns are often subtle and heterogeneous, requiring the ability of the
desired method to represent anatomy-specific features that are sensitive to
abnormal changes in different body parts. In this work, we present a novel SSL
framework, named DrasCLR, for 3D medical imaging to overcome these challenges.
We propose two domain-specific contrastive learning strategies: one aims to
capture subtle disease patterns inside a local anatomical region, and the other
aims to represent severe disease patterns that span larger regions. We
formulate the encoder using conditional hyper-parameterized network, in which
the parameters are dependant on the anatomical location, to extract
anatomically sensitive features. Extensive experiments on large-scale computer
tomography (CT) datasets of lung images show that our method improves the
performance of many downstream prediction and segmentation tasks. The
patient-level representation improves the performance of the patient survival
prediction task. We show how our method can detect emphysema subtypes via dense
prediction. We demonstrate that fine-tuning the pre-trained model can
significantly reduce annotation efforts without sacrificing emphysema detection
accuracy. Our ablation study highlights the importance of incorporating
anatomical context into the SSL framework.
| no_new_dataset | 0.947817 |
2302.11217 | Paul Voigtlaender | Paul Voigtlaender and Soravit Changpinyo and Jordi Pont-Tuset and Radu
Soricut and Vittorio Ferrari | Connecting Vision and Language with Video Localized Narratives | Accepted at CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose Video Localized Narratives, a new form of multimodal video
annotations connecting vision and language. In the original Localized
Narratives, annotators speak and move their mouse simultaneously on an image,
thus grounding each word with a mouse trace segment. However, this is
challenging on a video. Our new protocol empowers annotators to tell the story
of a video with Localized Narratives, capturing even complex events involving
multiple actors interacting with each other and with several passive objects.
We annotated 20k videos of the OVIS, UVO, and Oops datasets, totalling 1.7M
words. Based on this data, we also construct new benchmarks for the video
narrative grounding and video question answering tasks, and provide reference
results from strong baseline models. Our annotations are available at
https://google.github.io/video-localized-narratives/.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 09:04:00 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 10:30:18 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Voigtlaender",
"Paul",
""
],
[
"Changpinyo",
"Soravit",
""
],
[
"Pont-Tuset",
"Jordi",
""
],
[
"Soricut",
"Radu",
""
],
[
"Ferrari",
"Vittorio",
""
]
] | TITLE: Connecting Vision and Language with Video Localized Narratives
ABSTRACT: We propose Video Localized Narratives, a new form of multimodal video
annotations connecting vision and language. In the original Localized
Narratives, annotators speak and move their mouse simultaneously on an image,
thus grounding each word with a mouse trace segment. However, this is
challenging on a video. Our new protocol empowers annotators to tell the story
of a video with Localized Narratives, capturing even complex events involving
multiple actors interacting with each other and with several passive objects.
We annotated 20k videos of the OVIS, UVO, and Oops datasets, totalling 1.7M
words. Based on this data, we also construct new benchmarks for the video
narrative grounding and video question answering tasks, and provide reference
results from strong baseline models. Our annotations are available at
https://google.github.io/video-localized-narratives/.
| no_new_dataset | 0.954351 |
2302.11799 | Qichen Ye | Qichen Ye, Bowen Cao, Nuo Chen, Weiyuan Xu, Yuexian Zou | FiTs: Fine-grained Two-stage Training for Knowledge-aware Question
Answering | Accepted in AAAI 2023, oral | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Knowledge-aware question answering (KAQA) requires the model to answer
questions over a knowledge base, which is essential for both open-domain QA and
domain-specific QA, especially when language models alone cannot provide all
the knowledge needed. Despite the promising result of recent KAQA systems which
tend to integrate linguistic knowledge from pre-trained language models (PLM)
and factual knowledge from knowledge graphs (KG) to answer complex questions, a
bottleneck exists in effectively fusing the representations from PLMs and KGs
because of (i) the semantic and distributional gaps between them, and (ii) the
difficulties in joint reasoning over the provided knowledge from both
modalities. To address the above two problems, we propose a Fine-grained
Two-stage training framework (FiTs) to boost the KAQA system performance: The
first stage aims at aligning representations from the PLM and the KG, thus
bridging the modality gaps between them, named knowledge adaptive
post-training. The second stage, called knowledge-aware fine-tuning, aims to
improve the model's joint reasoning ability based on the aligned
representations. In detail, we fine-tune the post-trained model via two
auxiliary self-supervised tasks in addition to the QA supervision. Extensive
experiments demonstrate that our approach achieves state-of-the-art performance
on three benchmarks in the commonsense reasoning (i.e., CommonsenseQA,
OpenbookQA) and medical question answering (i.e., MedQA-USMILE) domains.
| [
{
"version": "v1",
"created": "Thu, 23 Feb 2023 06:25:51 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 14:31:56 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Ye",
"Qichen",
""
],
[
"Cao",
"Bowen",
""
],
[
"Chen",
"Nuo",
""
],
[
"Xu",
"Weiyuan",
""
],
[
"Zou",
"Yuexian",
""
]
] | TITLE: FiTs: Fine-grained Two-stage Training for Knowledge-aware Question
Answering
ABSTRACT: Knowledge-aware question answering (KAQA) requires the model to answer
questions over a knowledge base, which is essential for both open-domain QA and
domain-specific QA, especially when language models alone cannot provide all
the knowledge needed. Despite the promising result of recent KAQA systems which
tend to integrate linguistic knowledge from pre-trained language models (PLM)
and factual knowledge from knowledge graphs (KG) to answer complex questions, a
bottleneck exists in effectively fusing the representations from PLMs and KGs
because of (i) the semantic and distributional gaps between them, and (ii) the
difficulties in joint reasoning over the provided knowledge from both
modalities. To address the above two problems, we propose a Fine-grained
Two-stage training framework (FiTs) to boost the KAQA system performance: The
first stage aims at aligning representations from the PLM and the KG, thus
bridging the modality gaps between them, named knowledge adaptive
post-training. The second stage, called knowledge-aware fine-tuning, aims to
improve the model's joint reasoning ability based on the aligned
representations. In detail, we fine-tune the post-trained model via two
auxiliary self-supervised tasks in addition to the QA supervision. Extensive
experiments demonstrate that our approach achieves state-of-the-art performance
on three benchmarks in the commonsense reasoning (i.e., CommonsenseQA,
OpenbookQA) and medical question answering (i.e., MedQA-USMILE) domains.
| no_new_dataset | 0.949576 |
2302.11814 | Qichen Ye | Bowen Cao, Qichen Ye, Weiyuan Xu, Yuexian Zou | FTM: A Frame-level Timeline Modeling Method for Temporal Graph
Representation Learning | Accepted in AAAI 2023, oral | null | null | null | cs.LG cs.SI | http://creativecommons.org/licenses/by/4.0/ | Learning representations for graph-structured data is essential for graph
analytical tasks. While remarkable progress has been made on static graphs,
researches on temporal graphs are still in its beginning stage. The bottleneck
of the temporal graph representation learning approach is the neighborhood
aggregation strategy, based on which graph attributes share and gather
information explicitly. Existing neighborhood aggregation strategies fail to
capture either the short-term features or the long-term features of temporal
graph attributes, leading to unsatisfactory model performance and even poor
robustness and domain generality of the representation learning method. To
address this problem, we propose a Frame-level Timeline Modeling (FTM) method
that helps to capture both short-term and long-term features and thus learns
more informative representations on temporal graphs. In particular, we present
a novel link-based framing technique to preserve the short-term features and
then incorporate a timeline aggregator module to capture the intrinsic dynamics
of graph evolution as long-term features. Our method can be easily assembled
with most temporal GNNs. Extensive experiments on common datasets show that our
method brings great improvements to the capability, robustness, and domain
generality of backbone methods in downstream tasks. Our code can be found at
https://github.com/yeeeqichen/FTM.
| [
{
"version": "v1",
"created": "Thu, 23 Feb 2023 06:53:16 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 14:35:50 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Cao",
"Bowen",
""
],
[
"Ye",
"Qichen",
""
],
[
"Xu",
"Weiyuan",
""
],
[
"Zou",
"Yuexian",
""
]
] | TITLE: FTM: A Frame-level Timeline Modeling Method for Temporal Graph
Representation Learning
ABSTRACT: Learning representations for graph-structured data is essential for graph
analytical tasks. While remarkable progress has been made on static graphs,
researches on temporal graphs are still in its beginning stage. The bottleneck
of the temporal graph representation learning approach is the neighborhood
aggregation strategy, based on which graph attributes share and gather
information explicitly. Existing neighborhood aggregation strategies fail to
capture either the short-term features or the long-term features of temporal
graph attributes, leading to unsatisfactory model performance and even poor
robustness and domain generality of the representation learning method. To
address this problem, we propose a Frame-level Timeline Modeling (FTM) method
that helps to capture both short-term and long-term features and thus learns
more informative representations on temporal graphs. In particular, we present
a novel link-based framing technique to preserve the short-term features and
then incorporate a timeline aggregator module to capture the intrinsic dynamics
of graph evolution as long-term features. Our method can be easily assembled
with most temporal GNNs. Extensive experiments on common datasets show that our
method brings great improvements to the capability, robustness, and domain
generality of backbone methods in downstream tasks. Our code can be found at
https://github.com/yeeeqichen/FTM.
| no_new_dataset | 0.942876 |
2302.14680 | Holy Lovenia | Holy Lovenia, Samuel Cahyawijaya, Pascale Fung | Which One Are You Referring To? Multimodal Object Identification in
Situated Dialogue | Accepted at EACL SRW 2023 | null | null | null | cs.CL cs.AI cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | The demand for multimodal dialogue systems has been rising in various
domains, emphasizing the importance of interpreting multimodal inputs from
conversational and situational contexts. We explore three methods to tackle
this problem and evaluate them on the largest situated dialogue dataset, SIMMC
2.1. Our best method, scene-dialogue alignment, improves the performance by
~20% F1-score compared to the SIMMC 2.1 baselines. We provide analysis and
discussion regarding the limitation of our methods and the potential directions
for future works. Our code is publicly available at
https://github.com/holylovenia/multimodal-object-identification.
| [
{
"version": "v1",
"created": "Tue, 28 Feb 2023 15:45:20 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 14:38:21 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Lovenia",
"Holy",
""
],
[
"Cahyawijaya",
"Samuel",
""
],
[
"Fung",
"Pascale",
""
]
] | TITLE: Which One Are You Referring To? Multimodal Object Identification in
Situated Dialogue
ABSTRACT: The demand for multimodal dialogue systems has been rising in various
domains, emphasizing the importance of interpreting multimodal inputs from
conversational and situational contexts. We explore three methods to tackle
this problem and evaluate them on the largest situated dialogue dataset, SIMMC
2.1. Our best method, scene-dialogue alignment, improves the performance by
~20% F1-score compared to the SIMMC 2.1 baselines. We provide analysis and
discussion regarding the limitation of our methods and the potential directions
for future works. Our code is publicly available at
https://github.com/holylovenia/multimodal-object-identification.
| no_new_dataset | 0.884688 |
2303.00264 | Li Zhang | Li Zhang, Qing Wang, Hongji Wang, Yue Li, Wei Rao, Yannan Wang, Lei
Xie | Distance-based Weight Transfer from Near-field to Far-field Speaker
Verification | null | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | The scarcity of labeled far-field speech is a constraint for training
superior far-field speaker verification systems. Fine-tuning the model
pre-trained on large-scale near-field speech substantially outperforms training
from scratch. However, the fine-tuning method suffers from two
limitations--catastrophic forgetting and overfitting. In this paper, we propose
a weight transfer regularization(WTR) loss to constrain the distance of the
weights between the pre-trained model with large-scale near-field speech and
the fine-tuned model through a small number of far-field speech. With the WTR
loss, the fine-tuning process takes advantage of the previously acquired
discriminative ability from the large-scale near-field speech without
catastrophic forgetting. Meanwhile, we use the PAC-Bayes generalization theory
to analyze the generalization bound of the fine-tuned model with the WTR loss.
The analysis result indicates that the WTR term makes the fine-tuned model have
a tighter generalization upper bound. Moreover, we explore three kinds of norm
distance for weight transfer, which are L1-norm distance, L2-norm distance and
Max-norm distance. Finally, we evaluate the effectiveness of the WTR loss on
VoxCeleb (pre-trained dataset) and FFSVC (fine-tuned dataset) datasets.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 06:38:02 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 03:05:35 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zhang",
"Li",
""
],
[
"Wang",
"Qing",
""
],
[
"Wang",
"Hongji",
""
],
[
"Li",
"Yue",
""
],
[
"Rao",
"Wei",
""
],
[
"Wang",
"Yannan",
""
],
[
"Xie",
"Lei",
""
]
] | TITLE: Distance-based Weight Transfer from Near-field to Far-field Speaker
Verification
ABSTRACT: The scarcity of labeled far-field speech is a constraint for training
superior far-field speaker verification systems. Fine-tuning the model
pre-trained on large-scale near-field speech substantially outperforms training
from scratch. However, the fine-tuning method suffers from two
limitations--catastrophic forgetting and overfitting. In this paper, we propose
a weight transfer regularization(WTR) loss to constrain the distance of the
weights between the pre-trained model with large-scale near-field speech and
the fine-tuned model through a small number of far-field speech. With the WTR
loss, the fine-tuning process takes advantage of the previously acquired
discriminative ability from the large-scale near-field speech without
catastrophic forgetting. Meanwhile, we use the PAC-Bayes generalization theory
to analyze the generalization bound of the fine-tuned model with the WTR loss.
The analysis result indicates that the WTR term makes the fine-tuned model have
a tighter generalization upper bound. Moreover, we explore three kinds of norm
distance for weight transfer, which are L1-norm distance, L2-norm distance and
Max-norm distance. Finally, we evaluate the effectiveness of the WTR loss on
VoxCeleb (pre-trained dataset) and FFSVC (fine-tuned dataset) datasets.
| no_new_dataset | 0.951323 |
2303.01046 | Daizong Liu | Daizong Liu, Pan Zhou | Jointly Visual- and Semantic-Aware Graph Memory Networks for Temporal
Sentence Localization in Videos | Accepted by ICASSP2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Temporal sentence localization in videos (TSLV) aims to retrieve the most
interested segment in an untrimmed video according to a given sentence query.
However, almost of existing TSLV approaches suffer from the same limitations:
(1) They only focus on either frame-level or object-level visual representation
learning and corresponding correlation reasoning, but fail to integrate them
both; (2) They neglect to leverage the rich semantic contexts to further
benefit the query reasoning. To address these issues, in this paper, we propose
a novel Hierarchical Visual- and Semantic-Aware Reasoning Network (HVSARN),
which enables both visual- and semantic-aware query reasoning from object-level
to frame-level. Specifically, we present a new graph memory mechanism to
perform visual-semantic query reasoning: For visual reasoning, we design a
visual graph memory to leverage visual information of video; For semantic
reasoning, a semantic graph memory is also introduced to explicitly leverage
semantic knowledge contained in the classes and attributes of video objects,
and perform correlation reasoning in the semantic space. Experiments on three
datasets demonstrate that our HVSARN achieves a new state-of-the-art
performance.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 08:00:22 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 03:10:39 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Liu",
"Daizong",
""
],
[
"Zhou",
"Pan",
""
]
] | TITLE: Jointly Visual- and Semantic-Aware Graph Memory Networks for Temporal
Sentence Localization in Videos
ABSTRACT: Temporal sentence localization in videos (TSLV) aims to retrieve the most
interested segment in an untrimmed video according to a given sentence query.
However, almost of existing TSLV approaches suffer from the same limitations:
(1) They only focus on either frame-level or object-level visual representation
learning and corresponding correlation reasoning, but fail to integrate them
both; (2) They neglect to leverage the rich semantic contexts to further
benefit the query reasoning. To address these issues, in this paper, we propose
a novel Hierarchical Visual- and Semantic-Aware Reasoning Network (HVSARN),
which enables both visual- and semantic-aware query reasoning from object-level
to frame-level. Specifically, we present a new graph memory mechanism to
perform visual-semantic query reasoning: For visual reasoning, we design a
visual graph memory to leverage visual information of video; For semantic
reasoning, a semantic graph memory is also introduced to explicitly leverage
semantic knowledge contained in the classes and attributes of video objects,
and perform correlation reasoning in the semantic space. Experiments on three
datasets demonstrate that our HVSARN achieves a new state-of-the-art
performance.
| no_new_dataset | 0.952706 |
2303.02328 | Sangrok Lee | Sangrok Lee, Jongseong Bae, Ha Young Kim | Decompose, Adjust, Compose: Effective Normalization by Playing with
Frequency for Domain Generalization | 10 pages,6 figures, Conference on Computer Vision and Pattern
Recognition 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Domain generalization (DG) is a principal task to evaluate the robustness of
computer vision models. Many previous studies have used normalization for DG.
In normalization, statistics and normalized features are regarded as style and
content, respectively. However, it has a content variation problem when
removing style because the boundary between content and style is unclear. This
study addresses this problem from the frequency domain perspective, where
amplitude and phase are considered as style and content, respectively. First,
we verify the quantitative phase variation of normalization through the
mathematical derivation of the Fourier transform formula. Then, based on this,
we propose a novel normalization method, PCNorm, which eliminates style only as
the preserving content through spectral decomposition. Furthermore, we propose
advanced PCNorm variants, CCNorm and SCNorm, which adjust the degrees of
variations in content and style, respectively. Thus, they can learn
domain-agnostic representations for DG. With the normalization methods, we
propose ResNet-variant models, DAC-P and DAC-SC, which are robust to the domain
gap. The proposed models outperform other recent DG methods. The DAC-SC
achieves an average state-of-the-art performance of 65.6% on five datasets:
PACS, VLCS, Office-Home, DomainNet, and TerraIncognita.
| [
{
"version": "v1",
"created": "Sat, 4 Mar 2023 05:23:11 GMT"
},
{
"version": "v2",
"created": "Mon, 13 Mar 2023 16:04:17 GMT"
},
{
"version": "v3",
"created": "Wed, 15 Mar 2023 12:39:19 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Lee",
"Sangrok",
""
],
[
"Bae",
"Jongseong",
""
],
[
"Kim",
"Ha Young",
""
]
] | TITLE: Decompose, Adjust, Compose: Effective Normalization by Playing with
Frequency for Domain Generalization
ABSTRACT: Domain generalization (DG) is a principal task to evaluate the robustness of
computer vision models. Many previous studies have used normalization for DG.
In normalization, statistics and normalized features are regarded as style and
content, respectively. However, it has a content variation problem when
removing style because the boundary between content and style is unclear. This
study addresses this problem from the frequency domain perspective, where
amplitude and phase are considered as style and content, respectively. First,
we verify the quantitative phase variation of normalization through the
mathematical derivation of the Fourier transform formula. Then, based on this,
we propose a novel normalization method, PCNorm, which eliminates style only as
the preserving content through spectral decomposition. Furthermore, we propose
advanced PCNorm variants, CCNorm and SCNorm, which adjust the degrees of
variations in content and style, respectively. Thus, they can learn
domain-agnostic representations for DG. With the normalization methods, we
propose ResNet-variant models, DAC-P and DAC-SC, which are robust to the domain
gap. The proposed models outperform other recent DG methods. The DAC-SC
achieves an average state-of-the-art performance of 65.6% on five datasets:
PACS, VLCS, Office-Home, DomainNet, and TerraIncognita.
| no_new_dataset | 0.943556 |
2303.02489 | Yanxin Long | Yanxin Long, Youpeng Wen, Jianhua Han, Hang Xu, Pengzhen Ren, Wei
Zhang, Shen Zhao, Xiaodan Liang | CapDet: Unifying Dense Captioning and Open-World Detection Pretraining | Accepted by CVPR2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Benefiting from large-scale vision-language pre-training on image-text pairs,
open-world detection methods have shown superior generalization ability under
the zero-shot or few-shot detection settings. However, a pre-defined category
space is still required during the inference stage of existing methods and only
the objects belonging to that space will be predicted. To introduce a "real"
open-world detector, in this paper, we propose a novel method named CapDet to
either predict under a given category list or directly generate the category of
predicted bounding boxes. Specifically, we unify the open-world detection and
dense caption tasks into a single yet effective framework by introducing an
additional dense captioning head to generate the region-grounded captions.
Besides, adding the captioning task will in turn benefit the generalization of
detection performance since the captioning dataset covers more concepts.
Experiment results show that by unifying the dense caption task, our CapDet has
obtained significant performance improvements (e.g., +2.1% mAP on LVIS rare
classes) over the baseline method on LVIS (1203 classes). Besides, our CapDet
also achieves state-of-the-art performance on dense captioning tasks, e.g.,
15.44% mAP on VG V1.2 and 13.98% on the VG-COCO dataset.
| [
{
"version": "v1",
"created": "Sat, 4 Mar 2023 19:53:00 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 14:25:04 GMT"
},
{
"version": "v3",
"created": "Wed, 15 Mar 2023 13:45:48 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Long",
"Yanxin",
""
],
[
"Wen",
"Youpeng",
""
],
[
"Han",
"Jianhua",
""
],
[
"Xu",
"Hang",
""
],
[
"Ren",
"Pengzhen",
""
],
[
"Zhang",
"Wei",
""
],
[
"Zhao",
"Shen",
""
],
[
"Liang",
"Xiaodan",
""
]
] | TITLE: CapDet: Unifying Dense Captioning and Open-World Detection Pretraining
ABSTRACT: Benefiting from large-scale vision-language pre-training on image-text pairs,
open-world detection methods have shown superior generalization ability under
the zero-shot or few-shot detection settings. However, a pre-defined category
space is still required during the inference stage of existing methods and only
the objects belonging to that space will be predicted. To introduce a "real"
open-world detector, in this paper, we propose a novel method named CapDet to
either predict under a given category list or directly generate the category of
predicted bounding boxes. Specifically, we unify the open-world detection and
dense caption tasks into a single yet effective framework by introducing an
additional dense captioning head to generate the region-grounded captions.
Besides, adding the captioning task will in turn benefit the generalization of
detection performance since the captioning dataset covers more concepts.
Experiment results show that by unifying the dense caption task, our CapDet has
obtained significant performance improvements (e.g., +2.1% mAP on LVIS rare
classes) over the baseline method on LVIS (1203 classes). Besides, our CapDet
also achieves state-of-the-art performance on dense captioning tasks, e.g.,
15.44% mAP on VG V1.2 and 13.98% on the VG-COCO dataset.
| no_new_dataset | 0.942188 |
2303.06314 | Jian Xu | Jian Xu, Meiling Yang, Wenbo Ding, Shao-Lun Huang | Stabilizing and Improving Federated Learning with Non-IID Data and
Client Dropout | This work has been submitted to the IEEE for possible publication | null | null | null | cs.LG cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The label distribution skew induced data heterogeniety has been shown to be a
significant obstacle that limits the model performance in federated learning,
which is particularly developed for collaborative model training over
decentralized data sources while preserving user privacy. This challenge could
be more serious when the participating clients are in unstable circumstances
and dropout frequently. Previous work and our empirical observations
demonstrate that the classifier head for classification task is more sensitive
to label skew and the unstable performance of FedAvg mainly lies in the
imbalanced training samples across different classes. The biased classifier
head will also impact the learning of feature representations. Therefore,
maintaining a balanced classifier head is of significant importance for
building a better global model. To this end, we propose a simple yet effective
framework by introducing a prior-calibrated softmax function for computing the
cross-entropy loss and a prototype-based feature augmentation scheme to
re-balance the local training, which are lightweight for edge devices and can
facilitate the global model aggregation. The improved model performance over
existing baselines in the presence of non-IID data and client dropout is
demonstrated by conducting extensive experiments on benchmark classification
tasks.
| [
{
"version": "v1",
"created": "Sat, 11 Mar 2023 05:17:59 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 17:30:20 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Xu",
"Jian",
""
],
[
"Yang",
"Meiling",
""
],
[
"Ding",
"Wenbo",
""
],
[
"Huang",
"Shao-Lun",
""
]
] | TITLE: Stabilizing and Improving Federated Learning with Non-IID Data and
Client Dropout
ABSTRACT: The label distribution skew induced data heterogeniety has been shown to be a
significant obstacle that limits the model performance in federated learning,
which is particularly developed for collaborative model training over
decentralized data sources while preserving user privacy. This challenge could
be more serious when the participating clients are in unstable circumstances
and dropout frequently. Previous work and our empirical observations
demonstrate that the classifier head for classification task is more sensitive
to label skew and the unstable performance of FedAvg mainly lies in the
imbalanced training samples across different classes. The biased classifier
head will also impact the learning of feature representations. Therefore,
maintaining a balanced classifier head is of significant importance for
building a better global model. To this end, we propose a simple yet effective
framework by introducing a prior-calibrated softmax function for computing the
cross-entropy loss and a prototype-based feature augmentation scheme to
re-balance the local training, which are lightweight for edge devices and can
facilitate the global model aggregation. The improved model performance over
existing baselines in the presence of non-IID data and client dropout is
demonstrated by conducting extensive experiments on benchmark classification
tasks.
| no_new_dataset | 0.940953 |
2303.07641 | Nam Ly Tuan | Nam Tuan Ly, Atsuhiro Takasu, Phuc Nguyen, and Hideaki Takeda | Rethinking Image-based Table Recognition Using Weakly Supervised Methods | 10 pages, ICPRAM2023 | ICPRAM2023, pages 872-880, 2023 | 10.5220/0011682600003411 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most of the previous methods for table recognition rely on training datasets
containing many richly annotated table images. Detailed table image annotation,
e.g., cell or text bounding box annotation, however, is costly and often
subjective. In this paper, we propose a weakly supervised model named WSTabNet
for table recognition that relies only on HTML (or LaTeX) code-level
annotations of table images. The proposed model consists of three main parts:
an encoder for feature extraction, a structure decoder for generating table
structure, and a cell decoder for predicting the content of each cell in the
table. Our system is trained end-to-end by stochastic gradient descent
algorithms, requiring only table images and their ground-truth HTML (or LaTeX)
representations. To facilitate table recognition with deep learning, we create
and release WikiTableSet, the largest publicly available image-based table
recognition dataset built from Wikipedia. WikiTableSet contains nearly 4
million English table images, 590K Japanese table images, and 640k French table
images with corresponding HTML representation and cell bounding boxes. The
extensive experiments on WikiTableSet and two large-scale datasets: FinTabNet
and PubTabNet demonstrate that the proposed weakly supervised model achieves
better, or similar accuracies compared to the state-of-the-art models on all
benchmark datasets.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 06:03:57 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Ly",
"Nam Tuan",
""
],
[
"Takasu",
"Atsuhiro",
""
],
[
"Nguyen",
"Phuc",
""
],
[
"Takeda",
"Hideaki",
""
]
] | TITLE: Rethinking Image-based Table Recognition Using Weakly Supervised Methods
ABSTRACT: Most of the previous methods for table recognition rely on training datasets
containing many richly annotated table images. Detailed table image annotation,
e.g., cell or text bounding box annotation, however, is costly and often
subjective. In this paper, we propose a weakly supervised model named WSTabNet
for table recognition that relies only on HTML (or LaTeX) code-level
annotations of table images. The proposed model consists of three main parts:
an encoder for feature extraction, a structure decoder for generating table
structure, and a cell decoder for predicting the content of each cell in the
table. Our system is trained end-to-end by stochastic gradient descent
algorithms, requiring only table images and their ground-truth HTML (or LaTeX)
representations. To facilitate table recognition with deep learning, we create
and release WikiTableSet, the largest publicly available image-based table
recognition dataset built from Wikipedia. WikiTableSet contains nearly 4
million English table images, 590K Japanese table images, and 640k French table
images with corresponding HTML representation and cell bounding boxes. The
extensive experiments on WikiTableSet and two large-scale datasets: FinTabNet
and PubTabNet demonstrate that the proposed weakly supervised model achieves
better, or similar accuracies compared to the state-of-the-art models on all
benchmark datasets.
| new_dataset | 0.959193 |
2303.07938 | Zhaoyang Lyu | Zhaoyang Lyu, Jinyi Wang, Yuwei An, Ya Zhang, Dahua Lin, Bo Dai | Controllable Mesh Generation Through Sparse Latent Point Diffusion
Models | Accepted to CVPR 2023. Project page is at https://slide-3d.github.io | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Mesh generation is of great value in various applications involving computer
graphics and virtual content, yet designing generative models for meshes is
challenging due to their irregular data structure and inconsistent topology of
meshes in the same category. In this work, we design a novel sparse latent
point diffusion model for mesh generation. Our key insight is to regard point
clouds as an intermediate representation of meshes, and model the distribution
of point clouds instead. While meshes can be generated from point clouds via
techniques like Shape as Points (SAP), the challenges of directly generating
meshes can be effectively avoided. To boost the efficiency and controllability
of our mesh generation method, we propose to further encode point clouds to a
set of sparse latent points with point-wise semantic meaningful features, where
two DDPMs are trained in the space of sparse latent points to respectively
model the distribution of the latent point positions and features at these
latent points. We find that sampling in this latent space is faster than
directly sampling dense point clouds. Moreover, the sparse latent points also
enable us to explicitly control both the overall structures and local details
of the generated meshes. Extensive experiments are conducted on the ShapeNet
dataset, where our proposed sparse latent point diffusion model achieves
superior performance in terms of generation quality and controllability when
compared to existing methods.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 14:25:29 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 03:13:08 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Lyu",
"Zhaoyang",
""
],
[
"Wang",
"Jinyi",
""
],
[
"An",
"Yuwei",
""
],
[
"Zhang",
"Ya",
""
],
[
"Lin",
"Dahua",
""
],
[
"Dai",
"Bo",
""
]
] | TITLE: Controllable Mesh Generation Through Sparse Latent Point Diffusion
Models
ABSTRACT: Mesh generation is of great value in various applications involving computer
graphics and virtual content, yet designing generative models for meshes is
challenging due to their irregular data structure and inconsistent topology of
meshes in the same category. In this work, we design a novel sparse latent
point diffusion model for mesh generation. Our key insight is to regard point
clouds as an intermediate representation of meshes, and model the distribution
of point clouds instead. While meshes can be generated from point clouds via
techniques like Shape as Points (SAP), the challenges of directly generating
meshes can be effectively avoided. To boost the efficiency and controllability
of our mesh generation method, we propose to further encode point clouds to a
set of sparse latent points with point-wise semantic meaningful features, where
two DDPMs are trained in the space of sparse latent points to respectively
model the distribution of the latent point positions and features at these
latent points. We find that sampling in this latent space is faster than
directly sampling dense point clouds. Moreover, the sparse latent points also
enable us to explicitly control both the overall structures and local details
of the generated meshes. Extensive experiments are conducted on the ShapeNet
dataset, where our proposed sparse latent point diffusion model achieves
superior performance in terms of generation quality and controllability when
compared to existing methods.
| no_new_dataset | 0.951459 |
2303.08156 | Fei Zhu | Tingting Fang, Fei Zhu and Jie Chen | Nonlinear Hyperspectral Unmixing based on Multilinear Mixing Model using
Convolutional Autoencoders | null | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Unsupervised spectral unmixing consists of representing each observed pixel
as a combination of several pure materials called endmembers with their
corresponding abundance fractions. Beyond the linear assumption, various
nonlinear unmixing models have been proposed, with the associated optimization
problems solved either by traditional optimization algorithms or deep learning
techniques. Current deep learning-based nonlinear unmixing focuses on the
models in additive, bilinear-based formulations. By interpreting the reflection
process using the discrete Markov chain, the multilinear mixing model (MLM)
successfully accounts for the up to infinite-order interactions between
endmembers. However, to simulate the physics process of MLM by neural networks
explicitly is a challenging problem that has not been approached by far. In
this article, we propose a novel autoencoder-based network for unsupervised
unmixing based on MLM. Benefitting from an elaborate network design, the
relationships among all the model parameters {\em i.e.}, endmembers,
abundances, and transition probability parameters are explicitly modeled. There
are two modes: MLM-1DAE considers only pixel-wise spectral information, and
MLM-3DAE exploits the spectral-spatial correlations within input patches.
