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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2303.11715 | Shaohan Huang | Shaohan Huang, Yi Liu, Carol Fung, Jiaxing Qi, Hailong Yang, Zhongzhi
Luan | LogQA: Question Answering in Unstructured Logs | null | null | null | null | cs.NI | http://creativecommons.org/licenses/by/4.0/ | Modern systems produce a large volume of logs to record run-time status and
events. System operators use these raw logs to track a system in order to
obtain some useful information to diagnose system anomalies. One of the most
important problems in this area is to help operators find the answers to
log-based questions efficiently and user-friendly. In this work, we propose
LogQA, which aims at answering log-based questions in the form of natural
language based on large-scale unstructured log corpora. Our system presents the
answer to a question directly instead of returning a list of relevant snippets,
thus offering better user-friendliness and efficiency. LogQA represents the
first approach to solve question answering in lod domain. LogQA has two key
components: Log Retriever and Log Reader. Log Retriever aims at retrieving
relevant logs w.r.t. a given question, while Log Reader is responsible for
inferring the final answer. Given the lack of a public dataset for log questing
answering, we manually labelled a QA dataset of three open-source log corpus
and will make them publicly available. We evaluated our proposed model on these
datasets by comparing its performance with 6 other baseline methods. Our
experimental results demonstrate that LogQA has outperformed other baseline
methods.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 10:07:17 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Huang",
"Shaohan",
""
],
[
"Liu",
"Yi",
""
],
[
"Fung",
"Carol",
""
],
[
"Qi",
"Jiaxing",
""
],
[
"Yang",
"Hailong",
""
],
[
"Luan",
"Zhongzhi",
""
]
] | TITLE: LogQA: Question Answering in Unstructured Logs
ABSTRACT: Modern systems produce a large volume of logs to record run-time status and
events. System operators use these raw logs to track a system in order to
obtain some useful information to diagnose system anomalies. One of the most
important problems in this area is to help operators find the answers to
log-based questions efficiently and user-friendly. In this work, we propose
LogQA, which aims at answering log-based questions in the form of natural
language based on large-scale unstructured log corpora. Our system presents the
answer to a question directly instead of returning a list of relevant snippets,
thus offering better user-friendliness and efficiency. LogQA represents the
first approach to solve question answering in lod domain. LogQA has two key
components: Log Retriever and Log Reader. Log Retriever aims at retrieving
relevant logs w.r.t. a given question, while Log Reader is responsible for
inferring the final answer. Given the lack of a public dataset for log questing
answering, we manually labelled a QA dataset of three open-source log corpus
and will make them publicly available. We evaluated our proposed model on these
datasets by comparing its performance with 6 other baseline methods. Our
experimental results demonstrate that LogQA has outperformed other baseline
methods.
| new_dataset | 0.963541 |
2303.11730 | Kai Fong Ernest Chong | Jingyi Xu, Tushar Vaidya, Yufei Wu, Saket Chandra, Zhangsheng Lai, Kai
Fong Ernest Chong | Abstract Visual Reasoning: An Algebraic Approach for Solving Raven's
Progressive Matrices | Accepted at IEEE/CVF Conference on Computer Vision and Pattern
Recognition (CVPR) 2023. 30 pages, 7 figures (including supplementary
material). First three authors contributed equally. Code is available at:
https://github.com/Xu-Jingyi/AlgebraicMR | null | null | null | cs.CV cs.AI cs.SC math.AC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce algebraic machine reasoning, a new reasoning framework that is
well-suited for abstract reasoning. Effectively, algebraic machine reasoning
reduces the difficult process of novel problem-solving to routine algebraic
computation. The fundamental algebraic objects of interest are the ideals of
some suitably initialized polynomial ring. We shall explain how solving Raven's
Progressive Matrices (RPMs) can be realized as computational problems in
algebra, which combine various well-known algebraic subroutines that include:
Computing the Gr\"obner basis of an ideal, checking for ideal containment, etc.
Crucially, the additional algebraic structure satisfied by ideals allows for
more operations on ideals beyond set-theoretic operations.
Our algebraic machine reasoning framework is not only able to select the
correct answer from a given answer set, but also able to generate the correct
answer with only the question matrix given. Experiments on the I-RAVEN dataset
yield an overall $93.2\%$ accuracy, which significantly outperforms the current
state-of-the-art accuracy of $77.0\%$ and exceeds human performance at $84.4\%$
accuracy.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 10:34:39 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Xu",
"Jingyi",
""
],
[
"Vaidya",
"Tushar",
""
],
[
"Wu",
"Yufei",
""
],
[
"Chandra",
"Saket",
""
],
[
"Lai",
"Zhangsheng",
""
],
[
"Chong",
"Kai Fong Ernest",
""
]
] | TITLE: Abstract Visual Reasoning: An Algebraic Approach for Solving Raven's
Progressive Matrices
ABSTRACT: We introduce algebraic machine reasoning, a new reasoning framework that is
well-suited for abstract reasoning. Effectively, algebraic machine reasoning
reduces the difficult process of novel problem-solving to routine algebraic
computation. The fundamental algebraic objects of interest are the ideals of
some suitably initialized polynomial ring. We shall explain how solving Raven's
Progressive Matrices (RPMs) can be realized as computational problems in
algebra, which combine various well-known algebraic subroutines that include:
Computing the Gr\"obner basis of an ideal, checking for ideal containment, etc.
Crucially, the additional algebraic structure satisfied by ideals allows for
more operations on ideals beyond set-theoretic operations.
Our algebraic machine reasoning framework is not only able to select the
correct answer from a given answer set, but also able to generate the correct
answer with only the question matrix given. Experiments on the I-RAVEN dataset
yield an overall $93.2\%$ accuracy, which significantly outperforms the current
state-of-the-art accuracy of $77.0\%$ and exceeds human performance at $84.4\%$
accuracy.
| no_new_dataset | 0.945399 |
2303.11733 | Karthick Panner Selvam | Karthick Panner Selvam and Mats Brorsson | DIPPM: a Deep Learning Inference Performance Predictive Model using
Graph Neural Networks | null | null | null | null | cs.PF cs.AI cs.AR cs.DC cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Deep Learning (DL) has developed to become a corner-stone in many everyday
applications that we are now relying on. However, making sure that the DL model
uses the underlying hardware efficiently takes a lot of effort. Knowledge about
inference characteristics can help to find the right match so that enough
resources are given to the model, but not too much. We have developed a DL
Inference Performance Predictive Model (DIPPM) that predicts the inference
latency, energy, and memory usage of a given input DL model on the NVIDIA A100
GPU. We also devised an algorithm to suggest the appropriate A100
Multi-Instance GPU profile from the output of DIPPM. We developed a methodology
to convert DL models expressed in multiple frameworks to a generalized graph
structure that is used in DIPPM. It means DIPPM can parse input DL models from
various frameworks. Our DIPPM can be used not only helps to find suitable
hardware configurations but also helps to perform rapid design-space
exploration for the inference performance of a model. We constructed a graph
multi-regression dataset consisting of 10,508 different DL models to train and
evaluate the performance of DIPPM, and reached a resulting Mean Absolute
Percentage Error (MAPE) as low as 1.9%.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 10:43:41 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Selvam",
"Karthick Panner",
""
],
[
"Brorsson",
"Mats",
""
]
] | TITLE: DIPPM: a Deep Learning Inference Performance Predictive Model using
Graph Neural Networks
ABSTRACT: Deep Learning (DL) has developed to become a corner-stone in many everyday
applications that we are now relying on. However, making sure that the DL model
uses the underlying hardware efficiently takes a lot of effort. Knowledge about
inference characteristics can help to find the right match so that enough
resources are given to the model, but not too much. We have developed a DL
Inference Performance Predictive Model (DIPPM) that predicts the inference
latency, energy, and memory usage of a given input DL model on the NVIDIA A100
GPU. We also devised an algorithm to suggest the appropriate A100
Multi-Instance GPU profile from the output of DIPPM. We developed a methodology
to convert DL models expressed in multiple frameworks to a generalized graph
structure that is used in DIPPM. It means DIPPM can parse input DL models from
various frameworks. Our DIPPM can be used not only helps to find suitable
hardware configurations but also helps to perform rapid design-space
exploration for the inference performance of a model. We constructed a graph
multi-regression dataset consisting of 10,508 different DL models to train and
evaluate the performance of DIPPM, and reached a resulting Mean Absolute
Percentage Error (MAPE) as low as 1.9%.