Experiments on both the synthetic and real datasets demonstrate the
effectiveness of the proposed method as it achieves competitive performance to
the classic solutions of MLM.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 18:11:52 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Fang",
"Tingting",
""
],
[
"Zhu",
"Fei",
""
],
[
"Chen",
"Jie",
""
]
] | TITLE: Nonlinear Hyperspectral Unmixing based on Multilinear Mixing Model using
Convolutional Autoencoders
ABSTRACT: Unsupervised spectral unmixing consists of representing each observed pixel
as a combination of several pure materials called endmembers with their
corresponding abundance fractions. Beyond the linear assumption, various
nonlinear unmixing models have been proposed, with the associated optimization
problems solved either by traditional optimization algorithms or deep learning
techniques. Current deep learning-based nonlinear unmixing focuses on the
models in additive, bilinear-based formulations. By interpreting the reflection
process using the discrete Markov chain, the multilinear mixing model (MLM)
successfully accounts for the up to infinite-order interactions between
endmembers. However, to simulate the physics process of MLM by neural networks
explicitly is a challenging problem that has not been approached by far. In
this article, we propose a novel autoencoder-based network for unsupervised
unmixing based on MLM. Benefitting from an elaborate network design, the
relationships among all the model parameters {\em i.e.}, endmembers,
abundances, and transition probability parameters are explicitly modeled. There
are two modes: MLM-1DAE considers only pixel-wise spectral information, and
MLM-3DAE exploits the spectral-spatial correlations within input patches.
Experiments on both the synthetic and real datasets demonstrate the
effectiveness of the proposed method as it achieves competitive performance to
the classic solutions of MLM.
| no_new_dataset | 0.946448 |
2303.08207 | Arslan Chaudhry | Jiefeng Chen, Timothy Nguyen, Dilan Gorur, Arslan Chaudhry | Is forgetting less a good inductive bias for forward transfer? | Published as a conference paper at ICLR 2023 | ICLR 2023 | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | One of the main motivations of studying continual learning is that the
problem setting allows a model to accrue knowledge from past tasks to learn new
tasks more efficiently. However, recent studies suggest that the key metric
that continual learning algorithms optimize, reduction in catastrophic
forgetting, does not correlate well with the forward transfer of knowledge. We
believe that the conclusion previous works reached is due to the way they
measure forward transfer. We argue that the measure of forward transfer to a
task should not be affected by the restrictions placed on the continual learner
in order to preserve knowledge of previous tasks. Instead, forward transfer
should be measured by how easy it is to learn a new task given a set of
representations produced by continual learning on previous tasks. Under this
notion of forward transfer, we evaluate different continual learning algorithms
on a variety of image classification benchmarks. Our results indicate that less
forgetful representations lead to a better forward transfer suggesting a strong
correlation between retaining past information and learning efficiency on new
tasks. Further, we found less forgetful representations to be more diverse and
discriminative compared to their forgetful counterparts.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 19:52:09 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Chen",
"Jiefeng",
""
],
[
"Nguyen",
"Timothy",
""
],
[
"Gorur",
"Dilan",
""
],
[
"Chaudhry",
"Arslan",
""
]
] | TITLE: Is forgetting less a good inductive bias for forward transfer?
ABSTRACT: One of the main motivations of studying continual learning is that the
problem setting allows a model to accrue knowledge from past tasks to learn new
tasks more efficiently. However, recent studies suggest that the key metric
that continual learning algorithms optimize, reduction in catastrophic
forgetting, does not correlate well with the forward transfer of knowledge. We
believe that the conclusion previous works reached is due to the way they
measure forward transfer. We argue that the measure of forward transfer to a
task should not be affected by the restrictions placed on the continual learner
in order to preserve knowledge of previous tasks. Instead, forward transfer
should be measured by how easy it is to learn a new task given a set of
representations produced by continual learning on previous tasks. Under this
notion of forward transfer, we evaluate different continual learning algorithms
on a variety of image classification benchmarks. Our results indicate that less
forgetful representations lead to a better forward transfer suggesting a strong
correlation between retaining past information and learning efficiency on new
tasks. Further, we found less forgetful representations to be more diverse and
discriminative compared to their forgetful counterparts.
| no_new_dataset | 0.935759 |
2303.08216 | Nikhil Dhinagar | Nikhil J. Dhinagar, Sophia I. Thomopoulos, Emily Laltoo and Paul M.
Thompson | Efficiently Training Vision Transformers on Structural MRI Scans for
Alzheimer's Disease Detection | null | null | null | null | eess.IV cs.AI cs.CV cs.LG q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neuroimaging of large populations is valuable to identify factors that
promote or resist brain disease, and to assist diagnosis, subtyping, and
prognosis. Data-driven models such as convolutional neural networks (CNNs) have
increasingly been applied to brain images to perform diagnostic and prognostic
tasks by learning robust features. Vision transformers (ViT) - a new class of
deep learning architectures - have emerged in recent years as an alternative to
CNNs for several computer vision applications. Here we tested variants of the
ViT architecture for a range of desired neuroimaging downstream tasks based on
difficulty, in this case for sex and Alzheimer's disease (AD) classification
based on 3D brain MRI. In our experiments, two vision transformer architecture
variants achieved an AUC of 0.987 for sex and 0.892 for AD classification,
respectively. We independently evaluated our models on data from two benchmark
AD datasets. We achieved a performance boost of 5% and 9-10% upon fine-tuning
vision transformer models pre-trained on synthetic (generated by a latent
diffusion model) and real MRI scans, respectively. Our main contributions
include testing the effects of different ViT training strategies including
pre-training, data augmentation and learning rate warm-ups followed by
annealing, as pertaining to the neuroimaging domain. These techniques are
essential for training ViT-like models for neuroimaging applications where
training data is usually limited. We also analyzed the effect of the amount of
training data utilized on the test-time performance of the ViT via data-model
scaling curves.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 20:18:12 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Dhinagar",
"Nikhil J.",
""
],
[
"Thomopoulos",
"Sophia I.",
""
],
[
"Laltoo",
"Emily",
""
],
[
"Thompson",
"Paul M.",
""
]
] | TITLE: Efficiently Training Vision Transformers on Structural MRI Scans for
Alzheimer's Disease Detection
ABSTRACT: Neuroimaging of large populations is valuable to identify factors that
promote or resist brain disease, and to assist diagnosis, subtyping, and
prognosis. Data-driven models such as convolutional neural networks (CNNs) have
increasingly been applied to brain images to perform diagnostic and prognostic
tasks by learning robust features. Vision transformers (ViT) - a new class of
deep learning architectures - have emerged in recent years as an alternative to
CNNs for several computer vision applications. Here we tested variants of the
ViT architecture for a range of desired neuroimaging downstream tasks based on
difficulty, in this case for sex and Alzheimer's disease (AD) classification
based on 3D brain MRI. In our experiments, two vision transformer architecture
variants achieved an AUC of 0.987 for sex and 0.892 for AD classification,
respectively. We independently evaluated our models on data from two benchmark
AD datasets. We achieved a performance boost of 5% and 9-10% upon fine-tuning
vision transformer models pre-trained on synthetic (generated by a latent
diffusion model) and real MRI scans, respectively. Our main contributions
include testing the effects of different ViT training strategies including
pre-training, data augmentation and learning rate warm-ups followed by
annealing, as pertaining to the neuroimaging domain. These techniques are
essential for training ViT-like models for neuroimaging applications where
training data is usually limited. We also analyzed the effect of the amount of
training data utilized on the test-time performance of the ViT via data-model
scaling curves.
| no_new_dataset | 0.953013 |
2303.08225 | Hao Tang | Hao Tang, Zhenyu Zhang, Humphrey Shi, Bo Li, Ling Shao, Nicu Sebe,
Radu Timofte, Luc Van Gool | Graph Transformer GANs for Graph-Constrained House Generation | CVPR 2023 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel graph Transformer generative adversarial network (GTGAN)
to learn effective graph node relations in an end-to-end fashion for the
challenging graph-constrained house generation task. The proposed
graph-Transformer-based generator includes a novel graph Transformer encoder
that combines graph convolutions and self-attentions in a Transformer to model
both local and global interactions across connected and non-connected graph
nodes. Specifically, the proposed connected node attention (CNA) and
non-connected node attention (NNA) aim to capture the global relations across
connected nodes and non-connected nodes in the input graph, respectively. The
proposed graph modeling block (GMB) aims to exploit local vertex interactions
based on a house layout topology. Moreover, we propose a new node
classification-based discriminator to preserve the high-level semantic and
discriminative node features for different house components. Finally, we
propose a novel graph-based cycle-consistency loss that aims at maintaining the
relative spatial relationships between ground truth and predicted graphs.
Experiments on two challenging graph-constrained house generation tasks (i.e.,
house layout and roof generation) with two public datasets demonstrate the
effectiveness of GTGAN in terms of objective quantitative scores and subjective
visual realism. New state-of-the-art results are established by large margins
on both tasks.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 20:35:45 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Tang",
"Hao",
""
],
[
"Zhang",
"Zhenyu",
""
],
[
"Shi",
"Humphrey",
""
],
[
"Li",
"Bo",
""
],
[
"Shao",
"Ling",
""
],
[
"Sebe",
"Nicu",
""
],
[
"Timofte",
"Radu",
""
],
[
"Van Gool",
"Luc",
""
]
] | TITLE: Graph Transformer GANs for Graph-Constrained House Generation
ABSTRACT: We present a novel graph Transformer generative adversarial network (GTGAN)
to learn effective graph node relations in an end-to-end fashion for the
challenging graph-constrained house generation task. The proposed
graph-Transformer-based generator includes a novel graph Transformer encoder
that combines graph convolutions and self-attentions in a Transformer to model
both local and global interactions across connected and non-connected graph
nodes. Specifically, the proposed connected node attention (CNA) and
non-connected node attention (NNA) aim to capture the global relations across
connected nodes and non-connected nodes in the input graph, respectively. The
proposed graph modeling block (GMB) aims to exploit local vertex interactions
based on a house layout topology. Moreover, we propose a new node
classification-based discriminator to preserve the high-level semantic and
discriminative node features for different house components. Finally, we
propose a novel graph-based cycle-consistency loss that aims at maintaining the
relative spatial relationships between ground truth and predicted graphs.
Experiments on two challenging graph-constrained house generation tasks (i.e.,
house layout and roof generation) with two public datasets demonstrate the
effectiveness of GTGAN in terms of objective quantitative scores and subjective
visual realism. New state-of-the-art results are established by large margins
on both tasks.
| no_new_dataset | 0.948632 |
2303.08226 | Mahdi Taheri | Mahdi Taheri, Mohammad Riazati, Mohammad Hasan Ahmadilivani, Maksim
Jenihhin, Masoud Daneshtalab, Jaan Raik, Mikael Sjodin, and Bjorn Lisper | DeepAxe: A Framework for Exploration of Approximation and Reliability
Trade-offs in DNN Accelerators | This paper is accepted at the 24th International Symposium on Quality
Electronic Design (ISQED) 2023, 8 pages, 4 figures, 4 tables | null | null | null | cs.LG cs.AI cs.AR | http://creativecommons.org/licenses/by-nc-nd/4.0/ | While the role of Deep Neural Networks (DNNs) in a wide range of
safety-critical applications is expanding, emerging DNNs experience massive
growth in terms of computation power. It raises the necessity of improving the
reliability of DNN accelerators yet reducing the computational burden on the
hardware platforms, i.e. reducing the energy consumption and execution time as
well as increasing the efficiency of DNN accelerators. Therefore, the trade-off
between hardware performance, i.e. area, power and delay, and the reliability
of the DNN accelerator implementation becomes critical and requires tools for
analysis. In this paper, we propose a framework DeepAxe for design space
exploration for FPGA-based implementation of DNNs by considering the trilateral
impact of applying functional approximation on accuracy, reliability and
hardware performance. The framework enables selective approximation of
reliability-critical DNNs, providing a set of Pareto-optimal DNN implementation
design space points for the target resource utilization requirements. The
design flow starts with a pre-trained network in Keras, uses an innovative
high-level synthesis environment DeepHLS and results in a set of Pareto-optimal
design space points as a guide for the designer. The framework is demonstrated
in a case study of custom and state-of-the-art DNNs and datasets.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 20:42:38 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Taheri",
"Mahdi",
""
],
[
"Riazati",
"Mohammad",
""
],
[
"Ahmadilivani",
"Mohammad Hasan",
""
],
[
"Jenihhin",
"Maksim",
""
],
[
"Daneshtalab",
"Masoud",
""
],
[
"Raik",
"Jaan",
""
],
[
"Sjodin",
"Mikael",
""
],
[
"Lisper",
"Bjorn",
""
]
] | TITLE: DeepAxe: A Framework for Exploration of Approximation and Reliability
Trade-offs in DNN Accelerators
ABSTRACT: While the role of Deep Neural Networks (DNNs) in a wide range of
safety-critical applications is expanding, emerging DNNs experience massive
growth in terms of computation power. It raises the necessity of improving the
reliability of DNN accelerators yet reducing the computational burden on the
hardware platforms, i.e. reducing the energy consumption and execution time as
well as increasing the efficiency of DNN accelerators. Therefore, the trade-off
between hardware performance, i.e. area, power and delay, and the reliability
of the DNN accelerator implementation becomes critical and requires tools for
analysis. In this paper, we propose a framework DeepAxe for design space
exploration for FPGA-based implementation of DNNs by considering the trilateral
impact of applying functional approximation on accuracy, reliability and
hardware performance. The framework enables selective approximation of
reliability-critical DNNs, providing a set of Pareto-optimal DNN implementation
design space points for the target resource utilization requirements. The
design flow starts with a pre-trained network in Keras, uses an innovative
high-level synthesis environment DeepHLS and results in a set of Pareto-optimal
design space points as a guide for the designer. The framework is demonstrated
in a case study of custom and state-of-the-art DNNs and datasets.
| no_new_dataset | 0.949059 |
2303.08230 | Arunesh Mittal | Arunesh Mittal, Kai Yang, Paul Sajda, John Paisley | Bayesian Beta-Bernoulli Process Sparse Coding with Deep Neural Networks | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Several approximate inference methods have been proposed for deep discrete
latent variable models. However, non-parametric methods which have previously
been successfully employed for classical sparse coding models have largely been
unexplored in the context of deep models. We propose a non-parametric iterative
algorithm for learning discrete latent representations in such deep models.
Additionally, to learn scale invariant discrete features, we propose local data
scaling variables. Lastly, to encourage sparsity in our representations, we
propose a Beta-Bernoulli process prior on the latent factors. We evaluate our
spare coding model coupled with different likelihood models. We evaluate our
method across datasets with varying characteristics and compare our results to
current amortized approximate inference methods.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 20:50:12 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Mittal",
"Arunesh",
""
],
[
"Yang",
"Kai",
""
],
[
"Sajda",
"Paul",
""
],
[
"Paisley",
"John",
""
]
] | TITLE: Bayesian Beta-Bernoulli Process Sparse Coding with Deep Neural Networks
ABSTRACT: Several approximate inference methods have been proposed for deep discrete
latent variable models. However, non-parametric methods which have previously
been successfully employed for classical sparse coding models have largely been
unexplored in the context of deep models. We propose a non-parametric iterative
algorithm for learning discrete latent representations in such deep models.
Additionally, to learn scale invariant discrete features, we propose local data
scaling variables. Lastly, to encourage sparsity in our representations, we
propose a Beta-Bernoulli process prior on the latent factors. We evaluate our
spare coding model coupled with different likelihood models. We evaluate our
method across datasets with varying characteristics and compare our results to
current amortized approximate inference methods.
| no_new_dataset | 0.947088 |
2303.08242 | Rui Xie | Rui Xie, Shuyang Bai and Ping Ma | Optimal Sampling Designs for Multi-dimensional Streaming Time Series
with Application to Power Grid Sensor Data | Accepted by The Annals of Applied Statistics | null | null | null | stat.ML cs.LG stat.AP | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The Internet of Things (IoT) system generates massive high-speed temporally
correlated streaming data and is often connected with online inference tasks
under computational or energy constraints. Online analysis of these streaming
time series data often faces a trade-off between statistical efficiency and
computational cost. One important approach to balance this trade-off is
sampling, where only a small portion of the sample is selected for the model
fitting and update. Motivated by the demands of dynamic relationship analysis
of IoT system, we study the data-dependent sample selection and online
inference problem for a multi-dimensional streaming time series, aiming to
provide low-cost real-time analysis of high-speed power grid electricity
consumption data. Inspired by D-optimality criterion in design of experiments,
we propose a class of online data reduction methods that achieve an optimal
sampling criterion and improve the computational efficiency of the online
analysis. We show that the optimal solution amounts to a strategy that is a
mixture of Bernoulli sampling and leverage score sampling. The leverage score
sampling involves auxiliary estimations that have a computational advantage
over recursive least squares updates. Theoretical properties of the auxiliary
estimations involved are also discussed. When applied to European power grid
consumption data, the proposed leverage score based sampling methods outperform
the benchmark sampling method in online estimation and prediction. The general
applicability of the sampling-assisted online estimation method is assessed via
simulation studies.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 21:26:30 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Xie",
"Rui",
""
],
[
"Bai",
"Shuyang",
""
],
[
"Ma",
"Ping",
""
]
] | TITLE: Optimal Sampling Designs for Multi-dimensional Streaming Time Series
with Application to Power Grid Sensor Data
ABSTRACT: The Internet of Things (IoT) system generates massive high-speed temporally
correlated streaming data and is often connected with online inference tasks
under computational or energy constraints. Online analysis of these streaming
time series data often faces a trade-off between statistical efficiency and
computational cost. One important approach to balance this trade-off is
sampling, where only a small portion of the sample is selected for the model
fitting and update. Motivated by the demands of dynamic relationship analysis
of IoT system, we study the data-dependent sample selection and online
inference problem for a multi-dimensional streaming time series, aiming to
provide low-cost real-time analysis of high-speed power grid electricity
consumption data. Inspired by D-optimality criterion in design of experiments,
we propose a class of online data reduction methods that achieve an optimal
sampling criterion and improve the computational efficiency of the online
analysis. We show that the optimal solution amounts to a strategy that is a
mixture of Bernoulli sampling and leverage score sampling. The leverage score
sampling involves auxiliary estimations that have a computational advantage
over recursive least squares updates. Theoretical properties of the auxiliary
estimations involved are also discussed. When applied to European power grid
consumption data, the proposed leverage score based sampling methods outperform
the benchmark sampling method in online estimation and prediction. The general
applicability of the sampling-assisted online estimation method is assessed via
simulation studies.
| no_new_dataset | 0.949295 |
2303.08264 | David Chanin | David Chanin, Anthony Hunter | Neuro-symbolic Commonsense Social Reasoning | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Social norms underlie all human social interactions, yet formalizing and
reasoning with them remains a major challenge for AI systems. We present a
novel system for taking social rules of thumb (ROTs) in natural language from
the Social Chemistry 101 dataset and converting them to first-order logic where
reasoning is performed using a neuro-symbolic theorem prover. We accomplish
this in several steps. First, ROTs are converted into Abstract Meaning
Representation (AMR), which is a graphical representation of the concepts in a
sentence, and align the AMR with RoBERTa embeddings. We then generate alternate
simplified versions of the AMR via a novel algorithm, recombining and merging
embeddings for added robustness against different wordings of text, and
incorrect AMR parses. The AMR is then converted into first-order logic, and is
queried with a neuro-symbolic theorem prover. The goal of this paper is to
develop and evaluate a neuro-symbolic method which performs explicit reasoning
about social situations in a logical form.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 22:37:33 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Chanin",
"David",
""
],
[
"Hunter",
"Anthony",
""
]
] | TITLE: Neuro-symbolic Commonsense Social Reasoning
ABSTRACT: Social norms underlie all human social interactions, yet formalizing and
reasoning with them remains a major challenge for AI systems. We present a
novel system for taking social rules of thumb (ROTs) in natural language from
the Social Chemistry 101 dataset and converting them to first-order logic where
reasoning is performed using a neuro-symbolic theorem prover. We accomplish
this in several steps. First, ROTs are converted into Abstract Meaning
Representation (AMR), which is a graphical representation of the concepts in a
sentence, and align the AMR with RoBERTa embeddings. We then generate alternate
simplified versions of the AMR via a novel algorithm, recombining and merging
embeddings for added robustness against different wordings of text, and
incorrect AMR parses. The AMR is then converted into first-order logic, and is
queried with a neuro-symbolic theorem prover. The goal of this paper is to
develop and evaluate a neuro-symbolic method which performs explicit reasoning
about social situations in a logical form.
| no_new_dataset | 0.946151 |
2303.08274 | Zhening Huang | Zhening Huang, Xiaoyang Wu, Hengshuang Zhao, Lei Zhu, Shujun Wang,
Georgios Hadjidemetriou, Ioannis Brilakis | GeoSpark: Sparking up Point Cloud Segmentation with Geometry Clue | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Current point cloud segmentation architectures suffer from limited long-range
feature modeling, as they mostly rely on aggregating information with local
neighborhoods. Furthermore, in order to learn point features at multiple
scales, most methods utilize a data-agnostic sampling approach to decrease the
number of points after each stage. Such sampling methods, however, often
discard points for small objects in the early stages, leading to inadequate
feature learning. We believe these issues are can be mitigated by introducing
explicit geometry clues as guidance. To this end, we propose GeoSpark, a
Plug-in module that incorporates Geometry clues into the network to Spark up
feature learning and downsampling. GeoSpark can be easily integrated into
various backbones. For feature aggregation, it improves feature modeling by
allowing the network to learn from both local points and neighboring geometry
partitions, resulting in an enlarged data-tailored receptive field.
Additionally, GeoSpark utilizes geometry partition information to guide the
downsampling process, where points with unique features are preserved while
redundant points are fused, resulting in better preservation of key points
throughout the network. We observed consistent improvements after adding
GeoSpark to various backbones including PointNet++, KPConv, and
PointTransformer. Notably, when integrated with Point Transformer, our GeoSpark
module achieves a 74.7% mIoU on the ScanNetv2 dataset (4.1% improvement) and
71.5% mIoU on the S3DIS Area 5 dataset (1.1% improvement), ranking top on both
benchmarks. Code and models will be made publicly available.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 23:30:46 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Huang",
"Zhening",
""
],
[
"Wu",
"Xiaoyang",
""
],
[
"Zhao",
"Hengshuang",
""
],
[
"Zhu",
"Lei",
""
],
[
"Wang",
"Shujun",
""
],
[
"Hadjidemetriou",
"Georgios",
""
],
[
"Brilakis",
"Ioannis",
""
]
] | TITLE: GeoSpark: Sparking up Point Cloud Segmentation with Geometry Clue
ABSTRACT: Current point cloud segmentation architectures suffer from limited long-range
feature modeling, as they mostly rely on aggregating information with local
neighborhoods. Furthermore, in order to learn point features at multiple
scales, most methods utilize a data-agnostic sampling approach to decrease the
number of points after each stage. Such sampling methods, however, often
discard points for small objects in the early stages, leading to inadequate
feature learning. We believe these issues are can be mitigated by introducing
explicit geometry clues as guidance. To this end, we propose GeoSpark, a
Plug-in module that incorporates Geometry clues into the network to Spark up
feature learning and downsampling. GeoSpark can be easily integrated into
various backbones. For feature aggregation, it improves feature modeling by
allowing the network to learn from both local points and neighboring geometry
partitions, resulting in an enlarged data-tailored receptive field.
Additionally, GeoSpark utilizes geometry partition information to guide the
downsampling process, where points with unique features are preserved while
redundant points are fused, resulting in better preservation of key points
throughout the network. We observed consistent improvements after adding
GeoSpark to various backbones including PointNet++, KPConv, and
PointTransformer. Notably, when integrated with Point Transformer, our GeoSpark
module achieves a 74.7% mIoU on the ScanNetv2 dataset (4.1% improvement) and
71.5% mIoU on the S3DIS Area 5 dataset (1.1% improvement), ranking top on both
benchmarks. Code and models will be made publicly available.
| no_new_dataset | 0.951369 |
2303.08295 | Chong Tang | Yao Ge and Chong Tang and Haobo Li and Zikang Zhang and Wenda Li and
Kevin Chetty and Daniele Faccio and Qammer H. Abbasi and Muhammad Imran | A large-scale multimodal dataset of human speech recognition | null | null | null | null | eess.SP cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nowadays, non-privacy small-scale motion detection has attracted an
increasing amount of research in remote sensing in speech recognition. These
new modalities are employed to enhance and restore speech information from
speakers of multiple types of data. In this paper, we propose a dataset
contains 7.5 GHz Channel Impulse Response (CIR) data from ultra-wideband (UWB)
radars, 77-GHz frequency modulated continuous wave (FMCW) data from millimetre
wave (mmWave) radar, and laser data. Meanwhile, a depth camera is adopted to
record the landmarks of the subject's lip and voice. Approximately 400 minutes
of annotated speech profiles are provided, which are collected from 20
participants speaking 5 vowels, 15 words and 16 sentences. The dataset has been
validated and has potential for the research of lip reading and multimodal
speech recognition.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 01:06:17 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Ge",
"Yao",
""
],
[
"Tang",
"Chong",
""
],
[
"Li",
"Haobo",
""
],
[
"Zhang",
"Zikang",
""
],
[
"Li",
"Wenda",
""
],
[
"Chetty",
"Kevin",
""
],
[
"Faccio",
"Daniele",
""
],
[
"Abbasi",
"Qammer H.",
""
],
[
"Imran",
"Muhammad",
""
]
] | TITLE: A large-scale multimodal dataset of human speech recognition
ABSTRACT: Nowadays, non-privacy small-scale motion detection has attracted an
increasing amount of research in remote sensing in speech recognition. These
new modalities are employed to enhance and restore speech information from
speakers of multiple types of data. In this paper, we propose a dataset
contains 7.5 GHz Channel Impulse Response (CIR) data from ultra-wideband (UWB)
radars, 77-GHz frequency modulated continuous wave (FMCW) data from millimetre
wave (mmWave) radar, and laser data. Meanwhile, a depth camera is adopted to
record the landmarks of the subject's lip and voice. Approximately 400 minutes
of annotated speech profiles are provided, which are collected from 20
participants speaking 5 vowels, 15 words and 16 sentences. The dataset has been
validated and has potential for the research of lip reading and multimodal
speech recognition.
| new_dataset | 0.959497 |
2303.08301 | Ze Mao | Ze Mao, Yang Xu, Erick Suarez | Dataset Management Platform for Machine Learning | null | null | null | null | cs.DB cs.LG | http://creativecommons.org/licenses/by/4.0/ | The quality of the data in a dataset can have a substantial impact on the
performance of a machine learning model that is trained and/or evaluated using
the dataset. Effective dataset management, including tasks such as data
cleanup, versioning, access control, dataset transformation, automation,
integrity and security, etc., can help improve the efficiency and speed of the
machine learning process. Currently, engineers spend a substantial amount of
manual effort and time to manage dataset versions or to prepare datasets for
machine learning tasks. This disclosure describes a platform to manage and use
datasets effectively. The techniques integrate dataset management and dataset
transformation mechanisms. A storage engine is described that acts as a source
of truth for all data and handles versioning, access control etc. The dataset
transformation mechanism is a key part to generate a dataset (snapshot) to
serve different purposes. The described techniques can support different
workflows, pipelines, or data orchestration needs, e.g., for training and/or
evaluation of machine learning models.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 01:26:39 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Mao",
"Ze",
""
],
[
"Xu",
"Yang",
""
],
[
"Suarez",
"Erick",
""
]
] | TITLE: Dataset Management Platform for Machine Learning
ABSTRACT: The quality of the data in a dataset can have a substantial impact on the
performance of a machine learning model that is trained and/or evaluated using
the dataset. Effective dataset management, including tasks such as data
cleanup, versioning, access control, dataset transformation, automation,
integrity and security, etc., can help improve the efficiency and speed of the
machine learning process. Currently, engineers spend a substantial amount of
manual effort and time to manage dataset versions or to prepare datasets for
machine learning tasks. This disclosure describes a platform to manage and use
datasets effectively. The techniques integrate dataset management and dataset
transformation mechanisms. A storage engine is described that acts as a source
of truth for all data and handles versioning, access control etc. The dataset
transformation mechanism is a key part to generate a dataset (snapshot) to
serve different purposes. The described techniques can support different
workflows, pipelines, or data orchestration needs, e.g., for training and/or
evaluation of machine learning models.
| no_new_dataset | 0.940898 |
2303.08303 | Wei Zhu | Wei Zhu, Runtao Zhou, Yao Yuan, Campbell Timothy, Rajat Jain, Jiebo
Luo | SegPrompt: Using Segmentation Map as a Better Prompt to Finetune Deep
Models for Kidney Stone Classification | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recently, deep learning has produced encouraging results for kidney stone
classification using endoscope images. However, the shortage of annotated
training data poses a severe problem in improving the performance and
generalization ability of the trained model. It is thus crucial to fully
exploit the limited data at hand. In this paper, we propose SegPrompt to
alleviate the data shortage problems by exploiting segmentation maps from two
aspects. First, SegPrompt integrates segmentation maps to facilitate
classification training so that the classification model is aware of the
regions of interest. The proposed method allows the image and segmentation
tokens to interact with each other to fully utilize the segmentation map
information. Second, we use the segmentation maps as prompts to tune the
pretrained deep model, resulting in much fewer trainable parameters than
vanilla finetuning. We perform extensive experiments on the collected kidney
stone dataset. The results show that SegPrompt can achieve an advantageous
balance between the model fitting ability and the generalization ability,
eventually leading to an effective model with limited training data.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 01:30:48 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zhu",
"Wei",
""
],
[
"Zhou",
"Runtao",
""
],
[
"Yuan",
"Yao",
""
],
[
"Timothy",
"Campbell",
""
],
[
"Jain",
"Rajat",
""
],
[
"Luo",
"Jiebo",
""
]
] | TITLE: SegPrompt: Using Segmentation Map as a Better Prompt to Finetune Deep
Models for Kidney Stone Classification
ABSTRACT: Recently, deep learning has produced encouraging results for kidney stone
classification using endoscope images. However, the shortage of annotated
training data poses a severe problem in improving the performance and
generalization ability of the trained model. It is thus crucial to fully
exploit the limited data at hand. In this paper, we propose SegPrompt to
alleviate the data shortage problems by exploiting segmentation maps from two
aspects. First, SegPrompt integrates segmentation maps to facilitate
classification training so that the classification model is aware of the
regions of interest. The proposed method allows the image and segmentation
tokens to interact with each other to fully utilize the segmentation map
information. Second, we use the segmentation maps as prompts to tune the
pretrained deep model, resulting in much fewer trainable parameters than
vanilla finetuning. We perform extensive experiments on the collected kidney
stone dataset. The results show that SegPrompt can achieve an advantageous
balance between the model fitting ability and the generalization ability,
eventually leading to an effective model with limited training data.
| no_new_dataset | 0.950041 |
2303.08318 | Xiao Wang | Xiao Wang, Tian Gan, Yinwei Wei, Jianlong Wu, Dai Meng, Liqiang Nie | Micro-video Tagging via Jointly Modeling Social Influence and Tag
Relation | Accepted by Proceedings of the 30th ACM International Conference on
Multimedia (2022) | null | 10.1145/3503161.3548098 | null | cs.MM cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The last decade has witnessed the proliferation of micro-videos on various
user-generated content platforms. According to our statistics, around 85.7\% of
micro-videos lack annotation. In this paper, we focus on annotating
micro-videos with tags. Existing methods mostly focus on analyzing video
content, neglecting users' social influence and tag relation. Meanwhile,
existing tag relation construction methods suffer from either deficient
performance or low tag coverage. To jointly model social influence and tag
relation, we formulate micro-video tagging as a link prediction problem in a
constructed heterogeneous network. Specifically, the tag relation (represented
by tag ontology) is constructed in a semi-supervised manner. Then, we combine
tag relation, video-tag annotation, and user-follow relation to build the
network. Afterward, a better video and tag representation are derived through
Behavior Spread modeling and visual and linguistic knowledge aggregation.
Finally, the semantic similarity between each micro-video and all candidate
tags is calculated in this video-tag network. Extensive experiments on
industrial datasets of three verticals verify the superiority of our model
compared with several state-of-the-art baselines.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 02:13:34 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Wang",
"Xiao",
""
],
[
"Gan",
"Tian",
""
],
[
"Wei",
"Yinwei",
""
],
[
"Wu",
"Jianlong",
""
],
[
"Meng",
"Dai",
""
],
[
"Nie",
"Liqiang",
""
]
] | TITLE: Micro-video Tagging via Jointly Modeling Social Influence and Tag
Relation
ABSTRACT: The last decade has witnessed the proliferation of micro-videos on various
user-generated content platforms. According to our statistics, around 85.7\% of
micro-videos lack annotation. In this paper, we focus on annotating
micro-videos with tags. Existing methods mostly focus on analyzing video
content, neglecting users' social influence and tag relation. Meanwhile,
existing tag relation construction methods suffer from either deficient
performance or low tag coverage. To jointly model social influence and tag
relation, we formulate micro-video tagging as a link prediction problem in a
constructed heterogeneous network. Specifically, the tag relation (represented
by tag ontology) is constructed in a semi-supervised manner. Then, we combine
tag relation, video-tag annotation, and user-follow relation to build the
network. Afterward, a better video and tag representation are derived through
Behavior Spread modeling and visual and linguistic knowledge aggregation.
Finally, the semantic similarity between each micro-video and all candidate
tags is calculated in this video-tag network. Extensive experiments on
industrial datasets of three verticals verify the superiority of our model
compared with several state-of-the-art baselines.
| no_new_dataset | 0.946448 |
2303.08329 | Suhee Jo | Suhee Jo, Younggun Lee, Yookyung Shin, Yeongtae Hwang, Taesu Kim | Cross-speaker Emotion Transfer by Manipulating Speech Style Latents | accepted to ICASSP 2023 | null | null | null | cs.SD cs.CL eess.AS | http://creativecommons.org/licenses/by/4.0/ | In recent years, emotional text-to-speech has shown considerable progress.
However, it requires a large amount of labeled data, which is not easily
accessible. Even if it is possible to acquire an emotional speech dataset,
there is still a limitation in controlling emotion intensity. In this work, we
propose a novel method for cross-speaker emotion transfer and manipulation
using vector arithmetic in latent style space. By leveraging only a few labeled
samples, we generate emotional speech from reading-style speech without losing
the speaker identity. Furthermore, emotion strength is readily controllable
using a scalar value, providing an intuitive way for users to manipulate
speech. Experimental results show the proposed method affords superior
performance in terms of expressiveness, naturalness, and controllability,
preserving speaker identity.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 02:34:03 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Jo",
"Suhee",
""
],
[
"Lee",
"Younggun",
""
],
[
"Shin",
"Yookyung",
""
],
[
"Hwang",
"Yeongtae",
""
],
[
"Kim",
"Taesu",
""
]
] | TITLE: Cross-speaker Emotion Transfer by Manipulating Speech Style Latents
ABSTRACT: In recent years, emotional text-to-speech has shown considerable progress.
However, it requires a large amount of labeled data, which is not easily
accessible. Even if it is possible to acquire an emotional speech dataset,
there is still a limitation in controlling emotion intensity. In this work, we
propose a novel method for cross-speaker emotion transfer and manipulation
using vector arithmetic in latent style space. By leveraging only a few labeled
samples, we generate emotional speech from reading-style speech without losing
the speaker identity. Furthermore, emotion strength is readily controllable
using a scalar value, providing an intuitive way for users to manipulate
speech. Experimental results show the proposed method affords superior
performance in terms of expressiveness, naturalness, and controllability,
preserving speaker identity.
| no_new_dataset | 0.947478 |
2303.08333 | Jiayu Zou | Jiayu Zou, Zheng Zhu, Yun Ye, Xingang Wang | DiffBEV: Conditional Diffusion Model for Bird's Eye View Perception | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | BEV perception is of great importance in the field of autonomous driving,
serving as the cornerstone of planning, controlling, and motion prediction. The
quality of the BEV feature highly affects the performance of BEV perception.
However, taking the noises in camera parameters and LiDAR scans into
consideration, we usually obtain BEV representation with harmful noises.
Diffusion models naturally have the ability to denoise noisy samples to the
ideal data, which motivates us to utilize the diffusion model to get a better
BEV representation. In this work, we propose an end-to-end framework, named
DiffBEV, to exploit the potential of diffusion model to generate a more
comprehensive BEV representation. To the best of our knowledge, we are the
first to apply diffusion model to BEV perception. In practice, we design three
types of conditions to guide the training of the diffusion model which denoises
the coarse samples and refines the semantic feature in a progressive way.
What's more, a cross-attention module is leveraged to fuse the context of BEV
feature and the semantic content of conditional diffusion model. DiffBEV
achieves a 25.9% mIoU on the nuScenes dataset, which is 6.2% higher than the
best-performing existing approach. Quantitative and qualitative results on
multiple benchmarks demonstrate the effectiveness of DiffBEV in BEV semantic
segmentation and 3D object detection tasks. The code will be available soon.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 02:42:48 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zou",
"Jiayu",
""
],
[
"Zhu",
"Zheng",
""
],
[
"Ye",
"Yun",
""
],
[
"Wang",
"Xingang",
""
]
] | TITLE: DiffBEV: Conditional Diffusion Model for Bird's Eye View Perception
ABSTRACT: BEV perception is of great importance in the field of autonomous driving,
serving as the cornerstone of planning, controlling, and motion prediction. The
quality of the BEV feature highly affects the performance of BEV perception.
However, taking the noises in camera parameters and LiDAR scans into
consideration, we usually obtain BEV representation with harmful noises.
Diffusion models naturally have the ability to denoise noisy samples to the
ideal data, which motivates us to utilize the diffusion model to get a better
BEV representation. In this work, we propose an end-to-end framework, named
DiffBEV, to exploit the potential of diffusion model to generate a more
comprehensive BEV representation. To the best of our knowledge, we are the
first to apply diffusion model to BEV perception. In practice, we design three
types of conditions to guide the training of the diffusion model which denoises
the coarse samples and refines the semantic feature in a progressive way.
What's more, a cross-attention module is leveraged to fuse the context of BEV
feature and the semantic content of conditional diffusion model. DiffBEV
achieves a 25.9% mIoU on the nuScenes dataset, which is 6.2% higher than the
best-performing existing approach. Quantitative and qualitative results on
multiple benchmarks demonstrate the effectiveness of DiffBEV in BEV semantic
segmentation and 3D object detection tasks. The code will be available soon.
| no_new_dataset | 0.948346 |
2303.08343 | Steven M. Hernandez | Steven M. Hernandez, Ding Zhao, Shaojin Ding, Antoine Bruguier, Rohit
Prabhavalkar, Tara N. Sainath, Yanzhang He, Ian McGraw | Sharing Low Rank Conformer Weights for Tiny Always-On Ambient Speech
Recognition Models | Accepted to IEEE ICASSP 2023 | null | null | null | eess.AS cs.AI cs.LG cs.SD | http://creativecommons.org/licenses/by/4.0/ | Continued improvements in machine learning techniques offer exciting new
opportunities through the use of larger models and larger training datasets.
However, there is a growing need to offer these new capabilities on-board
low-powered devices such as smartphones, wearables and other embedded
environments where only low memory is available. Towards this, we consider
methods to reduce the model size of Conformer-based speech recognition models
which typically require models with greater than 100M parameters down to just
$5$M parameters while minimizing impact on model quality. Such a model allows
us to achieve always-on ambient speech recognition on edge devices with
low-memory neural processors. We propose model weight reuse at different levels
within our model architecture: (i) repeating full conformer block layers, (ii)
sharing specific conformer modules across layers, (iii) sharing sub-components
per conformer module, and (iv) sharing decomposed sub-component weights after
low-rank decomposition. By sharing weights at different levels of our model, we
can retain the full model in-memory while increasing the number of virtual
transformations applied to the input. Through a series of ablation studies and
evaluations, we find that with weight sharing and a low-rank architecture, we
can achieve a WER of 2.84 and 2.94 for Librispeech dev-clean and test-clean
respectively with a $5$M parameter model.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 03:21:38 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Hernandez",
"Steven M.",
""
],
[
"Zhao",
"Ding",
""
],
[
"Ding",
"Shaojin",
""
],
[
"Bruguier",
"Antoine",
""
],
[
"Prabhavalkar",
"Rohit",
""
],
[
"Sainath",
"Tara N.",
""
],
[
"He",
"Yanzhang",
""
],
[
"McGraw",
"Ian",
""
]
] | TITLE: Sharing Low Rank Conformer Weights for Tiny Always-On Ambient Speech
Recognition Models
ABSTRACT: Continued improvements in machine learning techniques offer exciting new
opportunities through the use of larger models and larger training datasets.
However, there is a growing need to offer these new capabilities on-board
low-powered devices such as smartphones, wearables and other embedded
environments where only low memory is available. Towards this, we consider
methods to reduce the model size of Conformer-based speech recognition models
which typically require models with greater than 100M parameters down to just
$5$M parameters while minimizing impact on model quality. Such a model allows
us to achieve always-on ambient speech recognition on edge devices with
low-memory neural processors. We propose model weight reuse at different levels
within our model architecture: (i) repeating full conformer block layers, (ii)
sharing specific conformer modules across layers, (iii) sharing sub-components
per conformer module, and (iv) sharing decomposed sub-component weights after
low-rank decomposition. By sharing weights at different levels of our model, we
can retain the full model in-memory while increasing the number of virtual
transformations applied to the input. Through a series of ablation studies and
evaluations, we find that with weight sharing and a low-rank architecture, we
can achieve a WER of 2.84 and 2.94 for Librispeech dev-clean and test-clean
respectively with a $5$M parameter model.
| no_new_dataset | 0.939248 |
2303.08346 | Yuhan Quan | Yuhan Quan, Jingtao Ding, Chen Gao, Lingling Yi, Depeng Jin, Yong Li | Robust Preference-Guided Denoising for Graph based Social Recommendation | null | null | 10.1145/3543507.3583374 | null | cs.IR | http://creativecommons.org/publicdomain/zero/1.0/ | Graph Neural Network(GNN) based social recommendation models improve the
prediction accuracy of user preference by leveraging GNN in exploiting
preference similarity contained in social relations. However, in terms of both
effectiveness and efficiency of recommendation, a large portion of social
relations can be redundant or even noisy, e.g., it is quite normal that friends
share no preference in a certain domain. Existing models do not fully solve
this problem of relation redundancy and noise, as they directly characterize
social influence over the full social network. In this paper, we instead
propose to improve graph based social recommendation by only retaining the
informative social relations to ensure an efficient and effective influence
diffusion, i.e., graph denoising. Our designed denoising method is
preference-guided to model social relation confidence and benefits user
preference learning in return by providing a denoised but more informative
social graph for recommendation models. Moreover, to avoid interference of
noisy social relations, it designs a self-correcting curriculum learning module
and an adaptive denoising strategy, both favoring highly-confident samples.
Experimental results on three public datasets demonstrate its consistent
capability of improving two state-of-the-art social recommendation models by
robustly removing 10-40% of original relations. We release the source code at
https://github.com/tsinghua-fib-lab/Graph-Denoising-SocialRec.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 03:57:09 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Quan",
"Yuhan",
""
],
[
"Ding",
"Jingtao",
""
],
[
"Gao",
"Chen",
""
],
[
"Yi",
"Lingling",
""
],
[
"Jin",
"Depeng",
""
],
[
"Li",
"Yong",
""
]
] | TITLE: Robust Preference-Guided Denoising for Graph based Social Recommendation
ABSTRACT: Graph Neural Network(GNN) based social recommendation models improve the
prediction accuracy of user preference by leveraging GNN in exploiting
preference similarity contained in social relations. However, in terms of both
effectiveness and efficiency of recommendation, a large portion of social
relations can be redundant or even noisy, e.g., it is quite normal that friends
share no preference in a certain domain. Existing models do not fully solve
this problem of relation redundancy and noise, as they directly characterize
social influence over the full social network. In this paper, we instead
propose to improve graph based social recommendation by only retaining the
informative social relations to ensure an efficient and effective influence
diffusion, i.e., graph denoising. Our designed denoising method is
preference-guided to model social relation confidence and benefits user
preference learning in return by providing a denoised but more informative
social graph for recommendation models. Moreover, to avoid interference of
noisy social relations, it designs a self-correcting curriculum learning module
and an adaptive denoising strategy, both favoring highly-confident samples.
Experimental results on three public datasets demonstrate its consistent
capability of improving two state-of-the-art social recommendation models by
robustly removing 10-40% of original relations. We release the source code at
https://github.com/tsinghua-fib-lab/Graph-Denoising-SocialRec.
| no_new_dataset | 0.945851 |
2303.08348 | Peng Mi | Peng Mi, Jianghang Lin, Yiyi Zhou, Yunhang Shen, Gen Luo, Xiaoshuai
Sun, Liujuan Cao, Rongrong Fu, Qiang Xu, Rongrong Ji | Active Teacher for Semi-Supervised Object Detection | 8 pages, 7 figures, CVPR2022 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study teacher-student learning from the perspective of data
initialization and propose a novel algorithm called Active Teacher(Source code
are available at: \url{https://github.com/HunterJ-Lin/ActiveTeacher}) for
semi-supervised object detection (SSOD). Active Teacher extends the
teacher-student framework to an iterative version, where the label set is
partially initialized and gradually augmented by evaluating three key factors
of unlabeled examples, including difficulty, information and diversity. With
this design, Active Teacher can maximize the effect of limited label
information while improving the quality of pseudo-labels. To validate our
approach, we conduct extensive experiments on the MS-COCO benchmark and compare
Active Teacher with a set of recently proposed SSOD methods. The experimental
results not only validate the superior performance gain of Active Teacher over
the compared methods, but also show that it enables the baseline network, ie,
Faster-RCNN, to achieve 100% supervised performance with much less label
expenditure, ie 40% labeled examples on MS-COCO. More importantly, we believe
that the experimental analyses in this paper can provide useful empirical
knowledge for data annotation in practical applications.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 03:59:27 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Mi",
"Peng",
""
],
[
"Lin",
"Jianghang",
""
],
[
"Zhou",
"Yiyi",
""
],
[
"Shen",
"Yunhang",
""
],
[
"Luo",
"Gen",
""
],
[
"Sun",
"Xiaoshuai",
""
],
[
"Cao",
"Liujuan",
""
],
[
"Fu",
"Rongrong",
""
],
[
"Xu",
"Qiang",
""
],
[
"Ji",
"Rongrong",
""
]
] | TITLE: Active Teacher for Semi-Supervised Object Detection
ABSTRACT: In this paper, we study teacher-student learning from the perspective of data
initialization and propose a novel algorithm called Active Teacher(Source code
are available at: \url{https://github.com/HunterJ-Lin/ActiveTeacher}) for
semi-supervised object detection (SSOD). Active Teacher extends the
teacher-student framework to an iterative version, where the label set is
partially initialized and gradually augmented by evaluating three key factors
of unlabeled examples, including difficulty, information and diversity. With
this design, Active Teacher can maximize the effect of limited label
information while improving the quality of pseudo-labels. To validate our
approach, we conduct extensive experiments on the MS-COCO benchmark and compare
Active Teacher with a set of recently proposed SSOD methods. The experimental
results not only validate the superior performance gain of Active Teacher over
the compared methods, but also show that it enables the baseline network, ie,
Faster-RCNN, to achieve 100% supervised performance with much less label
expenditure, ie 40% labeled examples on MS-COCO. More importantly, we believe
that the experimental analyses in this paper can provide useful empirical
knowledge for data annotation in practical applications.
| no_new_dataset | 0.954816 |
2303.08362 | Hafsa Gulzar | Hafsa Gulzar, Jiyun Li, Arslan Manzoor, Sadaf Rehmat, Usman Amjad and
Hadiqa Jalil Khan | Transfer Learning Based Diagnosis and Analysis of Lung Sound Aberrations | 12 pages, 9 figures | null | null | null | cs.SD cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | With the development of computer -systems that can collect and analyze
enormous volumes of data, the medical profession is establishing several
non-invasive tools. This work attempts to develop a non-invasive technique for
identifying respiratory sounds acquired by a stethoscope and voice recording
software via machine learning techniques. This study suggests a trained and
proven CNN-based approach for categorizing respiratory sounds. A visual
representation of each audio sample is constructed, allowing resource
identification for classification using methods like those used to effectively
describe visuals. We used a technique called Mel Frequency Cepstral
Coefficients (MFCCs). Here, features are retrieved and categorized via VGG16
(transfer learning) and prediction is accomplished using 5-fold
cross-validation. Employing various data splitting techniques, Respiratory
Sound Database obtained cutting-edge results, including accuracy of 95%,
precision of 88%, recall score of 86%, and F1 score of 81%. The ICBHI dataset
is used to train and test the model.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 04:46:57 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Gulzar",
"Hafsa",
""
],
[
"Li",
"Jiyun",
""
],
[
"Manzoor",
"Arslan",
""
],
[
"Rehmat",
"Sadaf",
""
],
[
"Amjad",
"Usman",
""
],
[
"Khan",
"Hadiqa Jalil",
""
]
] | TITLE: Transfer Learning Based Diagnosis and Analysis of Lung Sound Aberrations
ABSTRACT: With the development of computer -systems that can collect and analyze
enormous volumes of data, the medical profession is establishing several
non-invasive tools. This work attempts to develop a non-invasive technique for
identifying respiratory sounds acquired by a stethoscope and voice recording
software via machine learning techniques. This study suggests a trained and
proven CNN-based approach for categorizing respiratory sounds. A visual
representation of each audio sample is constructed, allowing resource
identification for classification using methods like those used to effectively
describe visuals. We used a technique called Mel Frequency Cepstral
Coefficients (MFCCs). Here, features are retrieved and categorized via VGG16
(transfer learning) and prediction is accomplished using 5-fold
cross-validation. Employing various data splitting techniques, Respiratory
Sound Database obtained cutting-edge results, including accuracy of 95%,
precision of 88%, recall score of 86%, and F1 score of 81%. The ICBHI dataset
is used to train and test the model.
| no_new_dataset | 0.945701 |
2303.08364 | Junbong Jang | Junbong Jang, Kwonmoo Lee and Tae-Kyun Kim | Unsupervised Contour Tracking of Live Cells by Mechanical and Cycle
Consistency Losses | 12 pages, 9 figures, Accepted to CVPR 2023 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Analyzing the dynamic changes of cellular morphology is important for
understanding the various functions and characteristics of live cells,
including stem cells and metastatic cancer cells. To this end, we need to track
all points on the highly deformable cellular contour in every frame of live
cell video. Local shapes and textures on the contour are not evident, and their
motions are complex, often with expansion and contraction of local contour
features. The prior arts for optical flow or deep point set tracking are
unsuited due to the fluidity of cells, and previous deep contour tracking does
not consider point correspondence. We propose the first deep learning-based
tracking of cellular (or more generally viscoelastic materials) contours with
point correspondence by fusing dense representation between two contours with
cross attention. Since it is impractical to manually label dense tracking
points on the contour, unsupervised learning comprised of the mechanical and
cyclical consistency losses is proposed to train our contour tracker. The
mechanical loss forcing the points to move perpendicular to the contour
effectively helps out. For quantitative evaluation, we labeled sparse tracking
points along the contour of live cells from two live cell datasets taken with
phase contrast and confocal fluorescence microscopes. Our contour tracker
quantitatively outperforms compared methods and produces qualitatively more
favorable results. Our code and data are publicly available at
https://github.com/JunbongJang/contour-tracking/
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 04:48:19 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Jang",
"Junbong",
""
],
[
"Lee",
"Kwonmoo",
""
],
[
"Kim",
"Tae-Kyun",
""
]
] | TITLE: Unsupervised Contour Tracking of Live Cells by Mechanical and Cycle
Consistency Losses
ABSTRACT: Analyzing the dynamic changes of cellular morphology is important for
understanding the various functions and characteristics of live cells,
including stem cells and metastatic cancer cells. To this end, we need to track
all points on the highly deformable cellular contour in every frame of live
cell video. Local shapes and textures on the contour are not evident, and their
motions are complex, often with expansion and contraction of local contour
features. The prior arts for optical flow or deep point set tracking are
unsuited due to the fluidity of cells, and previous deep contour tracking does
not consider point correspondence. We propose the first deep learning-based
tracking of cellular (or more generally viscoelastic materials) contours with
point correspondence by fusing dense representation between two contours with
cross attention. Since it is impractical to manually label dense tracking
points on the contour, unsupervised learning comprised of the mechanical and
cyclical consistency losses is proposed to train our contour tracker. The
mechanical loss forcing the points to move perpendicular to the contour
effectively helps out. For quantitative evaluation, we labeled sparse tracking
points along the contour of live cells from two live cell datasets taken with
phase contrast and confocal fluorescence microscopes. Our contour tracker
quantitatively outperforms compared methods and produces qualitatively more
favorable results. Our code and data are publicly available at
https://github.com/JunbongJang/contour-tracking/
| no_new_dataset | 0.949482 |
2303.08370 | Liangchen Song | Liangchen Song, Zhong Li, Xuan Gong, Lele Chen, Zhang Chen, Yi Xu,
Junsong Yuan | Harnessing Low-Frequency Neural Fields for Few-Shot View Synthesis | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural Radiance Fields (NeRF) have led to breakthroughs in the novel view
synthesis problem. Positional Encoding (P.E.) is a critical factor that brings
the impressive performance of NeRF, where low-dimensional coordinates are
mapped to high-dimensional space to better recover scene details. However,
blindly increasing the frequency of P.E. leads to overfitting when the
reconstruction problem is highly underconstrained, \eg, few-shot images for
training. We harness low-frequency neural fields to regularize high-frequency
neural fields from overfitting to better address the problem of few-shot view
synthesis. We propose reconstructing with a low-frequency only field and then
finishing details with a high-frequency equipped field. Unlike most existing
solutions that regularize the output space (\ie, rendered images), our
regularization is conducted in the input space (\ie, signal frequency). We
further propose a simple-yet-effective strategy for tuning the frequency to
avoid overfitting few-shot inputs: enforcing consistency among the frequency
domain of rendered 2D images. Thanks to the input space regularizing scheme,
our method readily applies to inputs beyond spatial locations, such as the time
dimension in dynamic scenes. Comparisons with state-of-the-art on both
synthetic and natural datasets validate the effectiveness of our proposed
solution for few-shot view synthesis. Code is available at
\href{https://github.com/lsongx/halo}{https://github.com/lsongx/halo}.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 05:15:21 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Song",
"Liangchen",
""
],
[
"Li",
"Zhong",
""
],
[
"Gong",
"Xuan",
""
],
[
"Chen",
"Lele",
""
],
[
"Chen",
"Zhang",
""
],
[
"Xu",
"Yi",
""
],
[
"Yuan",
"Junsong",
""
]
] | TITLE: Harnessing Low-Frequency Neural Fields for Few-Shot View Synthesis
ABSTRACT: Neural Radiance Fields (NeRF) have led to breakthroughs in the novel view
synthesis problem. Positional Encoding (P.E.) is a critical factor that brings
the impressive performance of NeRF, where low-dimensional coordinates are
mapped to high-dimensional space to better recover scene details. However,
blindly increasing the frequency of P.E. leads to overfitting when the
reconstruction problem is highly underconstrained, \eg, few-shot images for
training. We harness low-frequency neural fields to regularize high-frequency
neural fields from overfitting to better address the problem of few-shot view
synthesis. We propose reconstructing with a low-frequency only field and then
finishing details with a high-frequency equipped field. Unlike most existing
solutions that regularize the output space (\ie, rendered images), our
regularization is conducted in the input space (\ie, signal frequency). We
further propose a simple-yet-effective strategy for tuning the frequency to
avoid overfitting few-shot inputs: enforcing consistency among the frequency
domain of rendered 2D images. Thanks to the input space regularizing scheme,
our method readily applies to inputs beyond spatial locations, such as the time
dimension in dynamic scenes. Comparisons with state-of-the-art on both
synthetic and natural datasets validate the effectiveness of our proposed
solution for few-shot view synthesis. Code is available at
\href{https://github.com/lsongx/halo}{https://github.com/lsongx/halo}.
| no_new_dataset | 0.956513 |
2303.08371 | Muhammad Jahanzeb Khan | Muhammad Jahanzeb Khan, Rui Hu, Mohammad Sadoghi, Dongfang Zhao | Comparative Evaluation of Data Decoupling Techniques for Federated
Machine Learning with Database as a Service | 14 pages, 15 figures, 3 tables | null | null | null | cs.DB cs.AI cs.DC | http://creativecommons.org/licenses/by/4.0/ | Federated Learning (FL) is a machine learning approach that allows multiple
clients to collaboratively learn a shared model without sharing raw data.
However, current FL systems provide an all-in-one solution, which can hinder
the wide adoption of FL in certain domains such as scientific applications. To
overcome this limitation, this paper proposes a decoupling approach that
enables clients to customize FL applications with specific data subsystems. To
evaluate this approach, the authors develop a framework called Data-Decoupling
Federated Learning (DDFL) and compare it with state-of-the-art FL systems that
tightly couple data management and computation. Extensive experiments on
various datasets and data management subsystems show that DDFL achieves
comparable or better performance in terms of training time, inference accuracy,
and database query time. Moreover, DDFL provides clients with more options to
tune their FL applications regarding data-related metrics. The authors also
provide a detailed qualitative analysis of DDFL when integrated with mainstream
database systems.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 05:17:00 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Khan",
"Muhammad Jahanzeb",
""
],
[
"Hu",
"Rui",
""
],
[
"Sadoghi",
"Mohammad",
""
],
[
"Zhao",
"Dongfang",
""
]
] | TITLE: Comparative Evaluation of Data Decoupling Techniques for Federated
Machine Learning with Database as a Service
ABSTRACT: Federated Learning (FL) is a machine learning approach that allows multiple
clients to collaboratively learn a shared model without sharing raw data.
However, current FL systems provide an all-in-one solution, which can hinder
the wide adoption of FL in certain domains such as scientific applications. To
overcome this limitation, this paper proposes a decoupling approach that
enables clients to customize FL applications with specific data subsystems. To
evaluate this approach, the authors develop a framework called Data-Decoupling
Federated Learning (DDFL) and compare it with state-of-the-art FL systems that
tightly couple data management and computation. Extensive experiments on
various datasets and data management subsystems show that DDFL achieves
comparable or better performance in terms of training time, inference accuracy,
and database query time. Moreover, DDFL provides clients with more options to
tune their FL applications regarding data-related metrics. The authors also
provide a detailed qualitative analysis of DDFL when integrated with mainstream
database systems.
| no_new_dataset | 0.944125 |
2303.08372 | Chenda Li | Chenda Li, Yao Qian, Zhuo Chen, Dongmei Wang, Takuya Yoshioka, Shujie
Liu, Yanmin Qian, Michael Zeng | Target Sound Extraction with Variable Cross-modality Clues | Accepted by ICASSP 2023 | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatic target sound extraction (TSE) is a machine learning approach to
mimic the human auditory perception capability of attending to a sound source
of interest from a mixture of sources. It often uses a model conditioned on a
fixed form of target sound clues, such as a sound class label, which limits the
ways in which users can interact with the model to specify the target sounds.
To leverage variable number of clues cross modalities available in the
inference phase, including a video, a sound event class, and a text caption, we
propose a unified transformer-based TSE model architecture, where a multi-clue
attention module integrates all the clues across the modalities. Since there is
no off-the-shelf benchmark to evaluate our proposed approach, we build a
dataset based on public corpora, Audioset and AudioCaps. Experimental results
for seen and unseen target-sound evaluation sets show that our proposed TSE
model can effectively deal with a varying number of clues which improves the
TSE performance and robustness against partially compromised clues.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 05:17:53 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Li",
"Chenda",
""
],
[
"Qian",
"Yao",
""
],
[
"Chen",
"Zhuo",
""
],
[
"Wang",
"Dongmei",
""
],
[
"Yoshioka",
"Takuya",
""
],
[
"Liu",
"Shujie",
""
],
[
"Qian",
"Yanmin",
""
],
[
"Zeng",
"Michael",
""
]
] | TITLE: Target Sound Extraction with Variable Cross-modality Clues
ABSTRACT: Automatic target sound extraction (TSE) is a machine learning approach to
mimic the human auditory perception capability of attending to a sound source
of interest from a mixture of sources. It often uses a model conditioned on a
fixed form of target sound clues, such as a sound class label, which limits the
ways in which users can interact with the model to specify the target sounds.
To leverage variable number of clues cross modalities available in the
inference phase, including a video, a sound event class, and a text caption, we
propose a unified transformer-based TSE model architecture, where a multi-clue
attention module integrates all the clues across the modalities. Since there is
no off-the-shelf benchmark to evaluate our proposed approach, we build a
dataset based on public corpora, Audioset and AudioCaps. Experimental results
for seen and unseen target-sound evaluation sets show that our proposed TSE
model can effectively deal with a varying number of clues which improves the
TSE performance and robustness against partially compromised clues.
| new_dataset | 0.921711 |
2303.08384 | Qiaole Dong | Qiaole Dong and Chenjie Cao and Yanwei Fu | Rethinking Optical Flow from Geometric Matching Consistent Perspective | Accepted by CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Optical flow estimation is a challenging problem remaining unsolved. Recent
deep learning based optical flow models have achieved considerable success.