| new_dataset | 0.964689 |
2303.11745 | Mohamed Amine Ferrag | Mohamed Amine Ferrag and Burak Kantarci and Lucas C. Cordeiro and
Merouane Debbah and Kim-Kwang Raymond Choo | Poisoning Attacks in Federated Edge Learning for Digital Twin 6G-enabled
IoTs: An Anticipatory Study | The paper is accepted and will be published in the IEEE ICC 2023
Conference Proceedings | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by/4.0/ | Federated edge learning can be essential in supporting privacy-preserving,
artificial intelligence (AI)-enabled activities in digital twin 6G-enabled
Internet of Things (IoT) environments. However, we need to also consider the
potential of attacks targeting the underlying AI systems (e.g., adversaries
seek to corrupt data on the IoT devices during local updates or corrupt the
model updates); hence, in this article, we propose an anticipatory study for
poisoning attacks in federated edge learning for digital twin 6G-enabled IoT
environments. Specifically, we study the influence of adversaries on the
training and development of federated learning models in digital twin
6G-enabled IoT environments. We demonstrate that attackers can carry out
poisoning attacks in two different learning settings, namely: centralized
learning and federated learning, and successful attacks can severely reduce the
model's accuracy. We comprehensively evaluate the attacks on a new cyber
security dataset designed for IoT applications with three deep neural networks
under the non-independent and identically distributed (Non-IID) data and the
independent and identically distributed (IID) data. The poisoning attacks, on
an attack classification problem, can lead to a decrease in accuracy from
94.93% to 85.98% with IID data and from 94.18% to 30.04% with Non-IID.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:12:17 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Ferrag",
"Mohamed Amine",
""
],
[
"Kantarci",
"Burak",
""
],
[
"Cordeiro",
"Lucas C.",
""
],
[
"Debbah",
"Merouane",
""
],
[
"Choo",
"Kim-Kwang Raymond",
""
]
] | TITLE: Poisoning Attacks in Federated Edge Learning for Digital Twin 6G-enabled
IoTs: An Anticipatory Study
ABSTRACT: Federated edge learning can be essential in supporting privacy-preserving,
artificial intelligence (AI)-enabled activities in digital twin 6G-enabled
Internet of Things (IoT) environments. However, we need to also consider the
potential of attacks targeting the underlying AI systems (e.g., adversaries
seek to corrupt data on the IoT devices during local updates or corrupt the
model updates); hence, in this article, we propose an anticipatory study for
poisoning attacks in federated edge learning for digital twin 6G-enabled IoT
environments. Specifically, we study the influence of adversaries on the
training and development of federated learning models in digital twin
6G-enabled IoT environments. We demonstrate that attackers can carry out
poisoning attacks in two different learning settings, namely: centralized
learning and federated learning, and successful attacks can severely reduce the
model's accuracy. We comprehensively evaluate the attacks on a new cyber
security dataset designed for IoT applications with three deep neural networks
under the non-independent and identically distributed (Non-IID) data and the
independent and identically distributed (IID) data. The poisoning attacks, on
an attack classification problem, can lead to a decrease in accuracy from
94.93% to 85.98% with IID data and from 94.18% to 30.04% with Non-IID.
| new_dataset | 0.96738 |
2303.11746 | Luca Vassio Mr. | Alessandro Speciale, Greta Vallero, Luca Vassio, Marco Mellia | Recommendation Systems in Libraries: an Application with Heterogeneous
Data Sources | Accepted at 7th International workshop on Data Analytics solutions
for Real-LIfe APplications - 28th March-31st March, 2023, Ioannina, Greece.
The paper will be published in the Proceedings of EDBT/ICDT 2023 Joint
Conference | null | null | null | cs.IR cs.CY cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The Reading&Machine project exploits the support of digitalization to
increase the attractiveness of libraries and improve the users' experience. The
project implements an application that helps the users in their decision-making
process, providing recommendation system (RecSys)-generated lists of books the
users might be interested in, and showing them through an interactive Virtual
Reality (VR)-based Graphical User Interface (GUI). In this paper, we focus on
the design and testing of the recommendation system, employing data about all
users' loans over the past 9 years from the network of libraries located in
Turin, Italy. In addition, we use data collected by the Anobii online social
community of readers, who share their feedback and additional information about
books they read. Armed with this heterogeneous data, we build and evaluate
Content Based (CB) and Collaborative Filtering (CF) approaches. Our results
show that the CF outperforms the CB approach, improving by up to 47\% the
relevant recommendations provided to a reader. However, the performance of the
CB approach is heavily dependent on the number of books the reader has already
read, and it can work even better than CF for users with a large history.
Finally, our evaluations highlight that the performances of both approaches are
significantly improved if the system integrates and leverages the information
from the Anobii dataset, which allows us to include more user readings (for CF)
and richer book metadata (for CB).
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:13:01 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Speciale",
"Alessandro",
""
],
[
"Vallero",
"Greta",
""
],
[
"Vassio",
"Luca",
""
],
[
"Mellia",
"Marco",
""
]
] | TITLE: Recommendation Systems in Libraries: an Application with Heterogeneous
Data Sources
ABSTRACT: The Reading&Machine project exploits the support of digitalization to
increase the attractiveness of libraries and improve the users' experience. The
project implements an application that helps the users in their decision-making
process, providing recommendation system (RecSys)-generated lists of books the
users might be interested in, and showing them through an interactive Virtual
Reality (VR)-based Graphical User Interface (GUI). In this paper, we focus on
the design and testing of the recommendation system, employing data about all
users' loans over the past 9 years from the network of libraries located in
Turin, Italy. In addition, we use data collected by the Anobii online social
community of readers, who share their feedback and additional information about
books they read. Armed with this heterogeneous data, we build and evaluate
Content Based (CB) and Collaborative Filtering (CF) approaches. Our results
show that the CF outperforms the CB approach, improving by up to 47\% the
relevant recommendations provided to a reader. However, the performance of the
CB approach is heavily dependent on the number of books the reader has already
read, and it can work even better than CF for users with a large history.
Finally, our evaluations highlight that the performances of both approaches are
significantly improved if the system integrates and leverages the information
from the Anobii dataset, which allows us to include more user readings (for CF)
and richer book metadata (for CB).
| no_new_dataset | 0.944944 |
2303.11751 | Mohamed Amine Ferrag | Mohamed Amine Ferrag and Merouane Debbah and Muna Al-Hawawreh | Generative AI for Cyber Threat-Hunting in 6G-enabled IoT Networks | The paper is accepted and will be published in the IEEE/ACM CCGrid
2023 Conference Proceedings | null | null | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | The next generation of cellular technology, 6G, is being developed to enable
a wide range of new applications and services for the Internet of Things (IoT).
One of 6G's main advantages for IoT applications is its ability to support much
higher data rates and bandwidth as well as to support ultra-low latency.