However, these models often train networks from the scratch on standard optical
flow data, which restricts their ability to robustly and geometrically match
image features. In this paper, we propose a rethinking to previous optical flow
estimation. We particularly leverage Geometric Image Matching (GIM) as a
pre-training task for the optical flow estimation (MatchFlow) with better
feature representations, as GIM shares some common challenges as optical flow
estimation, and with massive labeled real-world data. Thus, matching static
scenes helps to learn more fundamental feature correlations of objects and
scenes with consistent displacements. Specifically, the proposed MatchFlow
model employs a QuadTree attention-based network pre-trained on MegaDepth to
extract coarse features for further flow regression. Extensive experiments show
that our model has great cross-dataset generalization. Our method achieves
11.5% and 10.1% error reduction from GMA on Sintel clean pass and KITTI test
set. At the time of anonymous submission, our MatchFlow(G) enjoys
state-of-the-art performance on Sintel clean and final pass compared to
published approaches with comparable computation and memory footprint. Codes
and models will be released in https://github.com/DQiaole/MatchFlow.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 06:00:38 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Dong",
"Qiaole",
""
],
[
"Cao",
"Chenjie",
""
],
[
"Fu",
"Yanwei",
""
]
] | TITLE: Rethinking Optical Flow from Geometric Matching Consistent Perspective
ABSTRACT: Optical flow estimation is a challenging problem remaining unsolved. Recent
deep learning based optical flow models have achieved considerable success.
However, these models often train networks from the scratch on standard optical
flow data, which restricts their ability to robustly and geometrically match
image features. In this paper, we propose a rethinking to previous optical flow
estimation. We particularly leverage Geometric Image Matching (GIM) as a
pre-training task for the optical flow estimation (MatchFlow) with better
feature representations, as GIM shares some common challenges as optical flow
estimation, and with massive labeled real-world data. Thus, matching static
scenes helps to learn more fundamental feature correlations of objects and
scenes with consistent displacements. Specifically, the proposed MatchFlow
model employs a QuadTree attention-based network pre-trained on MegaDepth to
extract coarse features for further flow regression. Extensive experiments show
that our model has great cross-dataset generalization. Our method achieves
11.5% and 10.1% error reduction from GMA on Sintel clean pass and KITTI test
set. At the time of anonymous submission, our MatchFlow(G) enjoys
state-of-the-art performance on Sintel clean and final pass compared to
published approaches with comparable computation and memory footprint. Codes
and models will be released in https://github.com/DQiaole/MatchFlow.
| no_new_dataset | 0.951594 |
2303.08389 | Yongil Kim | Yongil Kim, Yerin Hwang, Hyeongu Yun, Seunghyun Yoon, Trung Bui, and
Kyomin Jung | PR-MCS: Perturbation Robust Metric for MultiLingual Image Captioning | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Vulnerability to lexical perturbation is a critical weakness of automatic
evaluation metrics for image captioning. This paper proposes Perturbation
Robust Multi-Lingual CLIPScore(PR-MCS), which exhibits robustness to such
perturbations, as a novel reference-free image captioning metric applicable to
multiple languages. To achieve perturbation robustness, we fine-tune the text
encoder of CLIP with our language-agnostic method to distinguish the perturbed
text from the original text. To verify the robustness of PR-MCS, we introduce a
new fine-grained evaluation dataset consisting of detailed captions, critical
objects, and the relationships between the objects for 3, 000 images in five
languages. In our experiments, PR-MCS significantly outperforms baseline
metrics in capturing lexical noise of all various perturbation types in all
five languages, proving that PR-MCS is highly robust to lexical perturbations.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 06:37:26 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Kim",
"Yongil",
""
],
[
"Hwang",
"Yerin",
""
],
[
"Yun",
"Hyeongu",
""
],
[
"Yoon",
"Seunghyun",
""
],
[
"Bui",
"Trung",
""
],
[
"Jung",
"Kyomin",
""
]
] | TITLE: PR-MCS: Perturbation Robust Metric for MultiLingual Image Captioning
ABSTRACT: Vulnerability to lexical perturbation is a critical weakness of automatic
evaluation metrics for image captioning. This paper proposes Perturbation
Robust Multi-Lingual CLIPScore(PR-MCS), which exhibits robustness to such
perturbations, as a novel reference-free image captioning metric applicable to
multiple languages. To achieve perturbation robustness, we fine-tune the text
encoder of CLIP with our language-agnostic method to distinguish the perturbed
text from the original text. To verify the robustness of PR-MCS, we introduce a
new fine-grained evaluation dataset consisting of detailed captions, critical
objects, and the relationships between the objects for 3, 000 images in five
languages. In our experiments, PR-MCS significantly outperforms baseline
metrics in capturing lexical noise of all various perturbation types in all
five languages, proving that PR-MCS is highly robust to lexical perturbations.
| new_dataset | 0.955734 |
2303.08398 | Saeideh Yousefzadeh | Saeideh Yousefzadeh, Hamidreza Pourreza, Hamidreza Mahyar | A Triplet-loss Dilated Residual Network for High-Resolution
Representation Learning in Image Retrieval | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Content-based image retrieval is the process of retrieving a subset of images
from an extensive image gallery based on visual contents, such as color, shape
or spatial relations, and texture. In some applications, such as localization,
image retrieval is employed as the initial step. In such cases, the accuracy of
the top-retrieved images significantly affects the overall system accuracy. The
current paper introduces a simple yet efficient image retrieval system with a
fewer trainable parameters, which offers acceptable accuracy in top-retrieved
images. The proposed method benefits from a dilated residual convolutional
neural network with triplet loss. Experimental evaluations show that this model
can extract richer information (i.e., high-resolution representations) by
enlarging the receptive field, thus improving image retrieval accuracy without
increasing the depth or complexity of the model. To enhance the extracted
representations' robustness, the current research obtains candidate regions of
interest from each feature map and applies Generalized-Mean pooling to the
regions. As the choice of triplets in a triplet-based network affects the model
training, we employ a triplet online mining method. We test the performance of
the proposed method under various configurations on two of the challenging
image-retrieval datasets, namely Revisited Paris6k (RPar) and UKBench. The
experimental results show an accuracy of 94.54 and 80.23 (mean precision at
rank 10) in the RPar medium and hard modes and 3.86 (recall at rank 4) in the
UKBench dataset, respectively.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 07:01:44 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Yousefzadeh",
"Saeideh",
""
],
[
"Pourreza",
"Hamidreza",
""
],
[
"Mahyar",
"Hamidreza",
""
]
] | TITLE: A Triplet-loss Dilated Residual Network for High-Resolution
Representation Learning in Image Retrieval
ABSTRACT: Content-based image retrieval is the process of retrieving a subset of images
from an extensive image gallery based on visual contents, such as color, shape
or spatial relations, and texture. In some applications, such as localization,
image retrieval is employed as the initial step. In such cases, the accuracy of
the top-retrieved images significantly affects the overall system accuracy. The
current paper introduces a simple yet efficient image retrieval system with a
fewer trainable parameters, which offers acceptable accuracy in top-retrieved
images. The proposed method benefits from a dilated residual convolutional
neural network with triplet loss. Experimental evaluations show that this model
can extract richer information (i.e., high-resolution representations) by
enlarging the receptive field, thus improving image retrieval accuracy without
increasing the depth or complexity of the model. To enhance the extracted
representations' robustness, the current research obtains candidate regions of
interest from each feature map and applies Generalized-Mean pooling to the
regions. As the choice of triplets in a triplet-based network affects the model
training, we employ a triplet online mining method. We test the performance of
the proposed method under various configurations on two of the challenging
image-retrieval datasets, namely Revisited Paris6k (RPar) and UKBench. The
experimental results show an accuracy of 94.54 and 80.23 (mean precision at
rank 10) in the RPar medium and hard modes and 3.86 (recall at rank 4) in the
UKBench dataset, respectively.
| no_new_dataset | 0.949201 |
2303.08403 | Sungwon Han | Sungwon Han, Seungeon Lee, Fangzhao Wu, Sundong Kim, Chuhan Wu, Xiting
Wang, Xing Xie and Meeyoung Cha | DualFair: Fair Representation Learning at Both Group and Individual
Levels via Contrastive Self-supervision | Accepted and will be published at TheWebConf2023 (WWW2023) | null | null | null | cs.LG cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Algorithmic fairness has become an important machine learning problem,
especially for mission-critical Web applications. This work presents a
self-supervised model, called DualFair, that can debias sensitive attributes
like gender and race from learned representations. Unlike existing models that
target a single type of fairness, our model jointly optimizes for two fairness
criteria - group fairness and counterfactual fairness - and hence makes fairer
predictions at both the group and individual levels. Our model uses contrastive
loss to generate embeddings that are indistinguishable for each protected
group, while forcing the embeddings of counterfactual pairs to be similar. It
then uses a self-knowledge distillation method to maintain the quality of
representation for the downstream tasks. Extensive analysis over multiple
datasets confirms the model's validity and further shows the synergy of jointly
addressing two fairness criteria, suggesting the model's potential value in
fair intelligent Web applications.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 07:13:54 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Han",
"Sungwon",
""
],
[
"Lee",
"Seungeon",
""
],
[
"Wu",
"Fangzhao",
""
],
[
"Kim",
"Sundong",
""
],
[
"Wu",
"Chuhan",
""
],
[
"Wang",
"Xiting",
""
],
[
"Xie",
"Xing",
""
],
[
"Cha",
"Meeyoung",
""
]
] | TITLE: DualFair: Fair Representation Learning at Both Group and Individual
Levels via Contrastive Self-supervision
ABSTRACT: Algorithmic fairness has become an important machine learning problem,
especially for mission-critical Web applications. This work presents a
self-supervised model, called DualFair, that can debias sensitive attributes
like gender and race from learned representations. Unlike existing models that
target a single type of fairness, our model jointly optimizes for two fairness
criteria - group fairness and counterfactual fairness - and hence makes fairer
predictions at both the group and individual levels. Our model uses contrastive
loss to generate embeddings that are indistinguishable for each protected
group, while forcing the embeddings of counterfactual pairs to be similar. It
then uses a self-knowledge distillation method to maintain the quality of
representation for the downstream tasks. Extensive analysis over multiple
datasets confirms the model's validity and further shows the synergy of jointly
addressing two fairness criteria, suggesting the model's potential value in
fair intelligent Web applications.
| no_new_dataset | 0.949576 |
2303.08439 | Liang Shi | Liang Shi, Jie Zhang, Shiguang Shan | Real Face Foundation Representation Learning for Generalized Deepfake
Detection | 12 pages, 5 figures, 9 tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The emergence of deepfake technologies has become a matter of social concern
as they pose threats to individual privacy and public security. It is now of
great significance to develop reliable deepfake detectors. However, with
numerous face manipulation algorithms present, it is almost impossible to
collect sufficient representative fake faces, and it is hard for existing
detectors to generalize to all types of manipulation. Therefore, we turn to
learn the distribution of real faces, and indirectly identify fake images that
deviate from the real face distribution. In this study, we propose Real Face
Foundation Representation Learning (RFFR), which aims to learn a general
representation from large-scale real face datasets and detect potential
artifacts outside the distribution of RFFR. Specifically, we train a model on
real face datasets by masked image modeling (MIM), which results in a
discrepancy between input faces and the reconstructed ones when applying the
model on fake samples. This discrepancy reveals the low-level artifacts not
contained in RFFR, making it easier to build a deepfake detector sensitive to
all kinds of potential artifacts outside the distribution of RFFR. Extensive
experiments demonstrate that our method brings about better generalization
performance, as it significantly outperforms the state-of-the-art methods in
cross-manipulation evaluations, and has the potential to further improve by
introducing extra real faces for training RFFR.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 08:27:56 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Shi",
"Liang",
""
],
[
"Zhang",
"Jie",
""
],
[
"Shan",
"Shiguang",
""
]
] | TITLE: Real Face Foundation Representation Learning for Generalized Deepfake
Detection
ABSTRACT: The emergence of deepfake technologies has become a matter of social concern
as they pose threats to individual privacy and public security. It is now of
great significance to develop reliable deepfake detectors. However, with
numerous face manipulation algorithms present, it is almost impossible to
collect sufficient representative fake faces, and it is hard for existing
detectors to generalize to all types of manipulation. Therefore, we turn to
learn the distribution of real faces, and indirectly identify fake images that
deviate from the real face distribution. In this study, we propose Real Face
Foundation Representation Learning (RFFR), which aims to learn a general
representation from large-scale real face datasets and detect potential
artifacts outside the distribution of RFFR. Specifically, we train a model on
real face datasets by masked image modeling (MIM), which results in a
discrepancy between input faces and the reconstructed ones when applying the
model on fake samples. This discrepancy reveals the low-level artifacts not
contained in RFFR, making it easier to build a deepfake detector sensitive to
all kinds of potential artifacts outside the distribution of RFFR. Extensive
experiments demonstrate that our method brings about better generalization
performance, as it significantly outperforms the state-of-the-art methods in
cross-manipulation evaluations, and has the potential to further improve by
introducing extra real faces for training RFFR.
| no_new_dataset | 0.948728 |
2303.08446 | Honglin Li | Honglin Li, Chenglu Zhu, Yunlong Zhang, Yuxuan Sun, Zhongyi Shui,
Wenwei Kuang, Sunyi Zheng, Lin Yang | Task-specific Fine-tuning via Variational Information Bottleneck for
Weakly-supervised Pathology Whole Slide Image Classification | Accepted to CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While Multiple Instance Learning (MIL) has shown promising results in digital
Pathology Whole Slide Image (WSI) classification, such a paradigm still faces
performance and generalization problems due to challenges in high computational
costs on Gigapixel WSIs and limited sample size for model training. To deal
with the computation problem, most MIL methods utilize a frozen pretrained
model from ImageNet to obtain representations first. This process may lose
essential information owing to the large domain gap and hinder the
generalization of model due to the lack of image-level training-time
augmentations. Though Self-supervised Learning (SSL) proposes viable
representation learning schemes, the improvement of the downstream task still
needs to be further explored in the conversion from the task-agnostic features
of SSL to the task-specifics under the partial label supervised learning. To
alleviate the dilemma of computation cost and performance, we propose an
efficient WSI fine-tuning framework motivated by the Information Bottleneck
theory. The theory enables the framework to find the minimal sufficient
statistics of WSI, thus supporting us to fine-tune the backbone into a
task-specific representation only depending on WSI-level weak labels. The
WSI-MIL problem is further analyzed to theoretically deduce our fine-tuning
method. Our framework is evaluated on five pathology WSI datasets on various
WSI heads. The experimental results of our fine-tuned representations show
significant improvements in both accuracy and generalization compared with
previous works. Source code will be available at
https://github.com/invoker-LL/WSI-finetuning.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 08:41:57 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Li",
"Honglin",
""
],
[
"Zhu",
"Chenglu",
""
],
[
"Zhang",
"Yunlong",
""
],
[
"Sun",
"Yuxuan",
""
],
[
"Shui",
"Zhongyi",
""
],
[
"Kuang",
"Wenwei",
""
],
[
"Zheng",
"Sunyi",
""
],
[
"Yang",
"Lin",
""
]
] | TITLE: Task-specific Fine-tuning via Variational Information Bottleneck for
Weakly-supervised Pathology Whole Slide Image Classification
ABSTRACT: While Multiple Instance Learning (MIL) has shown promising results in digital
Pathology Whole Slide Image (WSI) classification, such a paradigm still faces
performance and generalization problems due to challenges in high computational
costs on Gigapixel WSIs and limited sample size for model training. To deal
with the computation problem, most MIL methods utilize a frozen pretrained
model from ImageNet to obtain representations first. This process may lose
essential information owing to the large domain gap and hinder the
generalization of model due to the lack of image-level training-time
augmentations. Though Self-supervised Learning (SSL) proposes viable
representation learning schemes, the improvement of the downstream task still
needs to be further explored in the conversion from the task-agnostic features
of SSL to the task-specifics under the partial label supervised learning. To
alleviate the dilemma of computation cost and performance, we propose an
efficient WSI fine-tuning framework motivated by the Information Bottleneck
theory. The theory enables the framework to find the minimal sufficient
statistics of WSI, thus supporting us to fine-tune the backbone into a
task-specific representation only depending on WSI-level weak labels. The
WSI-MIL problem is further analyzed to theoretically deduce our fine-tuning
method. Our framework is evaluated on five pathology WSI datasets on various
WSI heads. The experimental results of our fine-tuned representations show
significant improvements in both accuracy and generalization compared with
previous works. Source code will be available at
https://github.com/invoker-LL/WSI-finetuning.
| no_new_dataset | 0.9462 |
2303.08452 | Cosmin I. Bercea | Cosmin I Bercea and Benedikt Wiestler and Daniel Rueckert and Julia A
Schnabel | Reversing the Abnormal: Pseudo-Healthy Generative Networks for Anomaly
Detection | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Early and accurate disease detection is crucial for patient management and
successful treatment outcomes. However, the automatic identification of
anomalies in medical images can be challenging. Conventional methods rely on
large labeled datasets which are difficult to obtain. To overcome these
limitations, we introduce a novel unsupervised approach, called PHANES (Pseudo
Healthy generative networks for ANomaly Segmentation). Our method has the
capability of reversing anomalies, i.e., preserving healthy tissue and
replacing anomalous regions with pseudo-healthy (PH) reconstructions. Unlike
recent diffusion models, our method does not rely on a learned noise
distribution nor does it introduce random alterations to the entire image.
Instead, we use latent generative networks to create masks around possible
anomalies, which are refined using inpainting generative networks. We
demonstrate the effectiveness of PHANES in detecting stroke lesions in T1w
brain MRI datasets and show significant improvements over state-of-the-art
(SOTA) methods. We believe that our proposed framework will open new avenues
for interpretable, fast, and accurate anomaly segmentation with the potential
to support various clinical-oriented downstream tasks.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 08:54:20 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Bercea",
"Cosmin I",
""
],
[
"Wiestler",
"Benedikt",
""
],
[
"Rueckert",
"Daniel",
""
],
[
"Schnabel",
"Julia A",
""
]
] | TITLE: Reversing the Abnormal: Pseudo-Healthy Generative Networks for Anomaly
Detection
ABSTRACT: Early and accurate disease detection is crucial for patient management and
successful treatment outcomes. However, the automatic identification of
anomalies in medical images can be challenging. Conventional methods rely on
large labeled datasets which are difficult to obtain. To overcome these
limitations, we introduce a novel unsupervised approach, called PHANES (Pseudo
Healthy generative networks for ANomaly Segmentation). Our method has the
capability of reversing anomalies, i.e., preserving healthy tissue and
replacing anomalous regions with pseudo-healthy (PH) reconstructions. Unlike
recent diffusion models, our method does not rely on a learned noise
distribution nor does it introduce random alterations to the entire image.
Instead, we use latent generative networks to create masks around possible
anomalies, which are refined using inpainting generative networks. We
demonstrate the effectiveness of PHANES in detecting stroke lesions in T1w
brain MRI datasets and show significant improvements over state-of-the-art
(SOTA) methods. We believe that our proposed framework will open new avenues
for interpretable, fast, and accurate anomaly segmentation with the potential
to support various clinical-oriented downstream tasks.
| no_new_dataset | 0.952353 |
2303.08463 | Yizhe Wang | Congqi Cao, Yizhe Wang, Yue Lu, Xin Zhang and Yanning Zhang | Co-Occurrence Matters: Learning Action Relation for Temporal Action
Localization | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Temporal action localization (TAL) is a prevailing task due to its great
application potential. Existing works in this field mainly suffer from two
weaknesses: (1) They often neglect the multi-label case and only focus on
temporal modeling. (2) They ignore the semantic information in class labels and
only use the visual information. To solve these problems, we propose a novel
Co-Occurrence Relation Module (CORM) that explicitly models the co-occurrence
relationship between actions. Besides the visual information, it further
utilizes the semantic embeddings of class labels to model the co-occurrence
relationship. The CORM works in a plug-and-play manner and can be easily
incorporated with the existing sequence models. By considering both visual and
semantic co-occurrence, our method achieves high multi-label relationship
modeling capacity. Meanwhile, existing datasets in TAL always focus on
low-semantic atomic actions. Thus we construct a challenging multi-label
dataset UCF-Crime-TAL that focuses on high-semantic actions by annotating the
UCF-Crime dataset at frame level and considering the semantic overlap of
different events. Extensive experiments on two commonly used TAL datasets,
\textit{i.e.}, MultiTHUMOS and TSU, and our newly proposed UCF-Crime-TAL
demenstrate the effectiveness of the proposed CORM, which achieves
state-of-the-art performance on these datasets.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 09:07:04 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Cao",
"Congqi",
""
],
[
"Wang",
"Yizhe",
""
],
[
"Lu",
"Yue",
""
],
[
"Zhang",
"Xin",
""
],
[
"Zhang",
"Yanning",
""
]
] | TITLE: Co-Occurrence Matters: Learning Action Relation for Temporal Action
Localization
ABSTRACT: Temporal action localization (TAL) is a prevailing task due to its great
application potential. Existing works in this field mainly suffer from two
weaknesses: (1) They often neglect the multi-label case and only focus on
temporal modeling. (2) They ignore the semantic information in class labels and
only use the visual information. To solve these problems, we propose a novel
Co-Occurrence Relation Module (CORM) that explicitly models the co-occurrence
relationship between actions. Besides the visual information, it further
utilizes the semantic embeddings of class labels to model the co-occurrence
relationship. The CORM works in a plug-and-play manner and can be easily
incorporated with the existing sequence models. By considering both visual and
semantic co-occurrence, our method achieves high multi-label relationship
modeling capacity. Meanwhile, existing datasets in TAL always focus on
low-semantic atomic actions. Thus we construct a challenging multi-label
dataset UCF-Crime-TAL that focuses on high-semantic actions by annotating the
UCF-Crime dataset at frame level and considering the semantic overlap of
different events. Extensive experiments on two commonly used TAL datasets,
\textit{i.e.}, MultiTHUMOS and TSU, and our newly proposed UCF-Crime-TAL
demenstrate the effectiveness of the proposed CORM, which achieves
state-of-the-art performance on these datasets.
| no_new_dataset | 0.878575 |
2303.08490 | Chih-Chung Hsu | Chih-Chung Hsu, Chih-Yu Jian, Chia-Ming Lee, Chi-Han Tsai, and
Sheng-Chieh Dai | Strong Baseline and Bag of Tricks for COVID-19 Detection of CT Scans | technical report. Keywords: Spatial-Slice correlation, COVID-19
classification, convolutional neural networks, computed tomography | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | This paper investigates the application of deep learning models for lung
Computed Tomography (CT) image analysis. Traditional deep learning frameworks
encounter compatibility issues due to variations in slice numbers and
resolutions in CT images, which stem from the use of different machines.
Commonly, individual slices are predicted and subsequently merged to obtain the
final result; however, this approach lacks slice-wise feature learning and
consequently results in decreased performance. We propose a novel slice
selection method for each CT dataset to address this limitation, effectively
filtering out uncertain slices and enhancing the model's performance.
Furthermore, we introduce a spatial-slice feature learning (SSFL)
technique\cite{hsu2022} that employs a conventional and efficient backbone
model for slice feature training, followed by extracting one-dimensional data
from the trained model for COVID and non-COVID classification using a dedicated
classification model. Leveraging these experimental steps, we integrate
one-dimensional features with multiple slices for channel merging and employ a
2D convolutional neural network (CNN) model for classification. In addition to
the aforementioned methods, we explore various high-performance classification
models, ultimately achieving promising results.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 09:52:28 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Hsu",
"Chih-Chung",
""
],
[
"Jian",
"Chih-Yu",
""
],
[
"Lee",
"Chia-Ming",
""
],
[
"Tsai",
"Chi-Han",
""
],
[
"Dai",
"Sheng-Chieh",
""
]
] | TITLE: Strong Baseline and Bag of Tricks for COVID-19 Detection of CT Scans
ABSTRACT: This paper investigates the application of deep learning models for lung
Computed Tomography (CT) image analysis. Traditional deep learning frameworks
encounter compatibility issues due to variations in slice numbers and
resolutions in CT images, which stem from the use of different machines.
Commonly, individual slices are predicted and subsequently merged to obtain the
final result; however, this approach lacks slice-wise feature learning and
consequently results in decreased performance. We propose a novel slice
selection method for each CT dataset to address this limitation, effectively
filtering out uncertain slices and enhancing the model's performance.
Furthermore, we introduce a spatial-slice feature learning (SSFL)
technique\cite{hsu2022} that employs a conventional and efficient backbone
model for slice feature training, followed by extracting one-dimensional data
from the trained model for COVID and non-COVID classification using a dedicated
classification model. Leveraging these experimental steps, we integrate
one-dimensional features with multiple slices for channel merging and employ a
2D convolutional neural network (CNN) model for classification. In addition to
the aforementioned methods, we explore various high-performance classification
models, ultimately achieving promising results.
| no_new_dataset | 0.949949 |
2303.08514 | Jinghua Zhang | Yimin Yin, Siliang He, Renye Zhang, Hongli Chang, Xu Han, Jinghua
Zhang | Deep Learning for Iris Recognition: A Review | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Iris recognition is a secure biometric technology known for its stability and
privacy. With no two irises being identical and little change throughout a
person's lifetime, iris recognition is considered more reliable and less
susceptible to external factors than other biometric recognition methods.
Unlike traditional machine learning-based iris recognition methods, deep
learning technology does not rely on feature engineering and boasts excellent
performance. This paper collects 120 relevant papers to summarize the
development of iris recognition based on deep learning. We first introduce the
background of iris recognition and the motivation and contribution of this
survey. Then, we present the common datasets widely used in iris recognition.
After that, we summarize the key tasks involved in the process of iris
recognition based on deep learning technology, including identification,
segmentation, presentation attack detection, and localization. Finally, we
discuss the challenges and potential development of iris recognition. This
review provides a comprehensive sight of the research of iris recognition based
on deep learning.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 10:45:21 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Yin",
"Yimin",
""
],
[
"He",
"Siliang",
""
],
[
"Zhang",
"Renye",
""
],
[
"Chang",
"Hongli",
""
],
[
"Han",
"Xu",
""
],
[
"Zhang",
"Jinghua",
""
]
] | TITLE: Deep Learning for Iris Recognition: A Review
ABSTRACT: Iris recognition is a secure biometric technology known for its stability and
privacy. With no two irises being identical and little change throughout a
person's lifetime, iris recognition is considered more reliable and less
susceptible to external factors than other biometric recognition methods.
Unlike traditional machine learning-based iris recognition methods, deep
learning technology does not rely on feature engineering and boasts excellent
performance. This paper collects 120 relevant papers to summarize the
development of iris recognition based on deep learning. We first introduce the
background of iris recognition and the motivation and contribution of this
survey. Then, we present the common datasets widely used in iris recognition.
After that, we summarize the key tasks involved in the process of iris
recognition based on deep learning technology, including identification,
segmentation, presentation attack detection, and localization. Finally, we
discuss the challenges and potential development of iris recognition. This
review provides a comprehensive sight of the research of iris recognition based
on deep learning.
| no_new_dataset | 0.94868 |
2303.08538 | Yongming Li | Likun Tang, Jie Ma, Yongming Li | Health Monitoring of Movement Disorder Subject based on Diamond Stacked
Sparse Autoencoder Ensemble Model | 21 pages, 6 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The health monitoring of chronic diseases is very important for people with
movement disorders because of their limited mobility and long duration of
chronic diseases. Machine learning-based processing of data collected from the
human with movement disorders using wearable sensors is an effective method
currently available for health monitoring. However, wearable sensor systems are
difficult to obtain high-quality and large amounts of data, which cannot meet
the requirement for diagnostic accuracy. Moreover, existing machine learning
methods do not handle this problem well. Feature learning is key to machine
learning. To solve this problem, a health monitoring of movement disorder
subject based on diamond stacked sparse autoencoder ensemble model (DsaeEM) is
proposed in this paper. This algorithm has two major components. First, feature
expansion is designed using feature-embedded stacked sparse autoencoder
(FSSAE). Second, a feature reduction mechanism is designed to remove the
redundancy among the expanded features. This mechanism includes L1 regularized
feature-reduction algorithm and the improved manifold dimensionality reduction
algorithm. This paper refers to the combined feature expansion and feature
reduction mechanism as the diamond-like feature learning mechanism. The method
is experimentally verified with several state of art algorithms and on two
datasets. The results show that the proposed algorithm has higher accuracy
apparently. In conclusion, this study developed an effective and feasible
feature-learning algorithm for the recognition of chronic diseases.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 11:34:06 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Tang",
"Likun",
""
],
[
"Ma",
"Jie",
""
],
[
"Li",
"Yongming",
""
]
] | TITLE: Health Monitoring of Movement Disorder Subject based on Diamond Stacked
Sparse Autoencoder Ensemble Model
ABSTRACT: The health monitoring of chronic diseases is very important for people with
movement disorders because of their limited mobility and long duration of
chronic diseases. Machine learning-based processing of data collected from the
human with movement disorders using wearable sensors is an effective method
currently available for health monitoring. However, wearable sensor systems are
difficult to obtain high-quality and large amounts of data, which cannot meet
the requirement for diagnostic accuracy. Moreover, existing machine learning
methods do not handle this problem well. Feature learning is key to machine
learning. To solve this problem, a health monitoring of movement disorder
subject based on diamond stacked sparse autoencoder ensemble model (DsaeEM) is
proposed in this paper. This algorithm has two major components. First, feature
expansion is designed using feature-embedded stacked sparse autoencoder
(FSSAE). Second, a feature reduction mechanism is designed to remove the
redundancy among the expanded features. This mechanism includes L1 regularized
feature-reduction algorithm and the improved manifold dimensionality reduction
algorithm. This paper refers to the combined feature expansion and feature
reduction mechanism as the diamond-like feature learning mechanism. The method
is experimentally verified with several state of art algorithms and on two
datasets. The results show that the proposed algorithm has higher accuracy
apparently. In conclusion, this study developed an effective and feasible
feature-learning algorithm for the recognition of chronic diseases.
| no_new_dataset | 0.947624 |
2303.08541 | Mark Wenman Dr | J. D. Langcaster, D. S. Balint and M. R. Wenman | Adapting U-Net for linear elastic stress estimation in polycrystal Zr
microstructures | null | null | null | null | cond-mat.mtrl-sci cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A variant of the U-Net convolutional neural network architecture is proposed
to estimate linear elastic compatibility stresses in a-Zr (hcp) polycrystalline
grain structures. Training data was generated using VGrain software with a
regularity alpha of 0.73 and uniform random orientation for the grain
structures and ABAQUS to evaluate the stress welds using the finite element
method. The initial dataset contains 200 samples with 20 held from training for
validation. The network gives speedups of around 200x to 6000x using a CPU or
GPU, with signifcant memory savings, compared to finite element analysis with a
modest reduction in accuracy of up to 10%. Network performance is not
correlated with grain structure regularity or texture, showing generalisation
of the network beyond the training set to arbitrary Zr crystal structures.
Performance when trained with 200 and 400 samples was measured, finding an
improvement in accuracy of approximately 10% when the size of the dataset was
doubled.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 11:36:27 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Langcaster",
"J. D.",
""
],
[
"Balint",
"D. S.",
""
],
[
"Wenman",
"M. R.",
""
]
] | TITLE: Adapting U-Net for linear elastic stress estimation in polycrystal Zr
microstructures
ABSTRACT: A variant of the U-Net convolutional neural network architecture is proposed
to estimate linear elastic compatibility stresses in a-Zr (hcp) polycrystalline
grain structures. Training data was generated using VGrain software with a
regularity alpha of 0.73 and uniform random orientation for the grain
structures and ABAQUS to evaluate the stress welds using the finite element
method. The initial dataset contains 200 samples with 20 held from training for
validation. The network gives speedups of around 200x to 6000x using a CPU or
GPU, with signifcant memory savings, compared to finite element analysis with a
modest reduction in accuracy of up to 10%. Network performance is not
correlated with grain structure regularity or texture, showing generalisation
of the network beyond the training set to arbitrary Zr crystal structures.
Performance when trained with 200 and 400 samples was measured, finding an
improvement in accuracy of approximately 10% when the size of the dataset was
doubled.
| no_new_dataset | 0.952131 |
2303.08546 | Fuhui Zhou | Fuhui Zhou and Yihao Li and Ming Xu and Lu Yuan and Qihui Wu and Rose
Qingyang Hu and Naofal Al-Dhahir | Cognitive Semantic Communication Systems Driven by Knowledge Graph:
Principle, Implementation, and Performance Evaluation | arXiv admin note: text overlap with arXiv:2202.11958 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semantic communication is envisioned as a promising technique to break
through the Shannon limit. However, semantic inference and semantic error
correction have not been well studied. Moreover, error correction methods of
existing semantic communication frameworks are inexplicable and inflexible,
which limits the achievable performance. In this paper, to tackle this issue, a
knowledge graph is exploited to develop semantic communication systems. Two
cognitive semantic communication frameworks are proposed for the single-user
and multiple-user communication scenarios. Moreover, a simple, general, and
interpretable semantic alignment algorithm for semantic information detection
is proposed. Furthermore, an effective semantic correction algorithm is
proposed by mining the inference rule from the knowledge graph. Additionally,
the pre-trained model is fine-tuned to recover semantic information. For the
multi-user cognitive semantic communication system, a message recovery
algorithm is proposed to distinguish messages of different users by matching
the knowledge level between the source and the destination. Extensive
simulation results conducted on a public dataset demonstrate that our proposed
single-user and multi-user cognitive semantic communication systems are
superior to benchmark communication systems in terms of the data compression
rate and communication reliability. Finally, we present realistic single-user
and multi-user cognitive semantic communication systems results by building a
software-defined radio prototype system.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 12:01:43 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zhou",
"Fuhui",
""
],
[
"Li",
"Yihao",
""
],
[
"Xu",
"Ming",
""
],
[
"Yuan",
"Lu",
""
],
[
"Wu",
"Qihui",
""
],
[
"Hu",
"Rose Qingyang",
""
],
[
"Al-Dhahir",
"Naofal",
""
]
] | TITLE: Cognitive Semantic Communication Systems Driven by Knowledge Graph:
Principle, Implementation, and Performance Evaluation
ABSTRACT: Semantic communication is envisioned as a promising technique to break
through the Shannon limit. However, semantic inference and semantic error
correction have not been well studied. Moreover, error correction methods of
existing semantic communication frameworks are inexplicable and inflexible,
which limits the achievable performance. In this paper, to tackle this issue, a
knowledge graph is exploited to develop semantic communication systems. Two
cognitive semantic communication frameworks are proposed for the single-user
and multiple-user communication scenarios. Moreover, a simple, general, and
interpretable semantic alignment algorithm for semantic information detection
is proposed. Furthermore, an effective semantic correction algorithm is
proposed by mining the inference rule from the knowledge graph. Additionally,
the pre-trained model is fine-tuned to recover semantic information. For the
multi-user cognitive semantic communication system, a message recovery
algorithm is proposed to distinguish messages of different users by matching
the knowledge level between the source and the destination. Extensive
simulation results conducted on a public dataset demonstrate that our proposed
single-user and multi-user cognitive semantic communication systems are
superior to benchmark communication systems in terms of the data compression
rate and communication reliability. Finally, we present realistic single-user
and multi-user cognitive semantic communication systems results by building a
software-defined radio prototype system.
| no_new_dataset | 0.946399 |
2303.08550 | Zhe Jin | Chaoyang Jiang, Xiaoni Zheng, Zhe Jin, and Chengpu Yu | UMS-VINS: United Monocular-Stereo Features for Visual-Inertial Tightly
Coupled Odometry | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces the united monocular-stereo features into a
visual-inertial tightly coupled odometry (UMS-VINS) for robust pose estimation.
UMS-VINS requires two cameras and a low-cost inertial measurement unit (IMU).
The UMS-VINS is an evolution of VINS-FUSION, which modifies the VINS-FUSION
from the following three perspectives. 1) UMS-VINS extracts and tracks features
from the sub-pixel plane to achieve better positions of the features. 2)
UMS-VINS introduces additional 2-dimensional features from the left and/or
right cameras. 3) If the visual initialization fails, the IMU propagation is
directly used for pose estimation, and if the visual-IMU alignment fails,
UMS-VINS estimates the pose via the visual odometry. The performances on both
public datasets and new real-world experiments indicate that the proposed
UMS-VINS outperforms the VINS-FUSION from the perspective of localization
accuracy, localization robustness, and environmental adaptability.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 12:10:22 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Jiang",
"Chaoyang",
""
],
[
"Zheng",
"Xiaoni",
""
],
[
"Jin",
"Zhe",
""
],
[
"Yu",
"Chengpu",
""
]
] | TITLE: UMS-VINS: United Monocular-Stereo Features for Visual-Inertial Tightly
Coupled Odometry
ABSTRACT: This paper introduces the united monocular-stereo features into a
visual-inertial tightly coupled odometry (UMS-VINS) for robust pose estimation.
UMS-VINS requires two cameras and a low-cost inertial measurement unit (IMU).
The UMS-VINS is an evolution of VINS-FUSION, which modifies the VINS-FUSION
from the following three perspectives. 1) UMS-VINS extracts and tracks features
from the sub-pixel plane to achieve better positions of the features. 2)
UMS-VINS introduces additional 2-dimensional features from the left and/or
right cameras. 3) If the visual initialization fails, the IMU propagation is
directly used for pose estimation, and if the visual-IMU alignment fails,
UMS-VINS estimates the pose via the visual odometry. The performances on both
public datasets and new real-world experiments indicate that the proposed
UMS-VINS outperforms the VINS-FUSION from the perspective of localization
accuracy, localization robustness, and environmental adaptability.
| no_new_dataset | 0.94887 |
2303.08561 | Yulin Pan | Yulin Pan, Xiangteng He, Biao Gong, Yuxin Peng, Yiliang Lv | Enhancing Unsupervised Audio Representation Learning via Adversarial
Sample Generation | 8 pages, 4 figures | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing audio analysis methods generally first transform the audio stream to
spectrogram, and then feed it into CNN for further analysis. A standard CNN
recognizes specific visual patterns over feature map, then pools for high-level
representation, which overlooks the positional information of recognized
patterns. However, unlike natural image, the semantic of an audio spectrogram
is sensitive to positional change, as its vertical and horizontal axes indicate
the frequency and temporal information of the audio, instead of naive
rectangular coordinates. Thus, the insensitivity of CNN to positional change
plays a negative role on audio spectrogram encoding. To address this issue,
this paper proposes a new self-supervised learning mechanism, which enhances
the audio representation by first generating adversarial samples
(\textit{i.e.}, negative samples), then driving CNN to distinguish the
embeddings of negative pairs in the latent space. Extensive experiments show
that the proposed approach achieves best or competitive results on 9 downstream
datasets compared with previous methods, which verifies its effectiveness on
audio representation learning.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 12:25:29 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Pan",
"Yulin",
""
],
[
"He",
"Xiangteng",
""
],
[
"Gong",
"Biao",
""
],
[
"Peng",
"Yuxin",
""
],
[
"Lv",
"Yiliang",
""
]
] | TITLE: Enhancing Unsupervised Audio Representation Learning via Adversarial
Sample Generation
ABSTRACT: Existing audio analysis methods generally first transform the audio stream to
spectrogram, and then feed it into CNN for further analysis. A standard CNN
recognizes specific visual patterns over feature map, then pools for high-level
representation, which overlooks the positional information of recognized
patterns. However, unlike natural image, the semantic of an audio spectrogram
is sensitive to positional change, as its vertical and horizontal axes indicate
the frequency and temporal information of the audio, instead of naive
rectangular coordinates. Thus, the insensitivity of CNN to positional change
plays a negative role on audio spectrogram encoding. To address this issue,
this paper proposes a new self-supervised learning mechanism, which enhances
the audio representation by first generating adversarial samples
(\textit{i.e.}, negative samples), then driving CNN to distinguish the
embeddings of negative pairs in the latent space. Extensive experiments show
that the proposed approach achieves best or competitive results on 9 downstream
datasets compared with previous methods, which verifies its effectiveness on
audio representation learning.
| no_new_dataset | 0.950365 |
2303.08562 | Weijian Huang | Weijian Huang, Hao Yang, Cheng Li, Mingtong Dai, Rui Yang, Shanshan
Wang | MGA: Medical generalist agent through text-guided knowledge
transformation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-modal representation methods have achieved advanced performance in
medical applications by extracting more robust features from multi-domain data.
However, existing methods usually need to train additional branches for
downstream tasks, which may increase the model complexities in clinical
applications as well as introduce additional human inductive bias. Besides,
very few studies exploit the rich clinical knowledge embedded in clinical daily
reports. To this end, we propose a novel medical generalist agent, MGA, that
can address three kinds of common clinical tasks via clinical reports knowledge
transformation. Unlike the existing methods, MGA can easily adapt to different
tasks without specific downstream branches when their corresponding annotations
are missing. More importantly, we are the first attempt to use medical
professional language guidance as a transmission medium to guide the agent's
behavior. The proposed method is implemented on four well-known X-ray
open-source datasets, MIMIC-CXR, CheXpert, MIMIC-CXR-JPG, and MIMIC-CXR-MS.
Promising results are obtained, which validate the effectiveness of our
proposed MGA. Code is available at: https://github.com/SZUHvern/MGA
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 12:28:31 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Huang",
"Weijian",
""
],
[
"Yang",
"Hao",
""
],
[
"Li",
"Cheng",
""
],
[
"Dai",
"Mingtong",
""
],
[
"Yang",
"Rui",
""
],
[
"Wang",
"Shanshan",
""
]
] | TITLE: MGA: Medical generalist agent through text-guided knowledge
transformation
ABSTRACT: Multi-modal representation methods have achieved advanced performance in
medical applications by extracting more robust features from multi-domain data.
However, existing methods usually need to train additional branches for
downstream tasks, which may increase the model complexities in clinical
applications as well as introduce additional human inductive bias. Besides,
very few studies exploit the rich clinical knowledge embedded in clinical daily
reports. To this end, we propose a novel medical generalist agent, MGA, that
can address three kinds of common clinical tasks via clinical reports knowledge
transformation. Unlike the existing methods, MGA can easily adapt to different
tasks without specific downstream branches when their corresponding annotations
are missing. More importantly, we are the first attempt to use medical
professional language guidance as a transmission medium to guide the agent's
behavior. The proposed method is implemented on four well-known X-ray
open-source datasets, MIMIC-CXR, CheXpert, MIMIC-CXR-JPG, and MIMIC-CXR-MS.
Promising results are obtained, which validate the effectiveness of our
proposed MGA. Code is available at: https://github.com/SZUHvern/MGA
| no_new_dataset | 0.9462 |
2303.08572 | Mario Figueiredo | M\'ario A. T. Figueiredo and Catarina A. Oliveira | Distinguishing Cause from Effect on Categorical Data: The Uniform
Channel Model | 20 pages, 2 appendices | null | null | null | cs.LG cs.CL cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distinguishing cause from effect using observations of a pair of random
variables is a core problem in causal discovery. Most approaches proposed for
this task, namely additive noise models (ANM), are only adequate for
quantitative data. We propose a criterion to address the cause-effect problem
with categorical variables (living in sets with no meaningful order), inspired
by seeing a conditional probability mass function (pmf) as a discrete
memoryless channel. We select as the most likely causal direction the one in
which the conditional pmf is closer to a uniform channel (UC). The rationale is
that, in a UC, as in an ANM, the conditional entropy (of the effect given the
cause) is independent of the cause distribution, in agreement with the
principle of independence of cause and mechanism. Our approach, which we call
the uniform channel model (UCM), thus extends the ANM rationale to categorical
variables. To assess how close a conditional pmf (estimated from data) is to a
UC, we use statistical testing, supported by a closed-form estimate of a UC
channel. On the theoretical front, we prove identifiability of the UCM and show
its equivalence with a structural causal model with a low-cardinality exogenous
variable. Finally, the proposed method compares favorably with recent
state-of-the-art alternatives in experiments on synthetic, benchmark, and real
data.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 13:54:11 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Figueiredo",
"Mário A. T.",
""
],
[
"Oliveira",
"Catarina A.",
""
]
] | TITLE: Distinguishing Cause from Effect on Categorical Data: The Uniform
Channel Model
ABSTRACT: Distinguishing cause from effect using observations of a pair of random
variables is a core problem in causal discovery. Most approaches proposed for
this task, namely additive noise models (ANM), are only adequate for
quantitative data. We propose a criterion to address the cause-effect problem
with categorical variables (living in sets with no meaningful order), inspired
by seeing a conditional probability mass function (pmf) as a discrete
memoryless channel. We select as the most likely causal direction the one in
which the conditional pmf is closer to a uniform channel (UC). The rationale is
that, in a UC, as in an ANM, the conditional entropy (of the effect given the
cause) is independent of the cause distribution, in agreement with the
principle of independence of cause and mechanism. Our approach, which we call
the uniform channel model (UCM), thus extends the ANM rationale to categorical
variables. To assess how close a conditional pmf (estimated from data) is to a
UC, we use statistical testing, supported by a closed-form estimate of a UC
channel. On the theoretical front, we prove identifiability of the UCM and show
its equivalence with a structural causal model with a low-cardinality exogenous
variable. Finally, the proposed method compares favorably with recent
state-of-the-art alternatives in experiments on synthetic, benchmark, and real
data.
| no_new_dataset | 0.952882 |
2303.08577 | Giorgia Adorni | Giorgia Adorni, Felix Boelter, Stefano Carlo Lambertenghi | Investigating GANsformer: A Replication Study of a State-of-the-Art
Image Generation Model | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | The field of image generation through generative modelling is abundantly
discussed nowadays. It can be used for various applications, such as up-scaling
existing images, creating non-existing objects, such as interior design scenes,
products or even human faces, and achieving transfer-learning processes. In
this context, Generative Adversarial Networks (GANs) are a class of widely
studied machine learning frameworks first appearing in the paper "Generative
adversarial nets" by Goodfellow et al. that achieve the goal above. In our
work, we reproduce and evaluate a novel variation of the original GAN network,
the GANformer, proposed in "Generative Adversarial Transformers" by Hudson and
Zitnick. This project aimed to recreate the methods presented in this paper to
reproduce the original results and comment on the authors' claims. Due to
resources and time limitations, we had to constrain the network's training
times, dataset types, and sizes. Our research successfully recreated both
variations of the proposed GANformer model and found differences between the
authors' and our results. Moreover, discrepancies between the publication
methodology and the one implemented, made available in the code, allowed us to
study two undisclosed variations of the presented procedures.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 12:51:16 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Adorni",
"Giorgia",
""
],
[
"Boelter",
"Felix",
""
],
[
"Lambertenghi",
"Stefano Carlo",
""
]
] | TITLE: Investigating GANsformer: A Replication Study of a State-of-the-Art
Image Generation Model
ABSTRACT: The field of image generation through generative modelling is abundantly
discussed nowadays. It can be used for various applications, such as up-scaling
existing images, creating non-existing objects, such as interior design scenes,
products or even human faces, and achieving transfer-learning processes. In
this context, Generative Adversarial Networks (GANs) are a class of widely
studied machine learning frameworks first appearing in the paper "Generative
adversarial nets" by Goodfellow et al. that achieve the goal above. In our
work, we reproduce and evaluate a novel variation of the original GAN network,
the GANformer, proposed in "Generative Adversarial Transformers" by Hudson and
Zitnick. This project aimed to recreate the methods presented in this paper to
reproduce the original results and comment on the authors' claims. Due to
resources and time limitations, we had to constrain the network's training
times, dataset types, and sizes. Our research successfully recreated both
variations of the proposed GANformer model and found differences between the
authors' and our results. Moreover, discrepancies between the publication
methodology and the one implemented, made available in the code, allowed us to
study two undisclosed variations of the presented procedures.
| no_new_dataset | 0.9463 |
2303.08582 | Alaeddine Zahir | Alaeddine Zahir, Khalide Jbilou, Ahmed Ratnani | High-dimensional multi-view clustering methods | 4 figures | null | null | null | cs.LG cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | Multi-view clustering has been widely used in recent years in comparison to
single-view clustering, for clear reasons, as it offers more insights into the
data, which has brought with it some challenges, such as how to combine these
views or features. Most of recent work in this field focuses mainly on tensor
representation instead of treating the data as simple matrices. This permits to
deal with the high-order correlation between the data which the based matrix
approach struggles to capture. Accordingly, we will examine and compare these
approaches, particularly in two categories, namely graph-based clustering and
subspace-based clustering. We will conduct and report experiments of the main
clustering methods over a benchmark datasets.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 11:04:37 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zahir",
"Alaeddine",
""
],
[
"Jbilou",
"Khalide",
""
],
[
"Ratnani",
"Ahmed",
""
]
] | TITLE: High-dimensional multi-view clustering methods
ABSTRACT: Multi-view clustering has been widely used in recent years in comparison to
single-view clustering, for clear reasons, as it offers more insights into the
data, which has brought with it some challenges, such as how to combine these
views or features. Most of recent work in this field focuses mainly on tensor
representation instead of treating the data as simple matrices. This permits to
deal with the high-order correlation between the data which the based matrix
approach struggles to capture. Accordingly, we will examine and compare these
approaches, particularly in two categories, namely graph-based clustering and
subspace-based clustering. We will conduct and report experiments of the main
clustering methods over a benchmark datasets.
| no_new_dataset | 0.948106 |
2303.08599 | Tong Ye | Tong Ye, Zhitao Li, Jianzong Wang, Ning Cheng, Jing Xiao | Efficient Uncertainty Estimation with Gaussian Process for Reliable
Dialog Response Retrieval | Accepted by ICASSP 2023 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep neural networks have achieved remarkable performance in retrieval-based
dialogue systems, but they are shown to be ill calibrated. Though basic
calibration methods like Monte Carlo Dropout and Ensemble can calibrate well,
these methods are time-consuming in the training or inference stages. To tackle
these challenges, we propose an efficient uncertainty calibration framework
GPF-BERT for BERT-based conversational search, which employs a Gaussian Process
layer and the focal loss on top of the BERT architecture to achieve a
high-quality neural ranker. Extensive experiments are conducted to verify the
effectiveness of our method. In comparison with basic calibration methods,
GPF-BERT achieves the lowest empirical calibration error (ECE) in three
in-domain datasets and the distributional shift tasks, while yielding the
highest $R_{10}@1$ and MAP performance on most cases. In terms of time
consumption, our GPF-BERT has an 8$\times$ speedup.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 13:12:16 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Ye",
"Tong",
""
],
[
"Li",
"Zhitao",
""
],
[
"Wang",
"Jianzong",
""
],
[
"Cheng",
"Ning",
""
],
[
"Xiao",
"Jing",
""
]
] | TITLE: Efficient Uncertainty Estimation with Gaussian Process for Reliable
Dialog Response Retrieval
ABSTRACT: Deep neural networks have achieved remarkable performance in retrieval-based
dialogue systems, but they are shown to be ill calibrated. Though basic
calibration methods like Monte Carlo Dropout and Ensemble can calibrate well,
these methods are time-consuming in the training or inference stages. To tackle
these challenges, we propose an efficient uncertainty calibration framework
GPF-BERT for BERT-based conversational search, which employs a Gaussian Process
layer and the focal loss on top of the BERT architecture to achieve a
high-quality neural ranker. Extensive experiments are conducted to verify the
effectiveness of our method. In comparison with basic calibration methods,
GPF-BERT achieves the lowest empirical calibration error (ECE) in three
in-domain datasets and the distributional shift tasks, while yielding the
highest $R_{10}@1$ and MAP performance on most cases. In terms of time
consumption, our GPF-BERT has an 8$\times$ speedup.
| no_new_dataset | 0.946399 |
2303.08600 | Jiale Li | Jiale Li, Hang Dai, Hao Han, Yong Ding | MSeg3D: Multi-modal 3D Semantic Segmentation for Autonomous Driving | Accepted to CVPR 2023 (preprint) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | LiDAR and camera are two modalities available for 3D semantic segmentation in
autonomous driving. The popular LiDAR-only methods severely suffer from
inferior segmentation on small and distant objects due to insufficient laser
points, while the robust multi-modal solution is under-explored, where we
investigate three crucial inherent difficulties: modality heterogeneity,
limited sensor field of view intersection, and multi-modal data augmentation.
We propose a multi-modal 3D semantic segmentation model (MSeg3D) with joint
intra-modal feature extraction and inter-modal feature fusion to mitigate the
modality heterogeneity. The multi-modal fusion in MSeg3D consists of
geometry-based feature fusion GF-Phase, cross-modal feature completion, and
semantic-based feature fusion SF-Phase on all visible points. The multi-modal
data augmentation is reinvigorated by applying asymmetric transformations on
LiDAR point cloud and multi-camera images individually, which benefits the
model training with diversified augmentation transformations. MSeg3D achieves
state-of-the-art results on nuScenes, Waymo, and SemanticKITTI datasets. Under
the malfunctioning multi-camera input and the multi-frame point clouds input,
MSeg3D still shows robustness and improves the LiDAR-only baseline. Our code is
publicly available at \url{https://github.com/jialeli1/lidarseg3d}.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 13:13:03 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Li",
"Jiale",
""
],
[
"Dai",
"Hang",
""
],
[
"Han",
"Hao",
""
],
[
"Ding",
"Yong",
""
]
] | TITLE: MSeg3D: Multi-modal 3D Semantic Segmentation for Autonomous Driving
ABSTRACT: LiDAR and camera are two modalities available for 3D semantic segmentation in
autonomous driving. The popular LiDAR-only methods severely suffer from
inferior segmentation on small and distant objects due to insufficient laser
points, while the robust multi-modal solution is under-explored, where we
investigate three crucial inherent difficulties: modality heterogeneity,
limited sensor field of view intersection, and multi-modal data augmentation.
We propose a multi-modal 3D semantic segmentation model (MSeg3D) with joint
intra-modal feature extraction and inter-modal feature fusion to mitigate the
modality heterogeneity. The multi-modal fusion in MSeg3D consists of
geometry-based feature fusion GF-Phase, cross-modal feature completion, and
semantic-based feature fusion SF-Phase on all visible points. The multi-modal
data augmentation is reinvigorated by applying asymmetric transformations on
LiDAR point cloud and multi-camera images individually, which benefits the
model training with diversified augmentation transformations. MSeg3D achieves
state-of-the-art results on nuScenes, Waymo, and SemanticKITTI datasets. Under
the malfunctioning multi-camera input and the multi-frame point clouds input,
MSeg3D still shows robustness and improves the LiDAR-only baseline. Our code is
publicly available at \url{https://github.com/jialeli1/lidarseg3d}.
| no_new_dataset | 0.943815 |
2303.08606 | Tong Ye | Tong Ye, Shijing Si, Jianzong Wang, Ning Cheng, Zhitao Li, Jing Xiao | On the Calibration and Uncertainty with P\'{o}lya-Gamma Augmentation for
Dialog Retrieval Models | Accepted by AAAI 2023 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep neural retrieval models have amply demonstrated their power but
estimating the reliability of their predictions remains challenging. Most
dialog response retrieval models output a single score for a response on how
relevant it is to a given question. However, the bad calibration of deep neural
network results in various uncertainty for the single score such that the
unreliable predictions always misinform user decisions. To investigate these
issues, we present an efficient calibration and uncertainty estimation
framework PG-DRR for dialog response retrieval models which adds a Gaussian
Process layer to a deterministic deep neural network and recovers conjugacy for
tractable posterior inference by P\'{o}lya-Gamma augmentation. Finally, PG-DRR
achieves the lowest empirical calibration error (ECE) in the in-domain datasets
and the distributional shift task while keeping $R_{10}@1$ and MAP performance.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 13:26:25 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Ye",
"Tong",
""
],
[
"Si",
"Shijing",
""
],
[
"Wang",
"Jianzong",
""
],
[
"Cheng",
"Ning",
""
],
[
"Li",
"Zhitao",
""
],
[
"Xiao",
"Jing",
""
]
] | TITLE: On the Calibration and Uncertainty with P\'{o}lya-Gamma Augmentation for
Dialog Retrieval Models
ABSTRACT: Deep neural retrieval models have amply demonstrated their power but
estimating the reliability of their predictions remains challenging. Most
dialog response retrieval models output a single score for a response on how
relevant it is to a given question. However, the bad calibration of deep neural
network results in various uncertainty for the single score such that the
unreliable predictions always misinform user decisions. To investigate these
issues, we present an efficient calibration and uncertainty estimation
framework PG-DRR for dialog response retrieval models which adds a Gaussian
Process layer to a deterministic deep neural network and recovers conjugacy for
tractable posterior inference by P\'{o}lya-Gamma augmentation. Finally, PG-DRR
achieves the lowest empirical calibration error (ECE) in the in-domain datasets
and the distributional shift task while keeping $R_{10}@1$ and MAP performance.
| no_new_dataset | 0.942929 |
2303.08625 | Lev Utkin | Andrei V. Konstantinov and Lev V. Utkin | Interpretable Ensembles of Hyper-Rectangles as Base Models | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new extremely simple ensemble-based model with the uniformly generated
axis-parallel hyper-rectangles as base models (HRBM) is proposed. Two types of
HRBMs are studied: closed rectangles and corners. The main idea behind HRBM is
to consider and count training examples inside and outside each rectangle. It
is proposed to incorporate HRBMs into the gradient boosting machine (GBM).
Despite simplicity of HRBMs, it turns out that these simple base models allow
us to construct effective ensemble-based models and avoid overfitting. A simple
method for calculating optimal regularization parameters of the ensemble-based
model, which can be modified in the explicit way at each iteration of GBM, is
considered. Moreover, a new regularization called the "step height penalty" is
studied in addition to the standard L1 and L2 regularizations. An extremely
simple approach to the proposed ensemble-based model prediction interpretation
by using the well-known method SHAP is proposed. It is shown that GBM with HRBM
can be regarded as a model extending a set of interpretable models for
explaining black-box models. Numerical experiments with real datasets
illustrate the proposed GBM with HRBMs for regression and classification
problems. Experiments also illustrate computational efficiency of the proposed
SHAP modifications. The code of proposed algorithms implementing GBM with HRBM
is publicly available.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 13:50:36 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Konstantinov",
"Andrei V.",
""
],
[
"Utkin",
"Lev V.",
""
]
] | TITLE: Interpretable Ensembles of Hyper-Rectangles as Base Models
ABSTRACT: A new extremely simple ensemble-based model with the uniformly generated
axis-parallel hyper-rectangles as base models (HRBM) is proposed. Two types of
HRBMs are studied: closed rectangles and corners. The main idea behind HRBM is
to consider and count training examples inside and outside each rectangle. It
is proposed to incorporate HRBMs into the gradient boosting machine (GBM).
Despite simplicity of HRBMs, it turns out that these simple base models allow
us to construct effective ensemble-based models and avoid overfitting. A simple
method for calculating optimal regularization parameters of the ensemble-based
model, which can be modified in the explicit way at each iteration of GBM, is
considered. Moreover, a new regularization called the "step height penalty" is
studied in addition to the standard L1 and L2 regularizations. An extremely
simple approach to the proposed ensemble-based model prediction interpretation
by using the well-known method SHAP is proposed. It is shown that GBM with HRBM
can be regarded as a model extending a set of interpretable models for
explaining black-box models. Numerical experiments with real datasets
illustrate the proposed GBM with HRBMs for regression and classification
problems. Experiments also illustrate computational efficiency of the proposed
SHAP modifications. The code of proposed algorithms implementing GBM with HRBM
is publicly available.
| no_new_dataset | 0.945248 |
2303.08632 | Ario Sadafi | Ario Sadafi, Oleksandra Adonkina, Ashkan Khakzar, Peter Lienemann,
Rudolf Matthias Hehr, Daniel Rueckert, Nassir Navab, Carsten Marr | Pixel-Level Explanation of Multiple Instance Learning Models in
Biomedical Single Cell Images | Accepted for publication at the international conference on
Information Processing in Medical Imaging (IPMI 2023) | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Explainability is a key requirement for computer-aided diagnosis systems in
clinical decision-making. Multiple instance learning with attention pooling
provides instance-level explainability, however for many clinical applications
a deeper, pixel-level explanation is desirable, but missing so far. In this
work, we investigate the use of four attribution methods to explain a multiple
instance learning models: GradCAM, Layer-Wise Relevance Propagation (LRP),
Information Bottleneck Attribution (IBA), and InputIBA. With this collection of
methods, we can derive pixel-level explanations on for the task of diagnosing
blood cancer from patients' blood smears. We study two datasets of acute
myeloid leukemia with over 100 000 single cell images and observe how each
attribution method performs on the multiple instance learning architecture
focusing on different properties of the white blood single cells. Additionally,
we compare attribution maps with the annotations of a medical expert to see how
the model's decision-making differs from the human standard. Our study
addresses the challenge of implementing pixel-level explainability in multiple
instance learning models and provides insights for clinicians to better
understand and trust decisions from computer-aided diagnosis systems.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 14:00:11 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Sadafi",
"Ario",
""
],
[
"Adonkina",
"Oleksandra",
""
],
[
"Khakzar",
"Ashkan",
""
],
[
"Lienemann",
"Peter",
""
],
[
"Hehr",
"Rudolf Matthias",
""
],
[
"Rueckert",
"Daniel",
""
],
[
"Navab",
"Nassir",
""
],
[
"Marr",
"Carsten",
""
]
] | TITLE: Pixel-Level Explanation of Multiple Instance Learning Models in
Biomedical Single Cell Images
ABSTRACT: Explainability is a key requirement for computer-aided diagnosis systems in
clinical decision-making. Multiple instance learning with attention pooling
provides instance-level explainability, however for many clinical applications
a deeper, pixel-level explanation is desirable, but missing so far. In this
work, we investigate the use of four attribution methods to explain a multiple
instance learning models: GradCAM, Layer-Wise Relevance Propagation (LRP),
Information Bottleneck Attribution (IBA), and InputIBA. With this collection of
methods, we can derive pixel-level explanations on for the task of diagnosing
blood cancer from patients' blood smears. We study two datasets of acute
myeloid leukemia with over 100 000 single cell images and observe how each
attribution method performs on the multiple instance learning architecture
focusing on different properties of the white blood single cells. Additionally,
we compare attribution maps with the annotations of a medical expert to see how
the model's decision-making differs from the human standard. Our study
addresses the challenge of implementing pixel-level explainability in multiple
instance learning models and provides insights for clinicians to better
understand and trust decisions from computer-aided diagnosis systems.
| no_new_dataset | 0.945951 |
2303.08636 | Yang Yuguang | Yuguang Yang, Yu Pan, Jingjing Yin, Jiangyu Han, Lei Ma, Heng Lu | HYBRIDFORMER: improving SqueezeFormer with hybrid attention and NSR
mechanism | Accepted by ICASSP2023 | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | SqueezeFormer has recently shown impressive performance in automatic speech
recognition (ASR). However, its inference speed suffers from the quadratic
complexity of softmax-attention (SA). In addition, limited by the large
convolution kernel size, the local modeling ability of SqueezeFormer is
insufficient. In this paper, we propose a novel method HybridFormer to improve
SqueezeFormer in a fast and efficient way. Specifically, we first incorporate
linear attention (LA) and propose a hybrid LASA paradigm to increase the
model's inference speed. Second, a hybrid neural architecture search (NAS)
guided structural re-parameterization (SRep) mechanism, termed NSR, is proposed
to enhance the ability of the model to extract local interactions. Extensive
experiments conducted on the LibriSpeech dataset demonstrate that our proposed
HybridFormer can achieve a 9.1% relative word error rate (WER) reduction over
SqueezeFormer on the test-other dataset. Furthermore, when input speech is 30s,
the HybridFormer can improve the model's inference speed up to 18%. Our source
code is available online.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 14:03:16 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Yang",
"Yuguang",
""
],
[
"Pan",
"Yu",
""
],
[
"Yin",
"Jingjing",
""
],
[
"Han",
"Jiangyu",
""
],
[
"Ma",
"Lei",
""
],
[
"Lu",
"Heng",
""
]
] | TITLE: HYBRIDFORMER: improving SqueezeFormer with hybrid attention and NSR
mechanism
ABSTRACT: SqueezeFormer has recently shown impressive performance in automatic speech
recognition (ASR). However, its inference speed suffers from the quadratic
complexity of softmax-attention (SA). In addition, limited by the large
convolution kernel size, the local modeling ability of SqueezeFormer is
insufficient. In this paper, we propose a novel method HybridFormer to improve
SqueezeFormer in a fast and efficient way. Specifically, we first incorporate
linear attention (LA) and propose a hybrid LASA paradigm to increase the
model's inference speed. Second, a hybrid neural architecture search (NAS)
guided structural re-parameterization (SRep) mechanism, termed NSR, is proposed
to enhance the ability of the model to extract local interactions. Extensive
experiments conducted on the LibriSpeech dataset demonstrate that our proposed
HybridFormer can achieve a 9.1% relative word error rate (WER) reduction over
SqueezeFormer on the test-other dataset. Furthermore, when input speech is 30s,
the HybridFormer can improve the model's inference speed up to 18%. Our source
code is available online.
| no_new_dataset | 0.945601 |
2303.08652 | Nestor Prieto-Chavana | Nestor Prieto-Chavana, Julie Weeds, David Weir | Automated Query Generation for Evidence Collection from Web Search
Engines | null | null | null | null | cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is widely accepted that so-called facts can be checked by searching for
information on the Internet. This process requires a fact-checker to formulate
a search query based on the fact and to present it to a search engine. Then,
relevant and believable passages need to be identified in the search results
before a decision is made. This process is carried out by sub-editors at many
news and media organisations on a daily basis. Here, we ask the question as to
whether it is possible to automate the first step, that of query generation.