However, with this increased connectivity will come to an increased risk of
cyber threats, as attackers will be able to exploit the large network of
connected devices. Generative Artificial Intelligence (AI) can be used to
detect and prevent cyber attacks by continuously learning and adapting to new
threats and vulnerabilities. In this paper, we discuss the use of generative AI
for cyber threat-hunting (CTH) in 6G-enabled IoT networks. Then, we propose a
new generative adversarial network (GAN) and Transformer-based model for CTH in
6G-enabled IoT Networks. The experimental analysis results with a new cyber
security dataset demonstrate that the Transformer-based security model for CTH
can detect IoT attacks with a high overall accuracy of 95%. We examine the
challenges and opportunities and conclude by highlighting the potential of
generative AI in enhancing the security of 6G-enabled IoT networks and call for
further research to be conducted in this area.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:17:41 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Ferrag",
"Mohamed Amine",
""
],
[
"Debbah",
"Merouane",
""
],
[
"Al-Hawawreh",
"Muna",
""
]
] | TITLE: Generative AI for Cyber Threat-Hunting in 6G-enabled IoT Networks
ABSTRACT: The next generation of cellular technology, 6G, is being developed to enable
a wide range of new applications and services for the Internet of Things (IoT).
One of 6G's main advantages for IoT applications is its ability to support much
higher data rates and bandwidth as well as to support ultra-low latency.
However, with this increased connectivity will come to an increased risk of
cyber threats, as attackers will be able to exploit the large network of
connected devices. Generative Artificial Intelligence (AI) can be used to
detect and prevent cyber attacks by continuously learning and adapting to new
threats and vulnerabilities. In this paper, we discuss the use of generative AI
for cyber threat-hunting (CTH) in 6G-enabled IoT networks. Then, we propose a
new generative adversarial network (GAN) and Transformer-based model for CTH in
6G-enabled IoT Networks. The experimental analysis results with a new cyber
security dataset demonstrate that the Transformer-based security model for CTH
can detect IoT attacks with a high overall accuracy of 95%. We examine the
challenges and opportunities and conclude by highlighting the potential of
generative AI in enhancing the security of 6G-enabled IoT networks and call for
further research to be conducted in this area.
| new_dataset | 0.964589 |
2303.11765 | Achu Wilson | Achu Wilson, Helen Jiang, Wenzhao Lian, Wenzhen Yuan | Cable Routing and Assembly using Tactile-driven Motion Primitives | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Manipulating cables is challenging for robots because of the infinite degrees
of freedom of the cables and frequent occlusion by the gripper and the
environment. These challenges are further complicated by the dexterous nature
of the operations required for cable routing and assembly, such as weaving and
inserting, hampering common solutions with vision-only sensing. In this paper,
we propose to integrate tactile-guided low-level motion control with high-level
vision-based task parsing for a challenging task: cable routing and assembly on
a reconfigurable task board. Specifically, we build a library of tactile-guided
motion primitives using a fingertip GelSight sensor, where each primitive
reliably accomplishes an operation such as cable following and weaving. The
overall task is inferred via visual perception given a goal configuration
image, and then used to generate the primitive sequence. Experiments
demonstrate the effectiveness of individual tactile-guided primitives and the
integrated end-to-end solution, significantly outperforming the method without
tactile sensing. Our reconfigurable task setup and proposed baselines provide a
benchmark for future research in cable manipulation. More details and video are
presented in \url{https://helennn.github.io/cable-manip/}
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:36:37 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Wilson",
"Achu",
""
],
[
"Jiang",
"Helen",
""
],
[
"Lian",
"Wenzhao",
""
],
[
"Yuan",
"Wenzhen",
""
]
] | TITLE: Cable Routing and Assembly using Tactile-driven Motion Primitives
ABSTRACT: Manipulating cables is challenging for robots because of the infinite degrees
of freedom of the cables and frequent occlusion by the gripper and the
environment. These challenges are further complicated by the dexterous nature
of the operations required for cable routing and assembly, such as weaving and
inserting, hampering common solutions with vision-only sensing. In this paper,
we propose to integrate tactile-guided low-level motion control with high-level
vision-based task parsing for a challenging task: cable routing and assembly on
a reconfigurable task board. Specifically, we build a library of tactile-guided
motion primitives using a fingertip GelSight sensor, where each primitive
reliably accomplishes an operation such as cable following and weaving. The
overall task is inferred via visual perception given a goal configuration
image, and then used to generate the primitive sequence. Experiments
demonstrate the effectiveness of individual tactile-guided primitives and the
integrated end-to-end solution, significantly outperforming the method without
tactile sensing. Our reconfigurable task setup and proposed baselines provide a
benchmark for future research in cable manipulation. More details and video are
presented in \url{https://helennn.github.io/cable-manip/}
| no_new_dataset | 0.940463 |
2303.11771 | Youngjoon Jang | Youngjoon Jang, Youngtaek Oh, Jae Won Cho, Myungchul Kim, Dong-Jin
Kim, In So Kweon, Joon Son Chung | Self-Sufficient Framework for Continuous Sign Language Recognition | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The goal of this work is to develop self-sufficient framework for Continuous
Sign Language Recognition (CSLR) that addresses key issues of sign language
recognition. These include the need for complex multi-scale features such as
hands, face, and mouth for understanding, and absence of frame-level
annotations. To this end, we propose (1) Divide and Focus Convolution (DFConv)
which extracts both manual and non-manual features without the need for
additional networks or annotations, and (2) Dense Pseudo-Label Refinement
(DPLR) which propagates non-spiky frame-level pseudo-labels by combining the
ground truth gloss sequence labels with the predicted sequence. We demonstrate
that our model achieves state-of-the-art performance among RGB-based methods on
large-scale CSLR benchmarks, PHOENIX-2014 and PHOENIX-2014-T, while showing
comparable results with better efficiency when compared to other approaches
that use multi-modality or extra annotations.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:42:57 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Jang",
"Youngjoon",
""
],
[
"Oh",
"Youngtaek",
""
],
[
"Cho",
"Jae Won",
""
],
[
"Kim",
"Myungchul",
""
],
[
"Kim",
"Dong-Jin",
""
],
[
"Kweon",
"In So",
""
],
[
"Chung",
"Joon Son",
""
]
] | TITLE: Self-Sufficient Framework for Continuous Sign Language Recognition
ABSTRACT: The goal of this work is to develop self-sufficient framework for Continuous
Sign Language Recognition (CSLR) that addresses key issues of sign language
recognition. These include the need for complex multi-scale features such as
hands, face, and mouth for understanding, and absence of frame-level
annotations. To this end, we propose (1) Divide and Focus Convolution (DFConv)
which extracts both manual and non-manual features without the need for
additional networks or annotations, and (2) Dense Pseudo-Label Refinement
(DPLR) which propagates non-spiky frame-level pseudo-labels by combining the
ground truth gloss sequence labels with the predicted sequence. We demonstrate
that our model achieves state-of-the-art performance among RGB-based methods on
large-scale CSLR benchmarks, PHOENIX-2014 and PHOENIX-2014-T, while showing
comparable results with better efficiency when compared to other approaches
that use multi-modality or extra annotations.
| no_new_dataset | 0.950041 |
2303.11772 | Niklas Vogel | Tomas Hlavacek, Haya Shulman, Niklas Vogel, Michael Waidner | Keep Your Friends Close, but Your Routeservers Closer: Insights into
RPKI Validation in the Internet | Accepted for USENIX Security '23 | null | null | null | cs.NI cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | IP prefix hijacks allow adversaries to redirect and intercept traffic, posing
a threat to the stability and security of the Internet. To prevent prefix
hijacks, networks should deploy RPKI and filter bogus BGP announcements with
invalid routes.