Can we automatically formulate search queries based on factual statements which
are similar to those formulated by human experts? Here, we consider similarity
both in terms of textual similarity and with respect to relevant documents
being returned by a search engine. First, we introduce a moderate-sized
evidence collection dataset which includes 390 factual statements together with
associated human-generated search queries and search results. Then, we
investigate generating queries using a number of rule-based and automatic text
generation methods based on pre-trained large language models (LLMs). We show
that these methods have different merits and propose a hybrid approach which
has superior performance in practice.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 14:32:00 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Prieto-Chavana",
"Nestor",
""
],
[
"Weeds",
"Julie",
""
],
[
"Weir",
"David",
""
]
] | TITLE: Automated Query Generation for Evidence Collection from Web Search
Engines
ABSTRACT: It is widely accepted that so-called facts can be checked by searching for
information on the Internet. This process requires a fact-checker to formulate
a search query based on the fact and to present it to a search engine. Then,
relevant and believable passages need to be identified in the search results
before a decision is made. This process is carried out by sub-editors at many
news and media organisations on a daily basis. Here, we ask the question as to
whether it is possible to automate the first step, that of query generation.
Can we automatically formulate search queries based on factual statements which
are similar to those formulated by human experts? Here, we consider similarity
both in terms of textual similarity and with respect to relevant documents
being returned by a search engine. First, we introduce a moderate-sized
evidence collection dataset which includes 390 factual statements together with
associated human-generated search queries and search results. Then, we
investigate generating queries using a number of rule-based and automatic text
generation methods based on pre-trained large language models (LLMs). We show
that these methods have different merits and propose a hybrid approach which
has superior performance in practice.
| new_dataset | 0.961134 |
2303.08658 | Jiaxu Zhang | Jiaxu Zhang, Junwu Weng, Di Kang, Fang Zhao, Shaoli Huang, Xuefei Zhe,
Linchao Bao, Ying Shan, Jue Wang and Zhigang Tu | Skinned Motion Retargeting with Residual Perception of Motion Semantics
& Geometry | CVPR 2023 | null | null | null | cs.CV cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A good motion retargeting cannot be reached without reasonable consideration
of source-target differences on both the skeleton and shape geometry levels. In
this work, we propose a novel Residual RETargeting network (R2ET) structure,
which relies on two neural modification modules, to adjust the source motions
to fit the target skeletons and shapes progressively. In particular, a
skeleton-aware module is introduced to preserve the source motion semantics. A
shape-aware module is designed to perceive the geometries of target characters
to reduce interpenetration and contact-missing. Driven by our explored
distance-based losses that explicitly model the motion semantics and geometry,
these two modules can learn residual motion modifications on the source motion
to generate plausible retargeted motion in a single inference without
post-processing. To balance these two modifications, we further present a
balancing gate to conduct linear interpolation between them. Extensive
experiments on the public dataset Mixamo demonstrate that our R2ET achieves the
state-of-the-art performance, and provides a good balance between the
preservation of motion semantics as well as the attenuation of interpenetration
and contact-missing. Code is available at https://github.com/Kebii/R2ET.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 14:41:26 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zhang",
"Jiaxu",
""
],
[
"Weng",
"Junwu",
""
],
[
"Kang",
"Di",
""
],
[
"Zhao",
"Fang",
""
],
[
"Huang",
"Shaoli",
""
],
[
"Zhe",
"Xuefei",
""
],
[
"Bao",
"Linchao",
""
],
[
"Shan",
"Ying",
""
],
[
"Wang",
"Jue",
""
],
[
"Tu",
"Zhigang",
""
]
] | TITLE: Skinned Motion Retargeting with Residual Perception of Motion Semantics
& Geometry
ABSTRACT: A good motion retargeting cannot be reached without reasonable consideration
of source-target differences on both the skeleton and shape geometry levels. In
this work, we propose a novel Residual RETargeting network (R2ET) structure,
which relies on two neural modification modules, to adjust the source motions
to fit the target skeletons and shapes progressively. In particular, a
skeleton-aware module is introduced to preserve the source motion semantics. A
shape-aware module is designed to perceive the geometries of target characters
to reduce interpenetration and contact-missing. Driven by our explored
distance-based losses that explicitly model the motion semantics and geometry,
these two modules can learn residual motion modifications on the source motion
to generate plausible retargeted motion in a single inference without
post-processing. To balance these two modifications, we further present a
balancing gate to conduct linear interpolation between them. Extensive
experiments on the public dataset Mixamo demonstrate that our R2ET achieves the
state-of-the-art performance, and provides a good balance between the
preservation of motion semantics as well as the attenuation of interpenetration
and contact-missing. Code is available at https://github.com/Kebii/R2ET.
| no_new_dataset | 0.945298 |
2303.08660 | Raja CSP Raman | Vijitha Kannumuru, Santhosh Kannan S P, Krithiga Shankar, Joy Larnyoh,
Rohith Mahadevan, Raja CSP Raman | Fashion-model pose recommendation and generation using Machine Learning | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Fashion-model pose is an important attribute in the fashion industry.
Creative directors, modeling production houses, and top photographers always
look for professional models able to pose. without the skill to correctly pose,
their chances of landing professional modeling employment are regrettably quite
little. There are occasions when models and photographers are unsure of the
best pose to strike while taking photographs. This research concentrates on
suggesting the fashion personnel a series of similar images based on the input
image. The image is segmented into different parts and similar images are
suggested for the user. This was achieved by calculating the color histogram of
the input image and applying the same for all the images in the dataset and
comparing the histograms. Synthetic images have become popular to avoid privacy
concerns and to overcome the high cost of photoshoots. Hence, this paper also
extends the work of generating synthetic images from the recommendation engine
using styleGAN to an extent.
| [
{
"version": "v1",
"created": "Sun, 19 Feb 2023 09:12:46 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Kannumuru",
"Vijitha",
""
],
[
"P",
"Santhosh Kannan S",
""
],
[
"Shankar",
"Krithiga",
""
],
[
"Larnyoh",
"Joy",
""
],
[
"Mahadevan",
"Rohith",
""
],
[
"Raman",
"Raja CSP",
""
]
] | TITLE: Fashion-model pose recommendation and generation using Machine Learning
ABSTRACT: Fashion-model pose is an important attribute in the fashion industry.
Creative directors, modeling production houses, and top photographers always
look for professional models able to pose. without the skill to correctly pose,
their chances of landing professional modeling employment are regrettably quite
little. There are occasions when models and photographers are unsure of the
best pose to strike while taking photographs. This research concentrates on
suggesting the fashion personnel a series of similar images based on the input
image. The image is segmented into different parts and similar images are
suggested for the user. This was achieved by calculating the color histogram of
the input image and applying the same for all the images in the dataset and
comparing the histograms. Synthetic images have become popular to avoid privacy
concerns and to overcome the high cost of photoshoots. Hence, this paper also
extends the work of generating synthetic images from the recommendation engine
using styleGAN to an extent.
| no_new_dataset | 0.952264 |
2303.08678 | Li Shen | Guanghao Li, Wansen Wu, Yan Sun, Li Shen, Baoyuan Wu, Dacheng Tao | Visual Prompt Based Personalized Federated Learning | 14 pages | null | null | null | cs.LG cs.CV cs.DC | http://creativecommons.org/licenses/by/4.0/ | As a popular paradigm of distributed learning, personalized federated
learning (PFL) allows personalized models to improve generalization ability and
robustness by utilizing knowledge from all distributed clients. Most existing
PFL algorithms tackle personalization in a model-centric way, such as
personalized layer partition, model regularization, and model interpolation,
which all fail to take into account the data characteristics of distributed
clients. In this paper, we propose a novel PFL framework for image
classification tasks, dubbed pFedPT, that leverages personalized visual prompts
to implicitly represent local data distribution information of clients and
provides that information to the aggregation model to help with classification
tasks. Specifically, in each round of pFedPT training, each client generates a
local personalized prompt related to local data distribution. Then, the local
model is trained on the input composed of raw data and a visual prompt to learn
the distribution information contained in the prompt. During model testing, the
aggregated model obtains prior knowledge of the data distributions based on the
prompts, which can be seen as an adaptive fine-tuning of the aggregation model
to improve model performances on different clients. Furthermore, the visual
prompt can be added as an orthogonal method to implement personalization on the
client for existing FL methods to boost their performance. Experiments on the
CIFAR10 and CIFAR100 datasets show that pFedPT outperforms several
state-of-the-art (SOTA) PFL algorithms by a large margin in various settings.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 15:02:15 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Li",
"Guanghao",
""
],
[
"Wu",
"Wansen",
""
],
[
"Sun",
"Yan",
""
],
[
"Shen",
"Li",
""
],
[
"Wu",
"Baoyuan",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Visual Prompt Based Personalized Federated Learning
ABSTRACT: As a popular paradigm of distributed learning, personalized federated
learning (PFL) allows personalized models to improve generalization ability and
robustness by utilizing knowledge from all distributed clients. Most existing
PFL algorithms tackle personalization in a model-centric way, such as
personalized layer partition, model regularization, and model interpolation,
which all fail to take into account the data characteristics of distributed
clients. In this paper, we propose a novel PFL framework for image
classification tasks, dubbed pFedPT, that leverages personalized visual prompts
to implicitly represent local data distribution information of clients and
provides that information to the aggregation model to help with classification
tasks. Specifically, in each round of pFedPT training, each client generates a
local personalized prompt related to local data distribution. Then, the local
model is trained on the input composed of raw data and a visual prompt to learn
the distribution information contained in the prompt. During model testing, the
aggregated model obtains prior knowledge of the data distributions based on the
prompts, which can be seen as an adaptive fine-tuning of the aggregation model
to improve model performances on different clients. Furthermore, the visual
prompt can be added as an orthogonal method to implement personalization on the
client for existing FL methods to boost their performance. Experiments on the
CIFAR10 and CIFAR100 datasets show that pFedPT outperforms several
state-of-the-art (SOTA) PFL algorithms by a large margin in various settings.
| no_new_dataset | 0.944434 |
2303.08695 | Shuja Khalid | Shuja Khalid, Frank Rudzicz | RefiNeRF: Modelling dynamic neural radiance fields with inconsistent or
missing camera parameters | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Novel view synthesis (NVS) is a challenging task in computer vision that
involves synthesizing new views of a scene from a limited set of input images.
Neural Radiance Fields (NeRF) have emerged as a powerful approach to address
this problem, but they require accurate knowledge of camera \textit{intrinsic}
and \textit{extrinsic} parameters. Traditionally, structure-from-motion (SfM)
and multi-view stereo (MVS) approaches have been used to extract camera
parameters, but these methods can be unreliable and may fail in certain cases.
In this paper, we propose a novel technique that leverages unposed images from
dynamic datasets, such as the NVIDIA dynamic scenes dataset, to learn camera
parameters directly from data. Our approach is highly extensible and can be
integrated into existing NeRF architectures with minimal modifications. We
demonstrate the effectiveness of our method on a variety of static and dynamic
scenes and show that it outperforms traditional SfM and MVS approaches. The
code for our method is publicly available at
\href{https://github.com/redacted/refinerf}{https://github.com/redacted/refinerf}.
Our approach offers a promising new direction for improving the accuracy and
robustness of NVS using NeRF, and we anticipate that it will be a valuable tool
for a wide range of applications in computer vision and graphics.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 15:27:18 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Khalid",
"Shuja",
""
],
[
"Rudzicz",
"Frank",
""
]
] | TITLE: RefiNeRF: Modelling dynamic neural radiance fields with inconsistent or
missing camera parameters
ABSTRACT: Novel view synthesis (NVS) is a challenging task in computer vision that
involves synthesizing new views of a scene from a limited set of input images.
Neural Radiance Fields (NeRF) have emerged as a powerful approach to address
this problem, but they require accurate knowledge of camera \textit{intrinsic}
and \textit{extrinsic} parameters. Traditionally, structure-from-motion (SfM)
and multi-view stereo (MVS) approaches have been used to extract camera
parameters, but these methods can be unreliable and may fail in certain cases.
In this paper, we propose a novel technique that leverages unposed images from
dynamic datasets, such as the NVIDIA dynamic scenes dataset, to learn camera
parameters directly from data. Our approach is highly extensible and can be
integrated into existing NeRF architectures with minimal modifications. We
demonstrate the effectiveness of our method on a variety of static and dynamic
scenes and show that it outperforms traditional SfM and MVS approaches. The
code for our method is publicly available at
\href{https://github.com/redacted/refinerf}{https://github.com/redacted/refinerf}.
Our approach offers a promising new direction for improving the accuracy and
robustness of NVS using NeRF, and we anticipate that it will be a valuable tool
for a wide range of applications in computer vision and graphics.
| no_new_dataset | 0.951188 |
2303.08704 | Rui Zhou | Rui Zhou, Yan Niu | Multi-Exposure HDR Composition by Gated Swin Transformer | 7 pages, 4 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fusing a sequence of perfectly aligned images captured at various exposures,
has shown great potential to approach High Dynamic Range (HDR) imaging by
sensors with limited dynamic range. However, in the presence of large motion of
scene objects or the camera, mis-alignment is almost inevitable and leads to
the notorious ``ghost'' artifacts. Besides, factors such as the noise in the
dark region or color saturation in the over-bright region may also fail to fill
local image details to the HDR image. This paper provides a novel
multi-exposure fusion model based on Swin Transformer. Particularly, we design
feature selection gates, which are integrated with the feature extraction
layers to detect outliers and block them from HDR image synthesis. To
reconstruct the missing local details by well-aligned and properly-exposed
regions, we exploit the long distance contextual dependency in the
exposure-space pyramid by the self-attention mechanism. Extensive numerical and
visual evaluation has been conducted on a variety of benchmark datasets. The
experiments show that our model achieves the accuracy on par with current top
performing multi-exposure HDR imaging models, while gaining higher efficiency.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 15:38:43 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Zhou",
"Rui",
""
],
[
"Niu",
"Yan",
""
]
] | TITLE: Multi-Exposure HDR Composition by Gated Swin Transformer
ABSTRACT: Fusing a sequence of perfectly aligned images captured at various exposures,
has shown great potential to approach High Dynamic Range (HDR) imaging by
sensors with limited dynamic range. However, in the presence of large motion of
scene objects or the camera, mis-alignment is almost inevitable and leads to
the notorious ``ghost'' artifacts. Besides, factors such as the noise in the
dark region or color saturation in the over-bright region may also fail to fill
local image details to the HDR image. This paper provides a novel
multi-exposure fusion model based on Swin Transformer. Particularly, we design
feature selection gates, which are integrated with the feature extraction
layers to detect outliers and block them from HDR image synthesis. To
reconstruct the missing local details by well-aligned and properly-exposed
regions, we exploit the long distance contextual dependency in the
exposure-space pyramid by the self-attention mechanism. Extensive numerical and
visual evaluation has been conducted on a variety of benchmark datasets. The
experiments show that our model achieves the accuracy on par with current top
performing multi-exposure HDR imaging models, while gaining higher efficiency.
| no_new_dataset | 0.950273 |
2303.08709 | Carmine Dodaro | Matteo Cardellini, Paolo De Nardi, Carmine Dodaro, Giuseppe Galat\`a,
Anna Giardini, Marco Maratea, Ivan Porro | Solving Rehabilitation Scheduling problems via a Two-Phase ASP approach | Under consideration in Theory and Practice of Logic Programming
(TPLP) | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A core part of the rehabilitation scheduling process consists of planning
rehabilitation physiotherapy sessions for patients, by assigning proper
operators to them in a certain time slot of a given day, taking into account
several legal, medical and ethical requirements and optimizations, e.g.,
patient's preferences and operator's work balancing. Being able to efficiently
solve such problem is of upmost importance, in particular after the COVID-19
pandemic that significantly increased rehabilitation's needs.
In this paper, we present a two-phase solution to rehabilitation scheduling
based on Answer Set Programming, which proved to be an effective tool for
solving practical scheduling problems. We first present a general encoding, and
then add domain specific optimizations. Results of experiments performed on
both synthetic and real benchmarks, the latter provided by ICS Maugeri, show
the effectiveness of our solution as well as the impact of our domain specific
optimizations. Under consideration in Theory and Practice of Logic Programming
(TPLP).
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 15:43:46 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Cardellini",
"Matteo",
""
],
[
"De Nardi",
"Paolo",
""
],
[
"Dodaro",
"Carmine",
""
],
[
"Galatà",
"Giuseppe",
""
],
[
"Giardini",
"Anna",
""
],
[
"Maratea",
"Marco",
""
],
[
"Porro",
"Ivan",
""
]
] | TITLE: Solving Rehabilitation Scheduling problems via a Two-Phase ASP approach
ABSTRACT: A core part of the rehabilitation scheduling process consists of planning
rehabilitation physiotherapy sessions for patients, by assigning proper
operators to them in a certain time slot of a given day, taking into account
several legal, medical and ethical requirements and optimizations, e.g.,
patient's preferences and operator's work balancing. Being able to efficiently
solve such problem is of upmost importance, in particular after the COVID-19
pandemic that significantly increased rehabilitation's needs.
In this paper, we present a two-phase solution to rehabilitation scheduling
based on Answer Set Programming, which proved to be an effective tool for
solving practical scheduling problems. We first present a general encoding, and
then add domain specific optimizations. Results of experiments performed on
both synthetic and real benchmarks, the latter provided by ICS Maugeri, show
the effectiveness of our solution as well as the impact of our domain specific
optimizations. Under consideration in Theory and Practice of Logic Programming
(TPLP).
| no_new_dataset | 0.940408 |
2303.08722 | Chen Xu | Chen Xu, Jun Xu, Xu Chen, Zhenghua Dong, Ji-Rong Wen | Dually Enhanced Propensity Score Estimation in Sequential Recommendation | Accepted in CIKM2022 | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sequential recommender systems train their models based on a large amount of
implicit user feedback data and may be subject to biases when users are
systematically under/over-exposed to certain items. Unbiased learning based on
inverse propensity scores (IPS), which estimate the probability of observing a
user-item pair given the historical information, has been proposed to address
the issue. In these methods, propensity score estimation is usually limited to
the view of item, that is, treating the feedback data as sequences of items
that interacted with the users. However, the feedback data can also be treated
from the view of user, as the sequences of users that interact with the items.
Moreover, the two views can jointly enhance the propensity score estimation.
Inspired by the observation, we propose to estimate the propensity scores from
the views of user and item, called Dually Enhanced Propensity Score Estimation
(DEPS). Specifically, given a target user-item pair and the corresponding item
and user interaction sequences, DEPS firstly constructs a time-aware causal
graph to represent the user-item observational probability. According to the
graph, two complementary propensity scores are estimated from the views of item
and user, respectively, based on the same set of user feedback data. Finally,
two transformers are designed to make the final preference prediction.
Theoretical analysis showed the unbiasedness and variance of DEPS. Experimental
results on three publicly available and an industrial datasets demonstrated
that DEPS can significantly outperform the state-of-the-art baselines.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 16:08:40 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Xu",
"Chen",
""
],
[
"Xu",
"Jun",
""
],
[
"Chen",
"Xu",
""
],
[
"Dong",
"Zhenghua",
""
],
[
"Wen",
"Ji-Rong",
""
]
] | TITLE: Dually Enhanced Propensity Score Estimation in Sequential Recommendation
ABSTRACT: Sequential recommender systems train their models based on a large amount of
implicit user feedback data and may be subject to biases when users are
systematically under/over-exposed to certain items. Unbiased learning based on
inverse propensity scores (IPS), which estimate the probability of observing a
user-item pair given the historical information, has been proposed to address
the issue. In these methods, propensity score estimation is usually limited to
the view of item, that is, treating the feedback data as sequences of items
that interacted with the users. However, the feedback data can also be treated
from the view of user, as the sequences of users that interact with the items.
Moreover, the two views can jointly enhance the propensity score estimation.
Inspired by the observation, we propose to estimate the propensity scores from
the views of user and item, called Dually Enhanced Propensity Score Estimation
(DEPS). Specifically, given a target user-item pair and the corresponding item
and user interaction sequences, DEPS firstly constructs a time-aware causal
graph to represent the user-item observational probability. According to the
graph, two complementary propensity scores are estimated from the views of item
and user, respectively, based on the same set of user feedback data. Finally,
two transformers are designed to make the final preference prediction.
Theoretical analysis showed the unbiasedness and variance of DEPS. Experimental
results on three publicly available and an industrial datasets demonstrated
that DEPS can significantly outperform the state-of-the-art baselines.
| no_new_dataset | 0.941115 |
2303.08729 | Mootez Saad | Himesh Nandani, Mootez Saad, Tushar Sharma | DACOS-A Manually Annotated Dataset of Code Smells | 4 pages | null | null | null | cs.SE cs.AI cs.LG cs.PL | http://creativecommons.org/licenses/by/4.0/ | Researchers apply machine-learning techniques for code smell detection to
counter the subjectivity of many code smells. Such approaches need a large,
manually annotated dataset for training and benchmarking. Existing literature
offers a few datasets; however, they are small in size and, more importantly,
do not focus on the subjective code snippets. In this paper, we present DACOS,
a manually annotated dataset containing 10,267 annotations for 5,192 code
snippets. The dataset targets three kinds of code smells at different
granularity: multifaceted abstraction, complex method, and long parameter list.
The dataset is created in two phases. The first phase helps us identify the
code snippets that are potentially subjective by determining the thresholds of
metrics used to detect a smell. The second phase collects annotations for
potentially subjective snippets. We also offer an extended dataset DACOSX that
includes definitely benign and definitely smelly snippets by using the
thresholds identified in the first phase. We have developed TagMan, a web
application to help annotators view and mark the snippets one-by-one and record
the provided annotations. We make the datasets and the web application
accessible publicly. This dataset will help researchers working on smell
detection techniques to build relevant and context-aware machine-learning
models.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 16:13:40 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Nandani",
"Himesh",
""
],
[
"Saad",
"Mootez",
""
],
[
"Sharma",
"Tushar",
""
]
] | TITLE: DACOS-A Manually Annotated Dataset of Code Smells
ABSTRACT: Researchers apply machine-learning techniques for code smell detection to
counter the subjectivity of many code smells. Such approaches need a large,
manually annotated dataset for training and benchmarking. Existing literature
offers a few datasets; however, they are small in size and, more importantly,
do not focus on the subjective code snippets. In this paper, we present DACOS,
a manually annotated dataset containing 10,267 annotations for 5,192 code
snippets. The dataset targets three kinds of code smells at different
granularity: multifaceted abstraction, complex method, and long parameter list.
The dataset is created in two phases. The first phase helps us identify the
code snippets that are potentially subjective by determining the thresholds of
metrics used to detect a smell. The second phase collects annotations for
potentially subjective snippets. We also offer an extended dataset DACOSX that
includes definitely benign and definitely smelly snippets by using the
thresholds identified in the first phase. We have developed TagMan, a web
application to help annotators view and mark the snippets one-by-one and record
the provided annotations. We make the datasets and the web application
accessible publicly. This dataset will help researchers working on smell
detection techniques to build relevant and context-aware machine-learning
models.
| new_dataset | 0.96051 |
2303.08745 | Wail Gueaieb | Mohammed Abouheaf and Derek Boase and Wail Gueaieb and Davide Spinello
and Salah Al-Sharhan | Real-Time Measurement-Driven Reinforcement Learning Control Approach for
Uncertain Nonlinear Systems | 15 pages | Engineering Applications of Artificial Intelligence (Elsevier),
vol. 122, June 2023 | 10.1016/j.engappai.2023.106029 | null | eess.SY cs.RO cs.SY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The paper introduces an interactive machine learning mechanism to process the
measurements of an uncertain, nonlinear dynamic process and hence advise an
actuation strategy in real-time. For concept demonstration, a
trajectory-following optimization problem of a Kinova robotic arm is solved
using an integral reinforcement learning approach with guaranteed stability for
slowly varying dynamics. The solution is implemented using a model-free value
iteration process to solve the integral temporal difference equations of the
problem. The performance of the proposed technique is benchmarked against that
of another model-free high-order approach and is validated for dynamic payload
and disturbances. Unlike its benchmark, the proposed adaptive strategy is
capable of handling extreme process variations. This is experimentally
demonstrated by introducing static and time-varying payloads close to the rated
maximum payload capacity of the manipulator arm. The comparison algorithm
exhibited up to a seven-fold percent overshoot compared to the proposed
integral reinforcement learning solution. The robustness of the algorithm is
further validated by disturbing the real-time adapted strategy gains with a
white noise of a standard deviation as high as 5%.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 16:36:26 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Abouheaf",
"Mohammed",
""
],
[
"Boase",
"Derek",
""
],
[
"Gueaieb",
"Wail",
""
],
[
"Spinello",
"Davide",
""
],
[
"Al-Sharhan",
"Salah",
""
]
] | TITLE: Real-Time Measurement-Driven Reinforcement Learning Control Approach for
Uncertain Nonlinear Systems
ABSTRACT: The paper introduces an interactive machine learning mechanism to process the
measurements of an uncertain, nonlinear dynamic process and hence advise an
actuation strategy in real-time. For concept demonstration, a
trajectory-following optimization problem of a Kinova robotic arm is solved
using an integral reinforcement learning approach with guaranteed stability for
slowly varying dynamics. The solution is implemented using a model-free value
iteration process to solve the integral temporal difference equations of the
problem. The performance of the proposed technique is benchmarked against that
of another model-free high-order approach and is validated for dynamic payload
and disturbances. Unlike its benchmark, the proposed adaptive strategy is
capable of handling extreme process variations. This is experimentally
demonstrated by introducing static and time-varying payloads close to the rated
maximum payload capacity of the manipulator arm. The comparison algorithm
exhibited up to a seven-fold percent overshoot compared to the proposed
integral reinforcement learning solution. The robustness of the algorithm is
further validated by disturbing the real-time adapted strategy gains with a
white noise of a standard deviation as high as 5%.
| no_new_dataset | 0.94428 |
2303.08746 | Marcin Pietron | Krzysztof Stuglik, Piotr Listkiewicz, Mateusz Kulczyk, Marcin Pietron | J-Parallelio -- automatic parallelization framework for Java virtual
machine code | null | null | null | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | Manual translation of the algorithms from sequential version to its parallel
counterpart is time consuming and can be done only with the specific knowledge
of hardware accelerator architecture, parallel programming or programming
environment. The automation of this process makes porting the code much easier
and faster. The key aspect in this case is how efficient the generated parallel
code will be. The paper describes J-Parallelio, the framework for automatic
analysis of the bytecode source codes and its parallelisation on multicore
processors. The process consists of a few steps. First step is a process of
decompilation of JVM and its translation to internal abstract syntax tree, the
dependency extraction and memory analysis is performed. Finally, the mapping
process is performed which consists of a set of rules responsible for
translating the input virtual machine source code to its parallel version. The
main novelty is that it can deal with pure Java virtual machine and can
generate parallel code for multicore processors. This makes the system portable
and it can work with different languages based on JVM after some small
modifications. The efficiency of automatically translated source codes were
compared with their manually written counterparts on chosen benchmarks.
| [
{
"version": "v1",
"created": "Tue, 7 Feb 2023 19:10:36 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Stuglik",
"Krzysztof",
""
],
[
"Listkiewicz",
"Piotr",
""
],
[
"Kulczyk",
"Mateusz",
""
],
[
"Pietron",
"Marcin",
""
]
] | TITLE: J-Parallelio -- automatic parallelization framework for Java virtual
machine code
ABSTRACT: Manual translation of the algorithms from sequential version to its parallel
counterpart is time consuming and can be done only with the specific knowledge
of hardware accelerator architecture, parallel programming or programming
environment. The automation of this process makes porting the code much easier
and faster. The key aspect in this case is how efficient the generated parallel
code will be. The paper describes J-Parallelio, the framework for automatic
analysis of the bytecode source codes and its parallelisation on multicore
processors. The process consists of a few steps. First step is a process of
decompilation of JVM and its translation to internal abstract syntax tree, the
dependency extraction and memory analysis is performed. Finally, the mapping
process is performed which consists of a set of rules responsible for
translating the input virtual machine source code to its parallel version. The
main novelty is that it can deal with pure Java virtual machine and can
generate parallel code for multicore processors. This makes the system portable
and it can work with different languages based on JVM after some small
modifications. The efficiency of automatically translated source codes were
compared with their manually written counterparts on chosen benchmarks.
| no_new_dataset | 0.941547 |
2303.08747 | Akhil Meethal | Akhil Meethal, Eric Granger, Marco Pedersoli | Cascaded Zoom-in Detector for High Resolution Aerial Images | 12 pages, 7 figures | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Detecting objects in aerial images is challenging because they are typically
composed of crowded small objects distributed non-uniformly over
high-resolution images. Density cropping is a widely used method to improve
this small object detection where the crowded small object regions are
extracted and processed in high resolution. However, this is typically
accomplished by adding other learnable components, thus complicating the
training and inference over a standard detection process. In this paper, we
propose an efficient Cascaded Zoom-in (CZ) detector that re-purposes the
detector itself for density-guided training and inference. During training,
density crops are located, labeled as a new class, and employed to augment the
training dataset. During inference, the density crops are first detected along
with the base class objects, and then input for a second stage of inference.