In this work we evaluate the impact of RPKI deployments on the security and
resilience of the Internet. We aim to understand which networks filter invalid
routes and how effective that filtering is in blocking prefix hijacks. We
extend previous data acquisition and analysis methodologies to obtain more
accurate identification of networks that filter invalid routes with RPKI. We
find that more than 27% of networks enforce RPKI filtering and show for the
first time that deployments follow the business incentives of inter-domain
routing: providers have an increased motivation to filter in order to avoid
losing customers' traffic.
Analyzing the effectiveness of RPKI, we find that the current trend to deploy
RPKI on routeservers of Internet Exchange Points (IXPs) only provides a
localized protection against hijacks but has negligible impact on preventing
their spread globally. In contrast, we show that RPKI filtering in Tier-1
providers greatly benefits the security of the Internet as it limits the spread
of hijacks to a localized scope. Based on our observations, we provide
recommendations on the future roadmap of RPKI deployment.
We make our datasets available for public use [https://sit4.me/rpki].
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:43:09 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Hlavacek",
"Tomas",
""
],
[
"Shulman",
"Haya",
""
],
[
"Vogel",
"Niklas",
""
],
[
"Waidner",
"Michael",
""
]
] | TITLE: Keep Your Friends Close, but Your Routeservers Closer: Insights into
RPKI Validation in the Internet
ABSTRACT: IP prefix hijacks allow adversaries to redirect and intercept traffic, posing
a threat to the stability and security of the Internet. To prevent prefix
hijacks, networks should deploy RPKI and filter bogus BGP announcements with
invalid routes.
In this work we evaluate the impact of RPKI deployments on the security and
resilience of the Internet. We aim to understand which networks filter invalid
routes and how effective that filtering is in blocking prefix hijacks. We
extend previous data acquisition and analysis methodologies to obtain more
accurate identification of networks that filter invalid routes with RPKI. We
find that more than 27% of networks enforce RPKI filtering and show for the
first time that deployments follow the business incentives of inter-domain
routing: providers have an increased motivation to filter in order to avoid
losing customers' traffic.
Analyzing the effectiveness of RPKI, we find that the current trend to deploy
RPKI on routeservers of Internet Exchange Points (IXPs) only provides a
localized protection against hijacks but has negligible impact on preventing
their spread globally. In contrast, we show that RPKI filtering in Tier-1
providers greatly benefits the security of the Internet as it limits the spread
of hijacks to a localized scope. Based on our observations, we provide
recommendations on the future roadmap of RPKI deployment.
We make our datasets available for public use [https://sit4.me/rpki].
| no_new_dataset | 0.933188 |
2303.11780 | Lianghao Xia | Yuhao Yang, Chao Huang, Lianghao Xia, Chunzhen Huang, Da Luo and
Kangyi Lin | Debiased Contrastive Learning for Sequential Recommendation | This paper is accepted by WWW'2023 | null | 10.1145/3543507.3583361 | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current sequential recommender systems are proposed to tackle the dynamic
user preference learning with various neural techniques, such as Transformer
and Graph Neural Networks (GNNs). However, inference from the highly sparse
user behavior data may hinder the representation ability of sequential pattern
encoding. To address the label shortage issue, contrastive learning (CL)
methods are proposed recently to perform data augmentation in two fashions: (i)
randomly corrupting the sequence data (e.g. stochastic masking, reordering);
(ii) aligning representations across pre-defined contrastive views. Although
effective, we argue that current CL-based methods have limitations in
addressing popularity bias and disentangling of user conformity and real
interest. In this paper, we propose a new Debiased Contrastive learning
paradigm for Recommendation (DCRec) that unifies sequential pattern encoding
with global collaborative relation modeling through adaptive conformity-aware
augmentation. This solution is designed to tackle the popularity bias issue in
recommendation systems. Our debiased contrastive learning framework effectively
captures both the patterns of item transitions within sequences and the
dependencies between users across sequences. Our experiments on various
real-world datasets have demonstrated that DCRec significantly outperforms
state-of-the-art baselines, indicating its efficacy for recommendation. To
facilitate reproducibility of our results, we make our implementation of DCRec
publicly available at: https://github.com/HKUDS/DCRec.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 11:56:35 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Yang",
"Yuhao",
""
],
[
"Huang",
"Chao",
""
],
[
"Xia",
"Lianghao",
""
],
[
"Huang",
"Chunzhen",
""
],
[
"Luo",
"Da",
""
],
[
"Lin",
"Kangyi",
""
]
] | TITLE: Debiased Contrastive Learning for Sequential Recommendation
ABSTRACT: Current sequential recommender systems are proposed to tackle the dynamic
user preference learning with various neural techniques, such as Transformer
and Graph Neural Networks (GNNs). However, inference from the highly sparse
user behavior data may hinder the representation ability of sequential pattern
encoding. To address the label shortage issue, contrastive learning (CL)
methods are proposed recently to perform data augmentation in two fashions: (i)
randomly corrupting the sequence data (e.g. stochastic masking, reordering);
(ii) aligning representations across pre-defined contrastive views. Although
effective, we argue that current CL-based methods have limitations in
addressing popularity bias and disentangling of user conformity and real
interest. In this paper, we propose a new Debiased Contrastive learning
paradigm for Recommendation (DCRec) that unifies sequential pattern encoding
with global collaborative relation modeling through adaptive conformity-aware
augmentation. This solution is designed to tackle the popularity bias issue in
recommendation systems. Our debiased contrastive learning framework effectively
captures both the patterns of item transitions within sequences and the
dependencies between users across sequences. Our experiments on various
real-world datasets have demonstrated that DCRec significantly outperforms
state-of-the-art baselines, indicating its efficacy for recommendation. To
facilitate reproducibility of our results, we make our implementation of DCRec
publicly available at: https://github.com/HKUDS/DCRec.
| no_new_dataset | 0.94625 |
2303.11790 | Constantin Pape | Anwai Archit and Constantin Pape | Probabilistic Domain Adaptation for Biomedical Image Segmentation | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Segmentation is a key analysis tasks in biomedical imaging. Given the many
different experimental settings in this field, the lack of generalization
limits the use of deep learning in practice. Domain adaptation is a promising
remedy: it trains a model for a given task on a source dataset with labels and
adapts it to a target dataset without additional labels. We introduce a
probabilistic domain adaptation method, building on self-training approaches
and the Probabilistic UNet. We use the latter to sample multiple segmentation
hypothesis to implement better pseudo-label filtering. We further study joint
and separate source-target training strategies and evaluate our method on three
challenging domain adaptation tasks for biomedical segmentation.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 12:17:21 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Archit",
"Anwai",
""
],
[
"Pape",
"Constantin",
""
]
] | TITLE: Probabilistic Domain Adaptation for Biomedical Image Segmentation
ABSTRACT: Segmentation is a key analysis tasks in biomedical imaging. Given the many
different experimental settings in this field, the lack of generalization
limits the use of deep learning in practice. Domain adaptation is a promising
remedy: it trains a model for a given task on a source dataset with labels and
adapts it to a target dataset without additional labels. We introduce a
probabilistic domain adaptation method, building on self-training approaches
and the Probabilistic UNet. We use the latter to sample multiple segmentation
hypothesis to implement better pseudo-label filtering. We further study joint
and separate source-target training strategies and evaluate our method on three
challenging domain adaptation tasks for biomedical segmentation.
| no_new_dataset | 0.95222 |
2303.11828 | Caixia Zhou | Caixia Zhou and Yaping Huang and Mengyang Pu and Qingji Guan and Li
Huang and Haibin Ling | The Treasure Beneath Multiple Annotations: An Uncertainty-aware Edge
Detector | CVPR2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning-based edge detectors heavily rely on pixel-wise labels which
are often provided by multiple annotators. Existing methods fuse multiple
annotations using a simple voting process, ignoring the inherent ambiguity of
edges and labeling bias of annotators. In this paper, we propose a novel
uncertainty-aware edge detector (UAED), which employs uncertainty to
investigate the subjectivity and ambiguity of diverse annotations.