This approach is easily integrated into any detector, and creates no
significant change in the standard detection process, like the uniform cropping
approach popular in aerial image detection. Experimental results on the aerial
images of the challenging VisDrone and DOTA datasets verify the benefits of the
proposed approach. The proposed CZ detector also provides state-of-the-art
results over uniform cropping and other density cropping methods on the
VisDrone dataset, increasing the detection mAP of small objects by more than 3
points.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 16:39:21 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Meethal",
"Akhil",
""
],
[
"Granger",
"Eric",
""
],
[
"Pedersoli",
"Marco",
""
]
] | TITLE: Cascaded Zoom-in Detector for High Resolution Aerial Images
ABSTRACT: Detecting objects in aerial images is challenging because they are typically
composed of crowded small objects distributed non-uniformly over
high-resolution images. Density cropping is a widely used method to improve
this small object detection where the crowded small object regions are
extracted and processed in high resolution. However, this is typically
accomplished by adding other learnable components, thus complicating the
training and inference over a standard detection process. In this paper, we
propose an efficient Cascaded Zoom-in (CZ) detector that re-purposes the
detector itself for density-guided training and inference. During training,
density crops are located, labeled as a new class, and employed to augment the
training dataset. During inference, the density crops are first detected along
with the base class objects, and then input for a second stage of inference.
This approach is easily integrated into any detector, and creates no
significant change in the standard detection process, like the uniform cropping
approach popular in aerial image detection. Experimental results on the aerial
images of the challenging VisDrone and DOTA datasets verify the benefits of the
proposed approach. The proposed CZ detector also provides state-of-the-art
results over uniform cropping and other density cropping methods on the
VisDrone dataset, increasing the detection mAP of small objects by more than 3
points.
| no_new_dataset | 0.952042 |
2303.08760 | Jaehyung Choi | Young Shin Kim, Hyangju Kim, Jaehyung Choi | Deep Calibration With Artificial Neural Network: A Performance
Comparison on Option Pricing Models | 26 pages, 3 figures | null | null | null | q-fin.MF cs.CE cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper explores Artificial Neural Network (ANN) as a model-free solution
for a calibration algorithm of option pricing models. We construct ANNs to
calibrate parameters for two well-known GARCH-type option pricing models:
Duan's GARCH and the classical tempered stable GARCH that significantly improve
upon the limitation of the Black-Scholes model but have suffered from
computation complexity. To mitigate this technical difficulty, we train ANNs
with a dataset generated by Monte Carlo Simulation (MCS) method and apply them
to calibrate optimal parameters. The performance results indicate that the ANN
approach consistently outperforms MCS and takes advantage of faster computation
times once trained. The Greeks of options are also discussed.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 16:57:10 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Kim",
"Young Shin",
""
],
[
"Kim",
"Hyangju",
""
],
[
"Choi",
"Jaehyung",
""
]
] | TITLE: Deep Calibration With Artificial Neural Network: A Performance
Comparison on Option Pricing Models
ABSTRACT: This paper explores Artificial Neural Network (ANN) as a model-free solution
for a calibration algorithm of option pricing models. We construct ANNs to
calibrate parameters for two well-known GARCH-type option pricing models:
Duan's GARCH and the classical tempered stable GARCH that significantly improve
upon the limitation of the Black-Scholes model but have suffered from
computation complexity. To mitigate this technical difficulty, we train ANNs
with a dataset generated by Monte Carlo Simulation (MCS) method and apply them
to calibrate optimal parameters. The performance results indicate that the ANN
approach consistently outperforms MCS and takes advantage of faster computation
times once trained. The Greeks of options are also discussed.
| no_new_dataset | 0.951006 |
2303.08784 | Aditay Tripathi | Aditay Tripathi, Anand Mishra, Anirban Chakraborty | Query-guided Attention in Vision Transformers for Localizing Objects
Using a Single Sketch | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this work, we investigate the problem of sketch-based object localization
on natural images, where given a crude hand-drawn sketch of an object, the goal
is to localize all the instances of the same object on the target image. This
problem proves difficult due to the abstract nature of hand-drawn sketches,
variations in the style and quality of sketches, and the large domain gap
existing between the sketches and the natural images. To mitigate these
challenges, existing works proposed attention-based frameworks to incorporate
query information into the image features. However, in these works, the query
features are incorporated after the image features have already been
independently learned, leading to inadequate alignment. In contrast, we propose
a sketch-guided vision transformer encoder that uses cross-attention after each
block of the transformer-based image encoder to learn query-conditioned image
features leading to stronger alignment with the query sketch. Further, at the
output of the decoder, the object and the sketch features are refined to bring
the representation of relevant objects closer to the sketch query and thereby
improve the localization. The proposed model also generalizes to the object
categories not seen during training, as the target image features learned by
our method are query-aware. Our localization framework can also utilize
multiple sketch queries via a trainable novel sketch fusion strategy. The model
is evaluated on the images from the public object detection benchmark, namely
MS-COCO, using the sketch queries from QuickDraw! and Sketchy datasets.
Compared with existing localization methods, the proposed approach gives a
$6.6\%$ and $8.0\%$ improvement in mAP for seen objects using sketch queries
from QuickDraw! and Sketchy datasets, respectively, and a $12.2\%$ improvement
in AP@50 for large objects that are `unseen' during training.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 17:26:17 GMT"
}
] | 2023-03-16T00:00:00 | [
[
"Tripathi",
"Aditay",
""
],
[
"Mishra",
"Anand",
""
],
[
"Chakraborty",
"Anirban",
""
]
] | TITLE: Query-guided Attention in Vision Transformers for Localizing Objects
Using a Single Sketch
ABSTRACT: In this work, we investigate the problem of sketch-based object localization
on natural images, where given a crude hand-drawn sketch of an object, the goal
is to localize all the instances of the same object on the target image. This
problem proves difficult due to the abstract nature of hand-drawn sketches,
variations in the style and quality of sketches, and the large domain gap
existing between the sketches and the natural images. To mitigate these
challenges, existing works proposed attention-based frameworks to incorporate
query information into the image features. However, in these works, the query
features are incorporated after the image features have already been
independently learned, leading to inadequate alignment. In contrast, we propose
a sketch-guided vision transformer encoder that uses cross-attention after each
block of the transformer-based image encoder to learn query-conditioned image
features leading to stronger alignment with the query sketch. Further, at the
output of the decoder, the object and the sketch features are refined to bring
the representation of relevant objects closer to the sketch query and thereby
improve the localization. The proposed model also generalizes to the object
categories not seen during training, as the target image features learned by
our method are query-aware. Our localization framework can also utilize
multiple sketch queries via a trainable novel sketch fusion strategy. The model
is evaluated on the images from the public object detection benchmark, namely
MS-COCO, using the sketch queries from QuickDraw! and Sketchy datasets.
Compared with existing localization methods, the proposed approach gives a
$6.6\%$ and $8.0\%$ improvement in mAP for seen objects using sketch queries
from QuickDraw! and Sketchy datasets, respectively, and a $12.2\%$ improvement
in AP@50 for large objects that are `unseen' during training.
| no_new_dataset | 0.950457 |
2102.07725 | Berivan Isik | Berivan Isik, Kristy Choi, Xin Zheng, Tsachy Weissman, Stefano Ermon,
H.-S. Philip Wong, Armin Alaghi | Neural Network Compression for Noisy Storage Devices | Published at the ACM Transactions on Embedded Computing Systems
(TECS), 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Compression and efficient storage of neural network (NN) parameters is
critical for applications that run on resource-constrained devices. Despite the
significant progress in NN model compression, there has been considerably less
investigation in the actual \textit{physical} storage of NN parameters.
Conventionally, model compression and physical storage are decoupled, as
digital storage media with error-correcting codes (ECCs) provide robust
error-free storage. However, this decoupled approach is inefficient as it
ignores the overparameterization present in most NNs and forces the memory
device to allocate the same amount of resources to every bit of information
regardless of its importance. In this work, we investigate analog memory
devices as an alternative to digital media -- one that naturally provides a way
to add more protection for significant bits unlike its counterpart, but is
noisy and may compromise the stored model's performance if used naively. We
develop a variety of robust coding strategies for NN weight storage on analog
devices, and propose an approach to jointly optimize model compression and
memory resource allocation. We then demonstrate the efficacy of our approach on
models trained on MNIST, CIFAR-10 and ImageNet datasets for existing
compression techniques. Compared to conventional error-free digital storage,
our method reduces the memory footprint by up to one order of magnitude,
without significantly compromising the stored model's accuracy.
| [
{
"version": "v1",
"created": "Mon, 15 Feb 2021 18:19:07 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 02:45:14 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Isik",
"Berivan",
""
],
[
"Choi",
"Kristy",
""
],
[
"Zheng",
"Xin",
""
],
[
"Weissman",
"Tsachy",
""
],
[
"Ermon",
"Stefano",
""
],
[
"Wong",
"H. -S. Philip",
""
],
[
"Alaghi",
"Armin",
""
]
] | TITLE: Neural Network Compression for Noisy Storage Devices
ABSTRACT: Compression and efficient storage of neural network (NN) parameters is
critical for applications that run on resource-constrained devices. Despite the
significant progress in NN model compression, there has been considerably less
investigation in the actual \textit{physical} storage of NN parameters.
Conventionally, model compression and physical storage are decoupled, as
digital storage media with error-correcting codes (ECCs) provide robust
error-free storage. However, this decoupled approach is inefficient as it
ignores the overparameterization present in most NNs and forces the memory
device to allocate the same amount of resources to every bit of information
regardless of its importance. In this work, we investigate analog memory
devices as an alternative to digital media -- one that naturally provides a way
to add more protection for significant bits unlike its counterpart, but is
noisy and may compromise the stored model's performance if used naively. We
develop a variety of robust coding strategies for NN weight storage on analog
devices, and propose an approach to jointly optimize model compression and
memory resource allocation. We then demonstrate the efficacy of our approach on
models trained on MNIST, CIFAR-10 and ImageNet datasets for existing
compression techniques. Compared to conventional error-free digital storage,
our method reduces the memory footprint by up to one order of magnitude,
without significantly compromising the stored model's accuracy.
| no_new_dataset | 0.947672 |
2103.03076 | Yanghao Zhang | Fu Wang, Yanghao Zhang, Yanbin Zheng, Wenjie Ruan | Dynamic Efficient Adversarial Training Guided by Gradient Magnitude | 18 pages, 6 figures | null | null | null | cs.LG cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adversarial training is an effective but time-consuming way to train robust
deep neural networks that can withstand strong adversarial attacks. As a
response to its inefficiency, we propose Dynamic Efficient Adversarial Training
(DEAT), which gradually increases the adversarial iteration during training. We
demonstrate that the gradient's magnitude correlates with the curvature of the
trained model's loss landscape, allowing it to reflect the effect of
adversarial training. Therefore, based on the magnitude of the gradient, we
propose a general acceleration strategy, M+ acceleration, which enables an
automatic and highly effective method of adjusting the training procedure. M+
acceleration is computationally efficient and easy to implement. It is suited
for DEAT and compatible with the majority of existing adversarial training
techniques. Extensive experiments have been done on CIFAR-10 and ImageNet
datasets with various training environments. The results show that the proposed
M+ acceleration significantly improves the training efficiency of existing
adversarial training methods while achieving similar robustness performance.
This demonstrates that the strategy is highly adaptive and offers a valuable
solution for automatic adversarial training.
| [
{
"version": "v1",
"created": "Thu, 4 Mar 2021 14:57:53 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 13:05:12 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Wang",
"Fu",
""
],
[
"Zhang",
"Yanghao",
""
],
[
"Zheng",
"Yanbin",
""
],
[
"Ruan",
"Wenjie",
""
]
] | TITLE: Dynamic Efficient Adversarial Training Guided by Gradient Magnitude
ABSTRACT: Adversarial training is an effective but time-consuming way to train robust
deep neural networks that can withstand strong adversarial attacks. As a
response to its inefficiency, we propose Dynamic Efficient Adversarial Training
(DEAT), which gradually increases the adversarial iteration during training. We
demonstrate that the gradient's magnitude correlates with the curvature of the
trained model's loss landscape, allowing it to reflect the effect of
adversarial training. Therefore, based on the magnitude of the gradient, we
propose a general acceleration strategy, M+ acceleration, which enables an
automatic and highly effective method of adjusting the training procedure. M+
acceleration is computationally efficient and easy to implement. It is suited
for DEAT and compatible with the majority of existing adversarial training
techniques. Extensive experiments have been done on CIFAR-10 and ImageNet
datasets with various training environments. The results show that the proposed
M+ acceleration significantly improves the training efficiency of existing
adversarial training methods while achieving similar robustness performance.
This demonstrates that the strategy is highly adaptive and offers a valuable
solution for automatic adversarial training.
| no_new_dataset | 0.949809 |
2105.14655 | Zhuoran Qiao | Zhuoran Qiao, Anders S. Christensen, Matthew Welborn, Frederick R.
Manby, Anima Anandkumar, Thomas F. Miller III | Informing Geometric Deep Learning with Electronic Interactions to
Accelerate Quantum Chemistry | null | Proceedings of the National Academy of Sciences 119.31 (2022):
e2205221119 | 10.1073/pnas.2205221119 | null | cs.LG physics.chem-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Predicting electronic energies, densities, and related chemical properties
can facilitate the discovery of novel catalysts, medicines, and battery
materials. By developing a physics-inspired equivariant neural network, we
introduce a method to learn molecular representations based on the electronic
interactions among atomic orbitals. Our method, OrbNet-Equi, leverages
efficient tight-binding simulations and learned mappings to recover high
fidelity quantum chemical properties. OrbNet-Equi models a wide spectrum of
target properties with an accuracy consistently better than standard machine
learning methods and a speed orders of magnitude greater than density
functional theory. Despite only using training samples collected from readily
available small-molecule libraries, OrbNet-Equi outperforms traditional methods
on comprehensive downstream benchmarks that encompass diverse main-group
chemical processes. Our method also describes interactions in challenging
charge-transfer complexes and open-shell systems. We anticipate that the
strategy presented here will help to expand opportunities for studies in
chemistry and materials science, where the acquisition of experimental or
reference training data is costly.
| [
{
"version": "v1",
"created": "Mon, 31 May 2021 00:48:18 GMT"
},
{
"version": "v2",
"created": "Sun, 6 Jun 2021 02:47:31 GMT"
},
{
"version": "v3",
"created": "Mon, 25 Oct 2021 18:56:07 GMT"
},
{
"version": "v4",
"created": "Fri, 1 Apr 2022 22:54:06 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Qiao",
"Zhuoran",
""
],
[
"Christensen",
"Anders S.",
""
],
[
"Welborn",
"Matthew",
""
],
[
"Manby",
"Frederick R.",
""
],
[
"Anandkumar",
"Anima",
""
],
[
"Miller",
"Thomas F.",
"III"
]
] | TITLE: Informing Geometric Deep Learning with Electronic Interactions to
Accelerate Quantum Chemistry
ABSTRACT: Predicting electronic energies, densities, and related chemical properties
can facilitate the discovery of novel catalysts, medicines, and battery
materials. By developing a physics-inspired equivariant neural network, we
introduce a method to learn molecular representations based on the electronic
interactions among atomic orbitals. Our method, OrbNet-Equi, leverages
efficient tight-binding simulations and learned mappings to recover high
fidelity quantum chemical properties. OrbNet-Equi models a wide spectrum of
target properties with an accuracy consistently better than standard machine
learning methods and a speed orders of magnitude greater than density
functional theory. Despite only using training samples collected from readily
available small-molecule libraries, OrbNet-Equi outperforms traditional methods
on comprehensive downstream benchmarks that encompass diverse main-group
chemical processes. Our method also describes interactions in challenging
charge-transfer complexes and open-shell systems. We anticipate that the
strategy presented here will help to expand opportunities for studies in
chemistry and materials science, where the acquisition of experimental or
reference training data is costly.
| no_new_dataset | 0.946941 |
2106.07732 | Changan Chen | Changan Chen, Wei Sun, David Harwath, Kristen Grauman | Learning Audio-Visual Dereverberation | Accepted at ICASSP 2023. This is the longer version of the five-page
camera-ready paper. Project page:
https://vision.cs.utexas.edu/projects/learning-audio-visual-dereverberation | null | null | null | cs.SD cs.CV cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | Reverberation not only degrades the quality of speech for human perception,
but also severely impacts the accuracy of automatic speech recognition. Prior
work attempts to remove reverberation based on the audio modality only. Our
idea is to learn to dereverberate speech from audio-visual observations. The
visual environment surrounding a human speaker reveals important cues about the
room geometry, materials, and speaker location, all of which influence the
precise reverberation effects. We introduce Visually-Informed Dereverberation
of Audio (VIDA), an end-to-end approach that learns to remove reverberation
based on both the observed monaural sound and visual scene. In support of this
new task, we develop a large-scale dataset SoundSpaces-Speech that uses
realistic acoustic renderings of speech in real-world 3D scans of homes
offering a variety of room acoustics. Demonstrating our approach on both
simulated and real imagery for speech enhancement, speech recognition, and
speaker identification, we show it achieves state-of-the-art performance and
substantially improves over audio-only methods.
| [
{
"version": "v1",
"created": "Mon, 14 Jun 2021 20:01:24 GMT"
},
{
"version": "v2",
"created": "Mon, 13 Mar 2023 21:34:57 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Chen",
"Changan",
""
],
[
"Sun",
"Wei",
""
],
[
"Harwath",
"David",
""
],
[
"Grauman",
"Kristen",
""
]
] | TITLE: Learning Audio-Visual Dereverberation
ABSTRACT: Reverberation not only degrades the quality of speech for human perception,
but also severely impacts the accuracy of automatic speech recognition. Prior
work attempts to remove reverberation based on the audio modality only. Our
idea is to learn to dereverberate speech from audio-visual observations. The
visual environment surrounding a human speaker reveals important cues about the
room geometry, materials, and speaker location, all of which influence the
precise reverberation effects. We introduce Visually-Informed Dereverberation
of Audio (VIDA), an end-to-end approach that learns to remove reverberation
based on both the observed monaural sound and visual scene. In support of this
new task, we develop a large-scale dataset SoundSpaces-Speech that uses
realistic acoustic renderings of speech in real-world 3D scans of homes
offering a variety of room acoustics. Demonstrating our approach on both
simulated and real imagery for speech enhancement, speech recognition, and
speaker identification, we show it achieves state-of-the-art performance and
substantially improves over audio-only methods.
| new_dataset | 0.953188 |
2110.01303 | Jiahao Huo | Jiahao Huo, Terence L. van Zyl | Incremental Class Learning using Variational Autoencoders with
Similarity Learning | null | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Catastrophic forgetting in neural networks during incremental learning
remains a challenging problem. Previous research investigated catastrophic
forgetting in fully connected networks, with some earlier work exploring
activation functions and learning algorithms. Applications of neural networks
have been extended to include similarity learning. Understanding how similarity
learning loss functions would be affected by catastrophic forgetting is of
significant interest. Our research investigates catastrophic forgetting for
four well-known similarity-based loss functions during incremental class
learning. The loss functions are Angular, Contrastive, Center, and Triplet
loss. Our results show that the catastrophic forgetting rate differs across
loss functions on multiple datasets. The Angular loss was least affected,
followed by Contrastive, Triplet loss, and Center loss with good mining
techniques. We implemented three existing incremental learning techniques,
iCaRL, EWC, and EBLL. We further proposed a novel technique using Variational
Autoencoders (VAEs) to generate representation as exemplars passed through the
network's intermediate layers. Our method outperformed three existing
state-of-the-art techniques. We show that one does not require stored images
(exemplars) for incremental learning with similarity learning. The generated
representations from VAEs help preserve regions of the embedding space used by
prior knowledge so that new knowledge does not ``overwrite'' it.
| [
{
"version": "v1",
"created": "Mon, 4 Oct 2021 10:19:53 GMT"
},
{
"version": "v2",
"created": "Sat, 16 Oct 2021 12:04:56 GMT"
},
{
"version": "v3",
"created": "Tue, 14 Mar 2023 11:40:35 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Huo",
"Jiahao",
""
],
[
"van Zyl",
"Terence L.",
""
]
] | TITLE: Incremental Class Learning using Variational Autoencoders with
Similarity Learning
ABSTRACT: Catastrophic forgetting in neural networks during incremental learning
remains a challenging problem. Previous research investigated catastrophic
forgetting in fully connected networks, with some earlier work exploring
activation functions and learning algorithms. Applications of neural networks
have been extended to include similarity learning. Understanding how similarity
learning loss functions would be affected by catastrophic forgetting is of
significant interest. Our research investigates catastrophic forgetting for
four well-known similarity-based loss functions during incremental class
learning. The loss functions are Angular, Contrastive, Center, and Triplet
loss. Our results show that the catastrophic forgetting rate differs across
loss functions on multiple datasets. The Angular loss was least affected,
followed by Contrastive, Triplet loss, and Center loss with good mining
techniques. We implemented three existing incremental learning techniques,
iCaRL, EWC, and EBLL. We further proposed a novel technique using Variational
Autoencoders (VAEs) to generate representation as exemplars passed through the
network's intermediate layers. Our method outperformed three existing
state-of-the-art techniques. We show that one does not require stored images
(exemplars) for incremental learning with similarity learning. The generated
representations from VAEs help preserve regions of the embedding space used by
prior knowledge so that new knowledge does not ``overwrite'' it.
| no_new_dataset | 0.945651 |
2110.03540 | Hamed Bonab | Sepehr Bakhshi, Pouya Ghahramanian, Hamed Bonab, and Fazli Can | A Broad Ensemble Learning System for Drifting Stream Classification | Submitted to IEEE Access | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In a data stream environment, classification models must handle concept drift
efficiently and effectively. Ensemble methods are widely used for this purpose;
however, the ones available in the literature either use a large data chunk to
update the model or learn the data one by one. In the former, the model may
miss the changes in the data distribution, and in the latter, the model may
suffer from inefficiency and instability. To address these issues, we introduce
a novel ensemble approach based on the Broad Learning System (BLS), where mini
chunks are used at each update. BLS is an effective lightweight neural
architecture recently developed for incremental learning. Although it is fast,
it requires huge data chunks for effective updates, and is unable to handle
dynamic changes observed in data streams. Our proposed approach named Broad
Ensemble Learning System (BELS) uses a novel updating method that significantly
improves best-in-class model accuracy. It employs an ensemble of output layers
to address the limitations of BLS and handle drifts. Our model tracks the
changes in the accuracy of the ensemble components and react to these changes.
We present the mathematical derivation of BELS, perform comprehensive
experiments with 20 datasets that demonstrate the adaptability of our model to
various drift types, and provide hyperparameter and ablation analysis of our
proposed model. Our experiments show that the proposed approach outperforms
nine state-of-the-art baselines and supplies an overall improvement of 13.28%
in terms of average prequential accuracy.
| [
{
"version": "v1",
"created": "Thu, 7 Oct 2021 15:01:33 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 12:38:35 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Bakhshi",
"Sepehr",
""
],
[
"Ghahramanian",
"Pouya",
""
],
[
"Bonab",
"Hamed",
""
],
[
"Can",
"Fazli",
""
]
] | TITLE: A Broad Ensemble Learning System for Drifting Stream Classification
ABSTRACT: In a data stream environment, classification models must handle concept drift
efficiently and effectively. Ensemble methods are widely used for this purpose;
however, the ones available in the literature either use a large data chunk to
update the model or learn the data one by one. In the former, the model may
miss the changes in the data distribution, and in the latter, the model may
suffer from inefficiency and instability. To address these issues, we introduce
a novel ensemble approach based on the Broad Learning System (BLS), where mini
chunks are used at each update. BLS is an effective lightweight neural
architecture recently developed for incremental learning. Although it is fast,
it requires huge data chunks for effective updates, and is unable to handle
dynamic changes observed in data streams. Our proposed approach named Broad
Ensemble Learning System (BELS) uses a novel updating method that significantly
improves best-in-class model accuracy. It employs an ensemble of output layers
to address the limitations of BLS and handle drifts. Our model tracks the
changes in the accuracy of the ensemble components and react to these changes.
We present the mathematical derivation of BELS, perform comprehensive
experiments with 20 datasets that demonstrate the adaptability of our model to
various drift types, and provide hyperparameter and ablation analysis of our
proposed model. Our experiments show that the proposed approach outperforms
nine state-of-the-art baselines and supplies an overall improvement of 13.28%
in terms of average prequential accuracy.
| no_new_dataset | 0.945901 |
2110.08895 | Sreyan Ghosh | Sreyan Ghosh and Sandesh V Katta and Ashish Seth and S. Umesh | DECAR: Deep Clustering for learning general-purpose Audio
Representations | null | null | 10.1109/JSTSP.2022.3202093 | null | cs.SD cs.CL eess.AS | http://creativecommons.org/licenses/by/4.0/ | We introduce DECAR, a self-supervised pre-training approach for learning
general-purpose audio representations. Our system is based on clustering: it
utilizes an offline clustering step to provide target labels that act as
pseudo-labels for solving a prediction task. We develop on top of recent
advances in self-supervised learning for computer vision and design a
lightweight, easy-to-use self-supervised pre-training scheme. We pre-train
DECAR embeddings on a balanced subset of the large-scale Audioset dataset and
transfer those representations to 9 downstream classification tasks, including
speech, music, animal sounds, and acoustic scenes. Furthermore, we conduct
ablation studies identifying key design choices and also make all our code and
pre-trained models publicly available.
| [
{
"version": "v1",
"created": "Sun, 17 Oct 2021 19:03:51 GMT"
},
{
"version": "v2",
"created": "Sat, 6 Nov 2021 19:19:20 GMT"
},
{
"version": "v3",
"created": "Mon, 4 Apr 2022 14:22:18 GMT"
},
{
"version": "v4",
"created": "Tue, 14 Mar 2023 14:29:58 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Ghosh",
"Sreyan",
""
],
[
"Katta",
"Sandesh V",
""
],
[
"Seth",
"Ashish",
""
],
[
"Umesh",
"S.",
""
]
] | TITLE: DECAR: Deep Clustering for learning general-purpose Audio
Representations
ABSTRACT: We introduce DECAR, a self-supervised pre-training approach for learning
general-purpose audio representations. Our system is based on clustering: it
utilizes an offline clustering step to provide target labels that act as
pseudo-labels for solving a prediction task. We develop on top of recent
advances in self-supervised learning for computer vision and design a
lightweight, easy-to-use self-supervised pre-training scheme. We pre-train
DECAR embeddings on a balanced subset of the large-scale Audioset dataset and
transfer those representations to 9 downstream classification tasks, including
speech, music, animal sounds, and acoustic scenes. Furthermore, we conduct
ablation studies identifying key design choices and also make all our code and
pre-trained models publicly available.
| no_new_dataset | 0.947137 |
2201.04831 | Juhua Liu | Qihuang Zhong, Liang Ding, Juhua Liu, Bo Du, Hua Jin, Dacheng Tao | Knowledge Graph Augmented Network Towards Multiview Representation
Learning for Aspect-based Sentiment Analysis | Accepted by IEEE TKDE 2023 | null | 10.1109/TKDE.2023.3250499 | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Aspect-based sentiment analysis (ABSA) is a fine-grained task of sentiment
analysis. To better comprehend long complicated sentences and obtain accurate
aspect-specific information, linguistic and commonsense knowledge are generally
required in this task. However, most current methods employ complicated and
inefficient approaches to incorporate external knowledge, e.g., directly
searching the graph nodes. Additionally, the complementarity between external
knowledge and linguistic information has not been thoroughly studied. To this
end, we propose a knowledge graph augmented network KGAN, which aims to
effectively incorporate external knowledge with explicitly syntactic and
contextual information. In particular, KGAN captures the sentiment feature
representations from multiple different perspectives, i.e., context-, syntax-
and knowledge-based. First, KGAN learns the contextual and syntactic
representations in parallel to fully extract the semantic features. Then, KGAN
integrates the knowledge graphs into the embedding space, based on which the
aspect-specific knowledge representations are further obtained via an attention
mechanism. Last, we propose a hierarchical fusion module to complement these
multi-view representations in a local-to-global manner. Extensive experiments
on five popular ABSA benchmarks demonstrate the effectiveness and robustness of
our KGAN. Notably, with the help of the pretrained model of RoBERTa, KGAN
achieves a new record of state-of-the-art performance among all datasets.
| [
{
"version": "v1",
"created": "Thu, 13 Jan 2022 08:25:53 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 01:53:07 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Zhong",
"Qihuang",
""
],
[
"Ding",
"Liang",
""
],
[
"Liu",
"Juhua",
""
],
[
"Du",
"Bo",
""
],
[
"Jin",
"Hua",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Knowledge Graph Augmented Network Towards Multiview Representation
Learning for Aspect-based Sentiment Analysis
ABSTRACT: Aspect-based sentiment analysis (ABSA) is a fine-grained task of sentiment
analysis. To better comprehend long complicated sentences and obtain accurate
aspect-specific information, linguistic and commonsense knowledge are generally
required in this task. However, most current methods employ complicated and
inefficient approaches to incorporate external knowledge, e.g., directly
searching the graph nodes. Additionally, the complementarity between external
knowledge and linguistic information has not been thoroughly studied. To this
end, we propose a knowledge graph augmented network KGAN, which aims to
effectively incorporate external knowledge with explicitly syntactic and
contextual information. In particular, KGAN captures the sentiment feature
representations from multiple different perspectives, i.e., context-, syntax-
and knowledge-based. First, KGAN learns the contextual and syntactic
representations in parallel to fully extract the semantic features. Then, KGAN
integrates the knowledge graphs into the embedding space, based on which the
aspect-specific knowledge representations are further obtained via an attention
mechanism. Last, we propose a hierarchical fusion module to complement these
multi-view representations in a local-to-global manner. Extensive experiments
on five popular ABSA benchmarks demonstrate the effectiveness and robustness of
our KGAN. Notably, with the help of the pretrained model of RoBERTa, KGAN
achieves a new record of state-of-the-art performance among all datasets.
| no_new_dataset | 0.945096 |
2201.10037 | Steve Huntsman | Steve Huntsman | Diversity Enhancement via Magnitude | null | null | 10.1007/978-3-031-27250-9_27 | null | cs.NE math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Promoting and maintaining diversity of candidate solutions is a key
requirement of evolutionary algorithms in general and multi-objective
evolutionary algorithms in particular. In this paper, we use the recently
developed theory of magnitude to construct a gradient flow and similar notions
that systematically manipulate finite subsets of Euclidean space to enhance
their diversity, and apply the ideas in service of multi-objective evolutionary
algorithms. We demonstrate diversity enhancement on benchmark problems using
leading algorithms, and discuss extensions of the framework.
| [
{
"version": "v1",
"created": "Tue, 25 Jan 2022 01:38:53 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Huntsman",
"Steve",
""
]
] | TITLE: Diversity Enhancement via Magnitude
ABSTRACT: Promoting and maintaining diversity of candidate solutions is a key
requirement of evolutionary algorithms in general and multi-objective
evolutionary algorithms in particular. In this paper, we use the recently
developed theory of magnitude to construct a gradient flow and similar notions
that systematically manipulate finite subsets of Euclidean space to enhance
their diversity, and apply the ideas in service of multi-objective evolutionary
algorithms. We demonstrate diversity enhancement on benchmark problems using
leading algorithms, and discuss extensions of the framework.
| no_new_dataset | 0.95096 |
2202.02113 | Ningyu Zhang | Xin Xie, Ningyu Zhang, Zhoubo Li, Shumin Deng, Hui Chen, Feiyu Xiong,
Mosha Chen, Huajun Chen | From Discrimination to Generation: Knowledge Graph Completion with
Generative Transformer | Accepted by WWW 2022 Poster | null | 10.1145/3487553.3524238 | null | cs.CL cs.AI cs.DB cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Knowledge graph completion aims to address the problem of extending a KG with
missing triples. In this paper, we provide an approach GenKGC, which converts
knowledge graph completion to sequence-to-sequence generation task with the
pre-trained language model. We further introduce relation-guided demonstration
and entity-aware hierarchical decoding for better representation learning and
fast inference. Experimental results on three datasets show that our approach
can obtain better or comparable performance than baselines and achieve faster
inference speed compared with previous methods with pre-trained language
models. We also release a new large-scale Chinese knowledge graph dataset
AliopenKG500 for research purpose. Code and datasets are available in
https://github.com/zjunlp/PromptKG/tree/main/GenKGC.
| [
{
"version": "v1",
"created": "Fri, 4 Feb 2022 12:52:32 GMT"
},
{
"version": "v2",
"created": "Tue, 8 Feb 2022 14:07:30 GMT"
},
{
"version": "v3",
"created": "Mon, 7 Mar 2022 09:17:45 GMT"
},
{
"version": "v4",
"created": "Thu, 17 Mar 2022 11:32:50 GMT"
},
{
"version": "v5",
"created": "Fri, 25 Mar 2022 05:18:02 GMT"
},
{
"version": "v6",
"created": "Tue, 29 Mar 2022 07:17:43 GMT"
},
{
"version": "v7",
"created": "Tue, 14 Mar 2023 10:33:15 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Xie",
"Xin",
""
],
[
"Zhang",
"Ningyu",
""
],
[
"Li",
"Zhoubo",
""
],
[
"Deng",
"Shumin",
""
],
[
"Chen",
"Hui",
""
],
[
"Xiong",
"Feiyu",
""
],
[
"Chen",
"Mosha",
""
],
[
"Chen",
"Huajun",
""
]
] | TITLE: From Discrimination to Generation: Knowledge Graph Completion with
Generative Transformer
ABSTRACT: Knowledge graph completion aims to address the problem of extending a KG with
missing triples. In this paper, we provide an approach GenKGC, which converts
knowledge graph completion to sequence-to-sequence generation task with the
pre-trained language model. We further introduce relation-guided demonstration
and entity-aware hierarchical decoding for better representation learning and
fast inference. Experimental results on three datasets show that our approach
can obtain better or comparable performance than baselines and achieve faster
inference speed compared with previous methods with pre-trained language
models. We also release a new large-scale Chinese knowledge graph dataset
AliopenKG500 for research purpose. Code and datasets are available in
https://github.com/zjunlp/PromptKG/tree/main/GenKGC.
| new_dataset | 0.957198 |
2202.03008 | Gabriel Turinici | Gabriel Turinici | Algorithms that get old : the case of generative deep neural networks | null | Machine Learning, Optimization, and Data Science. LOD 2022.