Specifically, we first convert the deterministic label space into a learnable
Gaussian distribution, whose variance measures the degree of ambiguity among
different annotations. Then we regard the learned variance as the estimated
uncertainty of the predicted edge maps, and pixels with higher uncertainty are
likely to be hard samples for edge detection. Therefore we design an adaptive
weighting loss to emphasize the learning from those pixels with high
uncertainty, which helps the network to gradually concentrate on the important
pixels. UAED can be combined with various encoder-decoder backbones, and the
extensive experiments demonstrate that UAED achieves superior performance
consistently across multiple edge detection benchmarks. The source code is
available at \url{https://github.com/ZhouCX117/UAED}
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:14:36 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Zhou",
"Caixia",
""
],
[
"Huang",
"Yaping",
""
],
[
"Pu",
"Mengyang",
""
],
[
"Guan",
"Qingji",
""
],
[
"Huang",
"Li",
""
],
[
"Ling",
"Haibin",
""
]
] | TITLE: The Treasure Beneath Multiple Annotations: An Uncertainty-aware Edge
Detector
ABSTRACT: Deep learning-based edge detectors heavily rely on pixel-wise labels which
are often provided by multiple annotators. Existing methods fuse multiple
annotations using a simple voting process, ignoring the inherent ambiguity of
edges and labeling bias of annotators. In this paper, we propose a novel
uncertainty-aware edge detector (UAED), which employs uncertainty to
investigate the subjectivity and ambiguity of diverse annotations.
Specifically, we first convert the deterministic label space into a learnable
Gaussian distribution, whose variance measures the degree of ambiguity among
different annotations. Then we regard the learned variance as the estimated
uncertainty of the predicted edge maps, and pixels with higher uncertainty are
likely to be hard samples for edge detection. Therefore we design an adaptive
weighting loss to emphasize the learning from those pixels with high
uncertainty, which helps the network to gradually concentrate on the important
pixels. UAED can be combined with various encoder-decoder backbones, and the
extensive experiments demonstrate that UAED achieves superior performance
consistently across multiple edge detection benchmarks. The source code is
available at \url{https://github.com/ZhouCX117/UAED}
| no_new_dataset | 0.946695 |
2303.11853 | Donghwi Jung | Donghwi Jung, Jae-Kyung Cho, Younghwa Jung, Soohyun Shin, and
Seong-Woo Kim | LoRCoN-LO: Long-term Recurrent Convolutional Network-based LiDAR
Odometry | 4 pages, ICEIC 2023 | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | We propose a deep learning-based LiDAR odometry estimation method called
LoRCoN-LO that utilizes the long-term recurrent convolutional network (LRCN)
structure. The LRCN layer is a structure that can process spatial and temporal
information at once by using both CNN and LSTM layers. This feature is suitable
for predicting continuous robot movements as it uses point clouds that contain
spatial information. Therefore, we built a LoRCoN-LO model using the LRCN
layer, and predicted the pose of the robot through this model. For performance
verification, we conducted experiments exploiting a public dataset (KITTI). The
results of the experiment show that LoRCoN-LO displays accurate odometry
prediction in the dataset. The code is available at
https://github.com/donghwijung/LoRCoN-LO.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:52:00 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Jung",
"Donghwi",
""
],
[
"Cho",
"Jae-Kyung",
""
],
[
"Jung",
"Younghwa",
""
],
[
"Shin",
"Soohyun",
""
],
[
"Kim",
"Seong-Woo",
""
]
] | TITLE: LoRCoN-LO: Long-term Recurrent Convolutional Network-based LiDAR
Odometry
ABSTRACT: We propose a deep learning-based LiDAR odometry estimation method called
LoRCoN-LO that utilizes the long-term recurrent convolutional network (LRCN)
structure. The LRCN layer is a structure that can process spatial and temporal
information at once by using both CNN and LSTM layers. This feature is suitable
for predicting continuous robot movements as it uses point clouds that contain
spatial information. Therefore, we built a LoRCoN-LO model using the LRCN
layer, and predicted the pose of the robot through this model. For performance
verification, we conducted experiments exploiting a public dataset (KITTI). The
results of the experiment show that LoRCoN-LO displays accurate odometry
prediction in the dataset. The code is available at
https://github.com/donghwijung/LoRCoN-LO.
| no_new_dataset | 0.953923 |
2303.11854 | Yuanyuan Yang | Yuanyuan Yang and Bowen Xu and Yinjie Li and S\"oren Schwertfeger | The SLAM Hive Benchmarking Suite | 7 pages, 3 figures, IEEE Conference on Robotics and Automation (ICRA)
2023 | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Benchmarking Simultaneous Localization and Mapping (SLAM) algorithms is
important to scientists and users of robotic systems alike. But through their
many configuration options in hardware and software, SLAM systems feature a
vast parameter space that scientists up to now were not able to explore. The
proposed SLAM Hive Benchmarking Suite is able to analyze SLAM algorithms in
1000's of mapping runs, through its utilization of container technology and
deployment in a cluster. This paper presents the architecture and open source
implementation of SLAM Hive and compares it to existing efforts on SLAM
evaluation. Furthermore, we highlight the function of SLAM Hive by exploring
some open source algorithms on public datasets in terms of accuracy. We compare
the algorithms against each other and evaluate how parameters effect not only
accuracy but also CPU and memory usage. Through this we show that SLAM Hive can
become an essential tool for proper comparisons and evaluations of SLAM
algorithms and thus drive the scientific development in the research on SLAM.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:52:43 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Yang",
"Yuanyuan",
""
],
[
"Xu",
"Bowen",
""
],
[
"Li",
"Yinjie",
""
],
[
"Schwertfeger",
"Sören",
""
]
] | TITLE: The SLAM Hive Benchmarking Suite
ABSTRACT: Benchmarking Simultaneous Localization and Mapping (SLAM) algorithms is
important to scientists and users of robotic systems alike. But through their
many configuration options in hardware and software, SLAM systems feature a
vast parameter space that scientists up to now were not able to explore. The
proposed SLAM Hive Benchmarking Suite is able to analyze SLAM algorithms in
1000's of mapping runs, through its utilization of container technology and
deployment in a cluster. This paper presents the architecture and open source
implementation of SLAM Hive and compares it to existing efforts on SLAM
evaluation. Furthermore, we highlight the function of SLAM Hive by exploring
some open source algorithms on public datasets in terms of accuracy. We compare
the algorithms against each other and evaluate how parameters effect not only
accuracy but also CPU and memory usage. Through this we show that SLAM Hive can
become an essential tool for proper comparisons and evaluations of SLAM
algorithms and thus drive the scientific development in the research on SLAM.
| no_new_dataset | 0.943815 |
2303.11855 | Fabian Deuser | Konrad Habel, Fabian Deuser, Norbert Oswald | CLIP-ReIdent: Contrastive Training for Player Re-Identification | null | null | 10.1145/3552437.3555698 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Sports analytics benefits from recent advances in machine learning providing
a competitive advantage for teams or individuals. One important task in this
context is the performance measurement of individual players to provide reports
and log files for subsequent analysis. During sport events like basketball,
this involves the re-identification of players during a match either from
multiple camera viewpoints or from a single camera viewpoint at different
times. In this work, we investigate whether it is possible to transfer the
out-standing zero-shot performance of pre-trained CLIP models to the domain of
player re-identification. For this purpose we reformulate the contrastive
language-to-image pre-training approach from CLIP to a contrastive
image-to-image training approach using the InfoNCE loss as training objective.
Unlike previous work, our approach is entirely class-agnostic and benefits from
large-scale pre-training. With a fine-tuned CLIP ViT-L/14 model we achieve
98.44 % mAP on the MMSports 2022 Player Re-Identification challenge.
Furthermore we show that the CLIP Vision Transformers have already strong OCR
capabilities to identify useful player features like shirt numbers in a
zero-shot manner without any fine-tuning on the dataset. By applying the
Score-CAM algorithm we visualise the most important image regions that our
fine-tuned model identifies when calculating the similarity score between two
images of a player.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:55:27 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Habel",
"Konrad",
""
],
[
"Deuser",
"Fabian",
""
],
[
"Oswald",
"Norbert",
""
]
] | TITLE: CLIP-ReIdent: Contrastive Training for Player Re-Identification
ABSTRACT: Sports analytics benefits from recent advances in machine learning providing
a competitive advantage for teams or individuals. One important task in this
context is the performance measurement of individual players to provide reports
and log files for subsequent analysis. During sport events like basketball,
this involves the re-identification of players during a match either from
multiple camera viewpoints or from a single camera viewpoint at different
times. In this work, we investigate whether it is possible to transfer the
out-standing zero-shot performance of pre-trained CLIP models to the domain of
player re-identification. For this purpose we reformulate the contrastive
language-to-image pre-training approach from CLIP to a contrastive
image-to-image training approach using the InfoNCE loss as training objective.
Unlike previous work, our approach is entirely class-agnostic and benefits from
large-scale pre-training. With a fine-tuned CLIP ViT-L/14 model we achieve
98.44 % mAP on the MMSports 2022 Player Re-Identification challenge.
Furthermore we show that the CLIP Vision Transformers have already strong OCR
capabilities to identify useful player features like shirt numbers in a
zero-shot manner without any fine-tuning on the dataset. By applying the
Score-CAM algorithm we visualise the most important image regions that our
fine-tuned model identifies when calculating the similarity score between two
images of a player.
| no_new_dataset | 0.945901 |
2303.11859 | Jiale Cao | Zhiqiang Dong, Jiale Cao, Rao Muhammad Anwer, Jin Xie, Fahad Khan,
Yanwei Pang | LEAPS: End-to-End One-Step Person Search With Learnable Proposals | 11 pages, 9 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose an end-to-end one-step person search approach with learnable
proposals, named LEAPS. Given a set of sparse and learnable proposals, LEAPS
employs a dynamic person search head to directly perform person detection and
corresponding re-id feature generation without non-maximum suppression
post-processing. The dynamic person search head comprises a detection head and
a novel flexible re-id head. Our flexible re-id head first employs a dynamic
region-of-interest (RoI) operation to extract discriminative RoI features of
the proposals. Then, it generates re-id features using a plain and a
hierarchical interaction re-id module. To better guide discriminative re-id
feature learning, we introduce a diverse re-id sample matching strategy,
instead of bipartite matching in detection head. Comprehensive experiments
reveal the benefit of the proposed LEAPS, achieving a favorable performance on
two public person search benchmarks: CUHK-SYSU and PRW. When using the same
ResNet50 backbone, our LEAPS obtains a mAP score of 55.0%, outperforming the
best reported results in literature by 1.7%, while achieving around a two-fold
speedup on the challenging PRW dataset. Our source code and models will be
released.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:59:32 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Dong",
"Zhiqiang",
""
],
[
"Cao",
"Jiale",
""
],
[
"Anwer",
"Rao Muhammad",
""
],
[
"Xie",
"Jin",
""
],
[
"Khan",
"Fahad",
""
],
[
"Pang",
"Yanwei",
""
]
] | TITLE: LEAPS: End-to-End One-Step Person Search With Learnable Proposals
ABSTRACT: We propose an end-to-end one-step person search approach with learnable
proposals, named LEAPS. Given a set of sparse and learnable proposals, LEAPS
employs a dynamic person search head to directly perform person detection and
corresponding re-id feature generation without non-maximum suppression
post-processing. The dynamic person search head comprises a detection head and
a novel flexible re-id head. Our flexible re-id head first employs a dynamic
region-of-interest (RoI) operation to extract discriminative RoI features of
the proposals. Then, it generates re-id features using a plain and a
hierarchical interaction re-id module. To better guide discriminative re-id
feature learning, we introduce a diverse re-id sample matching strategy,
instead of bipartite matching in detection head. Comprehensive experiments
reveal the benefit of the proposed LEAPS, achieving a favorable performance on
two public person search benchmarks: CUHK-SYSU and PRW. When using the same
ResNet50 backbone, our LEAPS obtains a mAP score of 55.0%, outperforming the
best reported results in literature by 1.7%, while achieving around a two-fold
speedup on the challenging PRW dataset. Our source code and models will be
released.
| no_new_dataset | 0.942929 |
2303.11860 | Nathan Leroux | Nathan Leroux, Jan Finkbeiner, Emre Neftci | Online Transformers with Spiking Neurons for Fast Prosthetic Hand
Control | Preprint of 9 pages, 4 figures | null | null | null | cs.NE cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Transformers are state-of-the-art networks for most sequence processing
tasks. However, the self-attention mechanism often used in Transformers
requires large time windows for each computation step and thus makes them less
suitable for online signal processing compared to Recurrent Neural Networks
(RNNs). In this paper, instead of the self-attention mechanism, we use a
sliding window attention mechanism. We show that this mechanism is more
efficient for continuous signals with finite-range dependencies between input
and target, and that we can use it to process sequences element-by-element,
this making it compatible with online processing. We test our model on a finger
position regression dataset (NinaproDB8) with Surface Electromyographic (sEMG)
signals measured on the forearm skin to estimate muscle activities. Our
approach sets the new state-of-the-art in terms of accuracy on this dataset
while requiring only very short time windows of 3.5 ms at each inference step.
Moreover, we increase the sparsity of the network using Leaky-Integrate and
Fire (LIF) units, a bio-inspired neuron model that activates sparsely in time
solely when crossing a threshold. We thus reduce the number of synaptic
operations up to a factor of $\times5.3$ without loss of accuracy. Our results
hold great promises for accurate and fast online processing of sEMG signals for
smooth prosthetic hand control and is a step towards Transformers and Spiking
Neural Networks (SNNs) co-integration for energy efficient temporal signal
processing.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 13:59:35 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Leroux",
"Nathan",
""
],
[
"Finkbeiner",
"Jan",
""
],
[
"Neftci",
"Emre",
""
]
] | TITLE: Online Transformers with Spiking Neurons for Fast Prosthetic Hand
Control
ABSTRACT: Transformers are state-of-the-art networks for most sequence processing
tasks. However, the self-attention mechanism often used in Transformers
requires large time windows for each computation step and thus makes them less
suitable for online signal processing compared to Recurrent Neural Networks
(RNNs). In this paper, instead of the self-attention mechanism, we use a
sliding window attention mechanism. We show that this mechanism is more
efficient for continuous signals with finite-range dependencies between input
and target, and that we can use it to process sequences element-by-element,
this making it compatible with online processing. We test our model on a finger
position regression dataset (NinaproDB8) with Surface Electromyographic (sEMG)
signals measured on the forearm skin to estimate muscle activities. Our
approach sets the new state-of-the-art in terms of accuracy on this dataset
while requiring only very short time windows of 3.5 ms at each inference step.