Lecture Notes in Computer Science, vol 13811. Springer, Cham | 10.1007/978-3-031-25891-6_14 | null | stat.ML cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Generative deep neural networks used in machine learning, like the
Variational Auto-Encoders (VAE), and Generative Adversarial Networks (GANs)
produce new objects each time when asked to do so with the constraint that the
new objects remain similar to some list of examples given as input. However,
this behavior is unlike that of human artists that change their style as time
goes by and seldom return to the style of the initial creations.
We investigate a situation where VAEs are used to sample from a probability
measure described by some empirical dataset.
Based on recent works on Radon-Sobolev statistical distances, we propose a
numerical paradigm, to be used in conjunction with a generative algorithm, that
satisfies the two following requirements: the objects created do not repeat and
evolve to fill the entire target probability distribution.
| [
{
"version": "v1",
"created": "Mon, 7 Feb 2022 08:55:37 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jun 2022 09:53:42 GMT"
},
{
"version": "v3",
"created": "Thu, 29 Sep 2022 12:17:47 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Turinici",
"Gabriel",
""
]
] | TITLE: Algorithms that get old : the case of generative deep neural networks
ABSTRACT: Generative deep neural networks used in machine learning, like the
Variational Auto-Encoders (VAE), and Generative Adversarial Networks (GANs)
produce new objects each time when asked to do so with the constraint that the
new objects remain similar to some list of examples given as input. However,
this behavior is unlike that of human artists that change their style as time
goes by and seldom return to the style of the initial creations.
We investigate a situation where VAEs are used to sample from a probability
measure described by some empirical dataset.
Based on recent works on Radon-Sobolev statistical distances, we propose a
numerical paradigm, to be used in conjunction with a generative algorithm, that
satisfies the two following requirements: the objects created do not repeat and
evolve to fill the entire target probability distribution.
| no_new_dataset | 0.947866 |
2205.13158 | Yuan Hu | Yuan Hu, Lei Chen, Zhibin Wang, Hao Li | SwinVRNN: A Data-Driven Ensemble Forecasting Model via Learned
Distribution Perturbation | null | null | 10.1029/2022MS003211 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data-driven approaches for medium-range weather forecasting are recently
shown extraordinarily promising for ensemble forecasting for their fast
inference speed compared to traditional numerical weather prediction (NWP)
models, but their forecast accuracy can hardly match the state-of-the-art
operational ECMWF Integrated Forecasting System (IFS) model. Previous
data-driven attempts achieve ensemble forecast using some simple perturbation
methods, like initial condition perturbation and Monte Carlo dropout. However,
they mostly suffer unsatisfactory ensemble performance, which is arguably
attributed to the sub-optimal ways of applying perturbation. We propose a Swin
Transformer-based Variational Recurrent Neural Network (SwinVRNN), which is a
stochastic weather forecasting model combining a SwinRNN predictor with a
perturbation module. SwinRNN is designed as a Swin Transformer-based recurrent
neural network, which predicts future states deterministically. Furthermore, to
model the stochasticity in prediction, we design a perturbation module
following the Variational Auto-Encoder paradigm to learn multivariate Gaussian
distributions of a time-variant stochastic latent variable from data. Ensemble
forecasting can be easily achieved by perturbing the model features leveraging
noise sampled from the learned distribution. We also compare four categories of
perturbation methods for ensemble forecasting, i.e. fixed distribution
perturbation, learned distribution perturbation, MC dropout, and multi model
ensemble. Comparisons on WeatherBench dataset show the learned distribution
perturbation method using our SwinVRNN model achieves superior forecast
accuracy and reasonable ensemble spread due to joint optimization of the two
targets. More notably, SwinVRNN surpasses operational IFS on surface variables
of 2-m temperature and 6-hourly total precipitation at all lead times up to
five days.
| [
{
"version": "v1",
"created": "Thu, 26 May 2022 05:11:58 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Hu",
"Yuan",
""
],
[
"Chen",
"Lei",
""
],
[
"Wang",
"Zhibin",
""
],
[
"Li",
"Hao",
""
]
] | TITLE: SwinVRNN: A Data-Driven Ensemble Forecasting Model via Learned
Distribution Perturbation
ABSTRACT: Data-driven approaches for medium-range weather forecasting are recently
shown extraordinarily promising for ensemble forecasting for their fast
inference speed compared to traditional numerical weather prediction (NWP)
models, but their forecast accuracy can hardly match the state-of-the-art
operational ECMWF Integrated Forecasting System (IFS) model. Previous
data-driven attempts achieve ensemble forecast using some simple perturbation
methods, like initial condition perturbation and Monte Carlo dropout. However,
they mostly suffer unsatisfactory ensemble performance, which is arguably
attributed to the sub-optimal ways of applying perturbation. We propose a Swin
Transformer-based Variational Recurrent Neural Network (SwinVRNN), which is a
stochastic weather forecasting model combining a SwinRNN predictor with a
perturbation module. SwinRNN is designed as a Swin Transformer-based recurrent
neural network, which predicts future states deterministically. Furthermore, to
model the stochasticity in prediction, we design a perturbation module
following the Variational Auto-Encoder paradigm to learn multivariate Gaussian
distributions of a time-variant stochastic latent variable from data. Ensemble
forecasting can be easily achieved by perturbing the model features leveraging
noise sampled from the learned distribution. We also compare four categories of
perturbation methods for ensemble forecasting, i.e. fixed distribution
perturbation, learned distribution perturbation, MC dropout, and multi model
ensemble. Comparisons on WeatherBench dataset show the learned distribution
perturbation method using our SwinVRNN model achieves superior forecast
accuracy and reasonable ensemble spread due to joint optimization of the two
targets. More notably, SwinVRNN surpasses operational IFS on surface variables
of 2-m temperature and 6-hourly total precipitation at all lead times up to
five days.
| no_new_dataset | 0.948728 |
2205.15567 | Youngsik Yoon | Youngsik Yoon, Jinhwan Nam, Hyojeong Yun, Jaeho Lee, Dongwoo Kim,
Jungseul Ok | Few-Shot Unlearning by Model Inversion | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider a practical scenario of machine unlearning to erase a target
dataset, which causes unexpected behavior from the trained model. The target
dataset is often assumed to be fully identifiable in a standard unlearning
scenario. Such a flawless identification, however, is almost impossible if the
training dataset is inaccessible at the time of unlearning. Unlike previous
approaches requiring a complete set of targets, we consider few-shot unlearning
scenario when only a few samples of target data are available. To this end, we
formulate the few-shot unlearning problem specifying intentions behind the
unlearning request (e.g., purely unlearning, mislabel correction, privacy
protection), and we devise a straightforward framework that (i) retrieves a
proxy of the training data via model inversion fully exploiting information
available in the context of unlearning; (ii) adjusts the proxy according to the
unlearning intention; and (iii) updates the model with the adjusted proxy. We
demonstrate that our method using only a subset of target data can outperform
the state-of-the-art unlearning methods even with a complete indication of
target data.
| [
{
"version": "v1",
"created": "Tue, 31 May 2022 06:57:56 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 10:37:52 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Yoon",
"Youngsik",
""
],
[
"Nam",
"Jinhwan",
""
],
[
"Yun",
"Hyojeong",
""
],
[
"Lee",
"Jaeho",
""
],
[
"Kim",
"Dongwoo",
""
],
[
"Ok",
"Jungseul",
""
]
] | TITLE: Few-Shot Unlearning by Model Inversion
ABSTRACT: We consider a practical scenario of machine unlearning to erase a target
dataset, which causes unexpected behavior from the trained model. The target
dataset is often assumed to be fully identifiable in a standard unlearning
scenario. Such a flawless identification, however, is almost impossible if the
training dataset is inaccessible at the time of unlearning. Unlike previous
approaches requiring a complete set of targets, we consider few-shot unlearning
scenario when only a few samples of target data are available. To this end, we
formulate the few-shot unlearning problem specifying intentions behind the
unlearning request (e.g., purely unlearning, mislabel correction, privacy
protection), and we devise a straightforward framework that (i) retrieves a
proxy of the training data via model inversion fully exploiting information
available in the context of unlearning; (ii) adjusts the proxy according to the
unlearning intention; and (iii) updates the model with the adjusted proxy. We
demonstrate that our method using only a subset of target data can outperform
the state-of-the-art unlearning methods even with a complete indication of
target data.
| no_new_dataset | 0.945096 |
2206.12458 | Fagner Cunha | Fagner Cunha, Eulanda M. dos Santos, Juan G. Colonna | Bag of Tricks for Long-Tail Visual Recognition of Animal Species in
Camera-Trap Images | null | null | 10.1016/j.ecoinf.2023.102060 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Camera traps are a method for monitoring wildlife and they collect a large
number of pictures. The number of images collected of each species usually
follows a long-tail distribution, i.e., a few classes have a large number of
instances, while a lot of species have just a small percentage. Although in
most cases these rare species are the ones of interest to ecologists, they are
often neglected when using deep-learning models because these models require a
large number of images for the training. In this work, a simple and effective
framework called Square-Root Sampling Branch (SSB) is proposed, which combines
two classification branches that are trained using square-root sampling and
instance sampling to improve long-tail visual recognition, and this is compared
to state-of-the-art methods for handling this task: square-root sampling,
class-balanced focal loss, and balanced group softmax. To achieve a more
general conclusion, the methods for handling long-tail visual recognition were
systematically evaluated in four families of computer vision models (ResNet,
MobileNetV3, EfficientNetV2, and Swin Transformer) and four camera-trap
datasets with different characteristics. Initially, a robust baseline with the
most recent training tricks was prepared and, then, the methods for improving
long-tail recognition were applied. Our experiments show that square-root
sampling was the method that most improved the performance for minority classes
by around 15%; however, this was at the cost of reducing the majority classes'
accuracy by at least 3%. Our proposed framework (SSB) demonstrated itself to be
competitive with the other methods and achieved the best or the second-best
results for most of the cases for the tail classes; but, unlike the square-root
sampling, the loss in the performance of the head classes was minimal, thus
achieving the best trade-off among all the evaluated methods.
| [
{
"version": "v1",
"created": "Fri, 24 Jun 2022 18:30:26 GMT"
},
{
"version": "v2",
"created": "Mon, 31 Oct 2022 15:57:53 GMT"
},
{
"version": "v3",
"created": "Mon, 6 Mar 2023 21:26:26 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Cunha",
"Fagner",
""
],
[
"Santos",
"Eulanda M. dos",
""
],
[
"Colonna",
"Juan G.",
""
]
] | TITLE: Bag of Tricks for Long-Tail Visual Recognition of Animal Species in
Camera-Trap Images
ABSTRACT: Camera traps are a method for monitoring wildlife and they collect a large
number of pictures. The number of images collected of each species usually
follows a long-tail distribution, i.e., a few classes have a large number of
instances, while a lot of species have just a small percentage. Although in
most cases these rare species are the ones of interest to ecologists, they are
often neglected when using deep-learning models because these models require a
large number of images for the training. In this work, a simple and effective
framework called Square-Root Sampling Branch (SSB) is proposed, which combines
two classification branches that are trained using square-root sampling and
instance sampling to improve long-tail visual recognition, and this is compared
to state-of-the-art methods for handling this task: square-root sampling,
class-balanced focal loss, and balanced group softmax. To achieve a more
general conclusion, the methods for handling long-tail visual recognition were
systematically evaluated in four families of computer vision models (ResNet,
MobileNetV3, EfficientNetV2, and Swin Transformer) and four camera-trap
datasets with different characteristics. Initially, a robust baseline with the
most recent training tricks was prepared and, then, the methods for improving
long-tail recognition were applied. Our experiments show that square-root
sampling was the method that most improved the performance for minority classes
by around 15%; however, this was at the cost of reducing the majority classes'
accuracy by at least 3%. Our proposed framework (SSB) demonstrated itself to be
competitive with the other methods and achieved the best or the second-best
results for most of the cases for the tail classes; but, unlike the square-root
sampling, the loss in the performance of the head classes was minimal, thus
achieving the best trade-off among all the evaluated methods.
| no_new_dataset | 0.95253 |
2207.00576 | Justin Kottinger | Justin Kottinger, Shaull Almagor, Morteza Lahijanian | Conflict-based Search for Multi-Robot Motion Planning with Kinodynamic
Constraints | 7 pages, 6 figures, and 2 algorithms. To be presented at
International Conference on Intelligent Robots and Systems (IROS) in October
2022 | 2022 IEEE/RSJ International Conference on Intelligent Robots and
Systems (IROS) | 10.1109/IROS47612.2022.9982018 | null | cs.RO cs.MA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-robot motion planning (MRMP) is the fundamental problem of finding
non-colliding trajectories for multiple robots acting in an environment, under
kinodynamic constraints. Due to its complexity, existing algorithms either
utilize simplifying assumptions or are incomplete. This work introduces
kinodynamic conflict-based search (K-CBS), a decentralized (decoupled) MRMP
algorithm that is general, scalable, and probabilistically complete. The
algorithm takes inspiration from successful solutions to the discrete analogue
of MRMP over finite graphs, known as multi-agent path finding (MAPF).
Specifically, we adapt ideas from conflict-based search (CBS) - a popular
decentralized MAPF algorithm - to the MRMP setting. The novelty in this
adaptation is that we work directly in the continuous domain, without the need
for discretization. In particular, the kinodynamic constraints are treated
natively. K-CBS plans for each robot individually using a low-level planner and
and grows a conflict tree to resolve collisions between robots by defining
constraints for individual robots. The low-level planner can be any
sampling-based, tree-search algorithm for kinodynamic robots, thus lifting
existing planners for single robots to the multi-robot settings. We show that
K-CBS inherits the (probabilistic) completeness of the low-level planner. We
illustrate the generality and performance of K-CBS in several case studies and
benchmarks.
| [
{
"version": "v1",
"created": "Fri, 1 Jul 2022 17:54:10 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Kottinger",
"Justin",
""
],
[
"Almagor",
"Shaull",
""
],
[
"Lahijanian",
"Morteza",
""
]
] | TITLE: Conflict-based Search for Multi-Robot Motion Planning with Kinodynamic
Constraints
ABSTRACT: Multi-robot motion planning (MRMP) is the fundamental problem of finding
non-colliding trajectories for multiple robots acting in an environment, under
kinodynamic constraints. Due to its complexity, existing algorithms either
utilize simplifying assumptions or are incomplete. This work introduces
kinodynamic conflict-based search (K-CBS), a decentralized (decoupled) MRMP
algorithm that is general, scalable, and probabilistically complete. The
algorithm takes inspiration from successful solutions to the discrete analogue
of MRMP over finite graphs, known as multi-agent path finding (MAPF).
Specifically, we adapt ideas from conflict-based search (CBS) - a popular
decentralized MAPF algorithm - to the MRMP setting. The novelty in this
adaptation is that we work directly in the continuous domain, without the need
for discretization. In particular, the kinodynamic constraints are treated
natively. K-CBS plans for each robot individually using a low-level planner and
and grows a conflict tree to resolve collisions between robots by defining
constraints for individual robots. The low-level planner can be any
sampling-based, tree-search algorithm for kinodynamic robots, thus lifting
existing planners for single robots to the multi-robot settings. We show that
K-CBS inherits the (probabilistic) completeness of the low-level planner. We
illustrate the generality and performance of K-CBS in several case studies and
benchmarks.
| no_new_dataset | 0.947866 |
2207.07506 | Guoxuan Xia | Guoxuan Xia and Christos-Savvas Bouganis | Augmenting Softmax Information for Selective Classification with
Out-of-Distribution Data | ACCV 2022 (Best Paper Award)
https://openaccess.thecvf.com/content/ACCV2022/html/Xia_Augmenting_Softmax_Information_for_Selective_Classification_with_Out-of-Distribution_Data_ACCV_2022_paper.html | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Detecting out-of-distribution (OOD) data is a task that is receiving an
increasing amount of research attention in the domain of deep learning for
computer vision. However, the performance of detection methods is generally
evaluated on the task in isolation, rather than also considering potential
downstream tasks in tandem. In this work, we examine selective classification
in the presence of OOD data (SCOD). That is to say, the motivation for
detecting OOD samples is to reject them so their impact on the quality of
predictions is reduced. We show under this task specification, that existing
post-hoc methods perform quite differently compared to when evaluated only on
OOD detection. This is because it is no longer an issue to conflate
in-distribution (ID) data with OOD data if the ID data is going to be
misclassified. However, the conflation within ID data of correct and incorrect
predictions becomes undesirable. We also propose a novel method for SCOD,
Softmax Information Retaining Combination (SIRC), that augments softmax-based
confidence scores with feature-agnostic information such that their ability to
identify OOD samples is improved without sacrificing separation between correct
and incorrect ID predictions. Experiments on a wide variety of ImageNet-scale
datasets and convolutional neural network architectures show that SIRC is able
to consistently match or outperform the baseline for SCOD, whilst existing OOD
detection methods fail to do so.
| [
{
"version": "v1",
"created": "Fri, 15 Jul 2022 14:39:57 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Oct 2022 09:57:08 GMT"
},
{
"version": "v3",
"created": "Tue, 14 Mar 2023 16:11:35 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Xia",
"Guoxuan",
""
],
[
"Bouganis",
"Christos-Savvas",
""
]
] | TITLE: Augmenting Softmax Information for Selective Classification with
Out-of-Distribution Data
ABSTRACT: Detecting out-of-distribution (OOD) data is a task that is receiving an
increasing amount of research attention in the domain of deep learning for
computer vision. However, the performance of detection methods is generally
evaluated on the task in isolation, rather than also considering potential
downstream tasks in tandem. In this work, we examine selective classification
in the presence of OOD data (SCOD). That is to say, the motivation for
detecting OOD samples is to reject them so their impact on the quality of
predictions is reduced. We show under this task specification, that existing
post-hoc methods perform quite differently compared to when evaluated only on
OOD detection. This is because it is no longer an issue to conflate
in-distribution (ID) data with OOD data if the ID data is going to be
misclassified. However, the conflation within ID data of correct and incorrect
predictions becomes undesirable. We also propose a novel method for SCOD,
Softmax Information Retaining Combination (SIRC), that augments softmax-based
confidence scores with feature-agnostic information such that their ability to
identify OOD samples is improved without sacrificing separation between correct
and incorrect ID predictions. Experiments on a wide variety of ImageNet-scale
datasets and convolutional neural network architectures show that SIRC is able
to consistently match or outperform the baseline for SCOD, whilst existing OOD
detection methods fail to do so.
| no_new_dataset | 0.942929 |
2207.10018 | Guanchu Wang | Guanchu Wang and Mengnan Du and Ninghao Liu and Na Zou and Xia Hu | Mitigating Algorithmic Bias with Limited Annotations | null | null | null | null | cs.LG cs.AI cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing work on fairness modeling commonly assumes that sensitive attributes
for all instances are fully available, which may not be true in many real-world
applications due to the high cost of acquiring sensitive information. When
sensitive attributes are not disclosed or available, it is needed to manually
annotate a small part of the training data to mitigate bias. However, the
skewed distribution across different sensitive groups preserves the skewness of
the original dataset in the annotated subset, which leads to non-optimal bias
mitigation. To tackle this challenge, we propose Active Penalization Of
Discrimination (APOD), an interactive framework to guide the limited
annotations towards maximally eliminating the effect of algorithmic bias. The
proposed APOD integrates discrimination penalization with active instance
selection to efficiently utilize the limited annotation budget, and it is
theoretically proved to be capable of bounding the algorithmic bias. According
to the evaluation on five benchmark datasets, APOD outperforms the
state-of-the-arts baseline methods under the limited annotation budget, and
shows comparable performance to fully annotated bias mitigation, which
demonstrates that APOD could benefit real-world applications when sensitive
information is limited.
| [
{
"version": "v1",
"created": "Wed, 20 Jul 2022 16:31:19 GMT"
},
{
"version": "v2",
"created": "Tue, 7 Feb 2023 17:03:44 GMT"
},
{
"version": "v3",
"created": "Mon, 13 Mar 2023 22:18:39 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Wang",
"Guanchu",
""
],
[
"Du",
"Mengnan",
""
],
[
"Liu",
"Ninghao",
""
],
[
"Zou",
"Na",
""
],
[
"Hu",
"Xia",
""
]
] | TITLE: Mitigating Algorithmic Bias with Limited Annotations
ABSTRACT: Existing work on fairness modeling commonly assumes that sensitive attributes
for all instances are fully available, which may not be true in many real-world
applications due to the high cost of acquiring sensitive information. When
sensitive attributes are not disclosed or available, it is needed to manually
annotate a small part of the training data to mitigate bias. However, the
skewed distribution across different sensitive groups preserves the skewness of
the original dataset in the annotated subset, which leads to non-optimal bias
mitigation. To tackle this challenge, we propose Active Penalization Of
Discrimination (APOD), an interactive framework to guide the limited
annotations towards maximally eliminating the effect of algorithmic bias. The
proposed APOD integrates discrimination penalization with active instance
selection to efficiently utilize the limited annotation budget, and it is
theoretically proved to be capable of bounding the algorithmic bias. According
to the evaluation on five benchmark datasets, APOD outperforms the
state-of-the-arts baseline methods under the limited annotation budget, and
shows comparable performance to fully annotated bias mitigation, which
demonstrates that APOD could benefit real-world applications when sensitive
information is limited.
| no_new_dataset | 0.945601 |
2207.11659 | Haibo Shen | Haibo Shen, Yihao Luo, Xiang Cao, Liangqi Zhang, Juyu Xiao, Tianjiang
Wang | Training Robust Spiking Neural Networks on Neuromorphic Data with
Spatiotemporal Fragments | Accepted by ICASSP 2023 | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Neuromorphic vision sensors (event cameras) are inherently suitable for
spiking neural networks (SNNs) and provide novel neuromorphic vision data for
this biomimetic model. Due to the spatiotemporal characteristics, novel data
augmentations are required to process the unconventional visual signals of
these cameras. In this paper, we propose a novel Event SpatioTemporal Fragments
(ESTF) augmentation method. It preserves the continuity of neuromorphic data by
drifting or inverting fragments of the spatiotemporal event stream to simulate
the disturbance of brightness variations, leading to more robust spiking neural
networks. Extensive experiments are performed on prevailing neuromorphic
datasets. It turns out that ESTF provides substantial improvements over pure
geometric transformations and outperforms other event data augmentation
methods. It is worth noting that the SNNs with ESTF achieve the
state-of-the-art accuracy of 83.9\% on the CIFAR10-DVS dataset.
| [
{
"version": "v1",
"created": "Sun, 24 Jul 2022 04:23:56 GMT"
},
{
"version": "v2",
"created": "Tue, 13 Sep 2022 08:02:21 GMT"
},
{
"version": "v3",
"created": "Tue, 14 Mar 2023 02:52:54 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Shen",
"Haibo",
""
],
[
"Luo",
"Yihao",
""
],
[
"Cao",
"Xiang",
""
],
[
"Zhang",
"Liangqi",
""
],
[
"Xiao",
"Juyu",
""
],
[
"Wang",
"Tianjiang",
""
]
] | TITLE: Training Robust Spiking Neural Networks on Neuromorphic Data with
Spatiotemporal Fragments
ABSTRACT: Neuromorphic vision sensors (event cameras) are inherently suitable for
spiking neural networks (SNNs) and provide novel neuromorphic vision data for
this biomimetic model. Due to the spatiotemporal characteristics, novel data
augmentations are required to process the unconventional visual signals of
these cameras. In this paper, we propose a novel Event SpatioTemporal Fragments
(ESTF) augmentation method. It preserves the continuity of neuromorphic data by
drifting or inverting fragments of the spatiotemporal event stream to simulate
the disturbance of brightness variations, leading to more robust spiking neural
networks. Extensive experiments are performed on prevailing neuromorphic
datasets. It turns out that ESTF provides substantial improvements over pure
geometric transformations and outperforms other event data augmentation
methods. It is worth noting that the SNNs with ESTF achieve the
state-of-the-art accuracy of 83.9\% on the CIFAR10-DVS dataset.
| no_new_dataset | 0.951774 |
2207.11670 | Haibo Shen | Haibo Shen, Yihao Luo, Xiang Cao, Liangqi Zhang, Juyu Xiao, Tianjiang
Wang | Training Stronger Spiking Neural Networks with Biomimetic Adaptive
Internal Association Neurons | Accepted by ICASSP 2023 | null | null | null | cs.NE cs.CV cs.LG q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | As the third generation of neural networks, spiking neural networks (SNNs)
are dedicated to exploring more insightful neural mechanisms to achieve
near-biological intelligence. Intuitively, biomimetic mechanisms are crucial to
understanding and improving SNNs. For example, the associative long-term
potentiation (ALTP) phenomenon suggests that in addition to learning mechanisms
between neurons, there are associative effects within neurons. However, most
existing methods only focus on the former and lack exploration of the internal
association effects. In this paper, we propose a novel Adaptive Internal
Association~(AIA) neuron model to establish previously ignored influences
within neurons. Consistent with the ALTP phenomenon, the AIA neuron model is
adaptive to input stimuli, and internal associative learning occurs only when
both dendrites are stimulated at the same time. In addition, we employ weighted
weights to measure internal associations and introduce intermediate caches to
reduce the volatility of associations. Extensive experiments on prevailing
neuromorphic datasets show that the proposed method can potentiate or depress
the firing of spikes more specifically, resulting in better performance with
fewer spikes. It is worth noting that without adding any parameters at
inference, the AIA model achieves state-of-the-art performance on
DVS-CIFAR10~(83.9\%) and N-CARS~(95.64\%) datasets.
| [
{
"version": "v1",
"created": "Sun, 24 Jul 2022 06:12:23 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 02:51:07 GMT"
}
] | 2023-03-15T00:00:00 | [
[
"Shen",
"Haibo",
""
],
[
"Luo",
"Yihao",
""
],
[
"Cao",
"Xiang",
""
],
[
"Zhang",
"Liangqi",
""
],
[
"Xiao",
"Juyu",
""
],
[
"Wang",
"Tianjiang",
""
]
] | TITLE: Training Stronger Spiking Neural Networks with Biomimetic Adaptive
Internal Association Neurons
ABSTRACT: As the third generation of neural networks, spiking neural networks (SNNs)
are dedicated to exploring more insightful neural mechanisms to achieve
near-biological intelligence. Intuitively, biomimetic mechanisms are crucial to
understanding and improving SNNs. For example, the associative long-term
potentiation (ALTP) phenomenon suggests that in addition to learning mechanisms
between neurons, there are associative effects within neurons. However, most
existing methods only focus on the former and lack exploration of the internal
association effects. In this paper, we propose a novel Adaptive Internal
Association~(AIA) neuron model to establish previously ignored influences
within neurons. Consistent with the ALTP phenomenon, the AIA neuron model is
adaptive to input stimuli, and internal associative learning occurs only when
both dendrites are stimulated at the same time. In addition, we employ weighted
weights to measure internal associations and introduce intermediate caches to
reduce the volatility of associations. Extensive experiments on prevailing
neuromorphic datasets show that the proposed method can potentiate or depress
the firing of spikes more specifically, resulting in better performance with
fewer spikes. It is worth noting that without adding any parameters at
inference, the AIA model achieves state-of-the-art performance on
DVS-CIFAR10~(83.9\%) and N-CARS~(95.64\%) datasets.
| no_new_dataset | 0.948106 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.