Moreover, we increase the sparsity of the network using Leaky-Integrate and
Fire (LIF) units, a bio-inspired neuron model that activates sparsely in time
solely when crossing a threshold. We thus reduce the number of synaptic
operations up to a factor of $\times5.3$ without loss of accuracy. Our results
hold great promises for accurate and fast online processing of sEMG signals for
smooth prosthetic hand control and is a step towards Transformers and Spiking
Neural Networks (SNNs) co-integration for energy efficient temporal signal
processing.
| no_new_dataset | 0.951684 |
2303.11862 | Jean-Christophe Sibel | Vincent Corlay and Jean-Christophe Sibel | An MDP approach for radio resource allocation in urban Future Railway
Mobile Communication System (FRMCS) scenarios | submitted to VTC-spring 2023 | null | null | null | cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | In the context of railway systems, the application performance can be very
critical and the radio conditions not advantageous. Hence, the communication
problem parameters include both a survival time stemming from the application
layer and a channel error probability stemming from the PHY layer. This paper
proposes to consider the framework of Markov Decision Process (MDP) to design a
strategy for scheduling radio resources based on both application and PHY layer
parameters. The MDP approach enables to obtain the optimal strategy via the
value iteration algorithm. The performance of this algorithm can thus serve as
a benchmark to assess lower complexity schedulers. We show numerical
evaluations where we compare the value iteration algorithm with other
schedulers, including one based on deep Q learning.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 14:05:55 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Corlay",
"Vincent",
""
],
[
"Sibel",
"Jean-Christophe",
""
]
] | TITLE: An MDP approach for radio resource allocation in urban Future Railway
Mobile Communication System (FRMCS) scenarios
ABSTRACT: In the context of railway systems, the application performance can be very
critical and the radio conditions not advantageous. Hence, the communication
problem parameters include both a survival time stemming from the application
layer and a channel error probability stemming from the PHY layer. This paper
proposes to consider the framework of Markov Decision Process (MDP) to design a
strategy for scheduling radio resources based on both application and PHY layer
parameters. The MDP approach enables to obtain the optimal strategy via the
value iteration algorithm. The performance of this algorithm can thus serve as
a benchmark to assess lower complexity schedulers. We show numerical
evaluations where we compare the value iteration algorithm with other
schedulers, including one based on deep Q learning.
| no_new_dataset | 0.94474 |
2303.11863 | Sangwon Jung | Donggyu Lee, Sangwon Jung, Taesup Moon | Continual Learning in the Presence of Spurious Correlation | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most continual learning (CL) algorithms have focused on tackling the
stability-plasticity dilemma, that is, the challenge of preventing the
forgetting of previous tasks while learning new ones. However, they have
overlooked the impact of the knowledge transfer when the dataset in a certain
task is biased - namely, when some unintended spurious correlations of the
tasks are learned from the biased dataset. In that case, how would they affect
learning future tasks or the knowledge already learned from the past tasks? In
this work, we carefully design systematic experiments using one synthetic and
two real-world datasets to answer the question from our empirical findings.
Specifically, we first show through two-task CL experiments that standard CL
methods, which are unaware of dataset bias, can transfer biases from one task
to another, both forward and backward, and this transfer is exacerbated
depending on whether the CL methods focus on the stability or the plasticity.
We then present that the bias transfer also exists and even accumulate in
longer sequences of tasks. Finally, we propose a simple, yet strong plug-in
method for debiasing-aware continual learning, dubbed as Group-class Balanced
Greedy Sampling (BGS). As a result, we show that our BGS can always reduce the
bias of a CL model, with a slight loss of CL performance at most.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 14:06:12 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Lee",
"Donggyu",
""
],
[
"Jung",
"Sangwon",
""
],
[
"Moon",
"Taesup",
""
]
] | TITLE: Continual Learning in the Presence of Spurious Correlation
ABSTRACT: Most continual learning (CL) algorithms have focused on tackling the
stability-plasticity dilemma, that is, the challenge of preventing the
forgetting of previous tasks while learning new ones. However, they have
overlooked the impact of the knowledge transfer when the dataset in a certain
task is biased - namely, when some unintended spurious correlations of the
tasks are learned from the biased dataset. In that case, how would they affect
learning future tasks or the knowledge already learned from the past tasks? In
this work, we carefully design systematic experiments using one synthetic and
two real-world datasets to answer the question from our empirical findings.
Specifically, we first show through two-task CL experiments that standard CL
methods, which are unaware of dataset bias, can transfer biases from one task
to another, both forward and backward, and this transfer is exacerbated
depending on whether the CL methods focus on the stability or the plasticity.
We then present that the bias transfer also exists and even accumulate in
longer sequences of tasks. Finally, we propose a simple, yet strong plug-in
method for debiasing-aware continual learning, dubbed as Group-class Balanced
Greedy Sampling (BGS). As a result, we show that our BGS can always reduce the
bias of a CL model, with a slight loss of CL performance at most.
| no_new_dataset | 0.938576 |
2303.11866 | Zaid Khan | Zaid Khan and Yun Fu | Contrastive Alignment of Vision to Language Through Parameter-Efficient
Transfer Learning | Accepted to ICLR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Contrastive vision-language models (e.g. CLIP) are typically created by
updating all the parameters of a vision model and language model through
contrastive training. Can such models be created by a small number of parameter
updates to an already-trained language model and vision model? The literature
describes techniques that can create vision-language models by updating a small
number of parameters in a language model, but these require already aligned
visual representations and are non-contrastive, hence unusable for
latency-sensitive applications such as neural search. We explore the
feasibility and benefits of parameter-efficient contrastive vision-language
alignment through transfer learning: creating a model such as CLIP by minimally
updating an already-trained vision and language model. We find that a minimal
set of parameter updates ($<$7%) can achieve the same performance as full-model
training, and updating specific components ($<$1% of parameters) can match 75%
of full-model training. We describe a series of experiments: we show that
existing knowledge is conserved more strongly in parameter-efficient training
and that parameter-efficient scaling scales with model and dataset size. Where
paired-image text data is scarce but strong multilingual language models exist
(e.g. low resource languages), parameter-efficient training is even preferable
to full-model training. Given a fixed compute budget, parameter-efficient
training allows training larger models on the same hardware, achieving
equivalent performance in less time. Parameter-efficient training hence
constitutes an energy-efficient and effective training strategy for contrastive
vision-language models that may be preferable to the full-model training
paradigm for common use cases. Code and weights at
https://github.com/codezakh/LilT.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 14:12:08 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Khan",
"Zaid",
""
],
[
"Fu",
"Yun",
""
]
] | TITLE: Contrastive Alignment of Vision to Language Through Parameter-Efficient
Transfer Learning
ABSTRACT: Contrastive vision-language models (e.g. CLIP) are typically created by
updating all the parameters of a vision model and language model through
contrastive training. Can such models be created by a small number of parameter
updates to an already-trained language model and vision model? The literature
describes techniques that can create vision-language models by updating a small
number of parameters in a language model, but these require already aligned
visual representations and are non-contrastive, hence unusable for
latency-sensitive applications such as neural search. We explore the
feasibility and benefits of parameter-efficient contrastive vision-language
alignment through transfer learning: creating a model such as CLIP by minimally
updating an already-trained vision and language model. We find that a minimal
set of parameter updates ($<$7%) can achieve the same performance as full-model
training, and updating specific components ($<$1% of parameters) can match 75%
of full-model training. We describe a series of experiments: we show that
existing knowledge is conserved more strongly in parameter-efficient training
and that parameter-efficient scaling scales with model and dataset size. Where
paired-image text data is scarce but strong multilingual language models exist
(e.g. low resource languages), parameter-efficient training is even preferable
to full-model training. Given a fixed compute budget, parameter-efficient
training allows training larger models on the same hardware, achieving
equivalent performance in less time. Parameter-efficient training hence
constitutes an energy-efficient and effective training strategy for contrastive
vision-language models that may be preferable to the full-model training
paradigm for common use cases. Code and weights at
https://github.com/codezakh/LilT.
| no_new_dataset | 0.958343 |
2303.11880 | Qiaoqiao Wei | Qiaoqiao Wei, Hui Zhang and Jun-Hai Yong | Focused and Collaborative Feedback Integration for Interactive Image
Segmentation | Accepted for publication at CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Interactive image segmentation aims at obtaining a segmentation mask for an
image using simple user annotations. During each round of interaction, the
segmentation result from the previous round serves as feedback to guide the
user's annotation and provides dense prior information for the segmentation
model, effectively acting as a bridge between interactions. Existing methods
overlook the importance of feedback or simply concatenate it with the original
input, leading to underutilization of feedback and an increase in the number of
required annotations. To address this, we propose an approach called Focused
and Collaborative Feedback Integration (FCFI) to fully exploit the feedback for
click-based interactive image segmentation. FCFI first focuses on a local area
around the new click and corrects the feedback based on the similarities of
high-level features. It then alternately and collaboratively updates the
feedback and deep features to integrate the feedback into the features. The
efficacy and efficiency of FCFI were validated on four benchmarks, namely
GrabCut, Berkeley, SBD, and DAVIS. Experimental results show that FCFI achieved
new state-of-the-art performance with less computational overhead than previous
methods. The source code is available at
https://github.com/veizgyauzgyauz/FCFI.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 14:24:06 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Wei",
"Qiaoqiao",
""
],
[
"Zhang",
"Hui",
""
],
[
"Yong",
"Jun-Hai",
""
]
] | TITLE: Focused and Collaborative Feedback Integration for Interactive Image
Segmentation
ABSTRACT: Interactive image segmentation aims at obtaining a segmentation mask for an
image using simple user annotations. During each round of interaction, the
segmentation result from the previous round serves as feedback to guide the
user's annotation and provides dense prior information for the segmentation
model, effectively acting as a bridge between interactions. Existing methods
overlook the importance of feedback or simply concatenate it with the original
input, leading to underutilization of feedback and an increase in the number of
required annotations. To address this, we propose an approach called Focused
and Collaborative Feedback Integration (FCFI) to fully exploit the feedback for
click-based interactive image segmentation. FCFI first focuses on a local area
around the new click and corrects the feedback based on the similarities of
high-level features. It then alternately and collaboratively updates the
feedback and deep features to integrate the feedback into the features. The
efficacy and efficiency of FCFI were validated on four benchmarks, namely
GrabCut, Berkeley, SBD, and DAVIS. Experimental results show that FCFI achieved
new state-of-the-art performance with less computational overhead than previous
methods. The source code is available at
https://github.com/veizgyauzgyauz/FCFI.
| no_new_dataset | 0.948632 |
2303.11911 | Wei Cheng | Dongsheng Luo, Wei Cheng, Yingheng Wang, Dongkuan Xu, Jingchao Ni,
Wenchao Yu, Xuchao Zhang, Yanchi Liu, Yuncong Chen, Haifeng Chen, Xiang Zhang | Time Series Contrastive Learning with Information-Aware Augmentations | Published in AAAI 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Various contrastive learning approaches have been proposed in recent years
and achieve significant empirical success. While effective and prevalent,
contrastive learning has been less explored for time series data. A key
component of contrastive learning is to select appropriate augmentations
imposing some priors to construct feasible positive samples, such that an
encoder can be trained to learn robust and discriminative representations.
Unlike image and language domains where ``desired'' augmented samples can be
generated with the rule of thumb guided by prefabricated human priors, the
ad-hoc manual selection of time series augmentations is hindered by their
diverse and human-unrecognizable temporal structures. How to find the desired
augmentations of time series data that are meaningful for given contrastive
learning tasks and datasets remains an open question. In this work, we address
the problem by encouraging both high \textit{fidelity} and \textit{variety}
based upon information theory. A theoretical analysis leads to the criteria for
selecting feasible data augmentations. On top of that, we propose a new
contrastive learning approach with information-aware augmentations, InfoTS,
that adaptively selects optimal augmentations for time series representation
learning. Experiments on various datasets show highly competitive performance
with up to 12.0\% reduction in MSE on forecasting tasks and up to 3.7\%
relative improvement in accuracy on classification tasks over the leading
baselines.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 15:02:50 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Luo",
"Dongsheng",
""
],
[
"Cheng",
"Wei",
""
],
[
"Wang",
"Yingheng",
""
],
[
"Xu",
"Dongkuan",
""
],
[
"Ni",
"Jingchao",
""
],
[
"Yu",
"Wenchao",
""
],
[
"Zhang",
"Xuchao",
""
],
[
"Liu",
"Yanchi",
""
],
[
"Chen",
"Yuncong",
""
],
[
"Chen",
"Haifeng",
""
],
[
"Zhang",
"Xiang",
""
]
] | TITLE: Time Series Contrastive Learning with Information-Aware Augmentations
ABSTRACT: Various contrastive learning approaches have been proposed in recent years
and achieve significant empirical success. While effective and prevalent,
contrastive learning has been less explored for time series data. A key
component of contrastive learning is to select appropriate augmentations
imposing some priors to construct feasible positive samples, such that an
encoder can be trained to learn robust and discriminative representations.
Unlike image and language domains where ``desired'' augmented samples can be
generated with the rule of thumb guided by prefabricated human priors, the
ad-hoc manual selection of time series augmentations is hindered by their
diverse and human-unrecognizable temporal structures. How to find the desired
augmentations of time series data that are meaningful for given contrastive
learning tasks and datasets remains an open question. In this work, we address
the problem by encouraging both high \textit{fidelity} and \textit{variety}
based upon information theory. A theoretical analysis leads to the criteria for
selecting feasible data augmentations. On top of that, we propose a new
contrastive learning approach with information-aware augmentations, InfoTS,
that adaptively selects optimal augmentations for time series representation
learning. Experiments on various datasets show highly competitive performance
with up to 12.0\% reduction in MSE on forecasting tasks and up to 3.7\%
relative improvement in accuracy on classification tasks over the leading
baselines.
| no_new_dataset | 0.947527 |
2303.11912 | Anirban Sarkar | Anirban Sarkar, Matthew Groth, Ian Mason, Tomotake Sasaki, Xavier Boix | Deephys: Deep Electrophysiology, Debugging Neural Networks under
Distribution Shifts | 12 pages, 8 figures | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep Neural Networks (DNNs) often fail in out-of-distribution scenarios. In
this paper, we introduce a tool to visualize and understand such failures. We
draw inspiration from concepts from neural electrophysiology, which are based
on inspecting the internal functioning of a neural networks by analyzing the
feature tuning and invariances of individual units. Deep Electrophysiology, in
short Deephys, provides insights of the DNN's failures in out-of-distribution
scenarios by comparative visualization of the neural activity in
in-distribution and out-of-distribution datasets. Deephys provides seamless
analyses of individual neurons, individual images, and a set of set of images
from a category, and it is capable of revealing failures due to the presence of
spurious features and novel features. We substantiate the validity of the
qualitative visualizations of Deephys thorough quantitative analyses using
convolutional and transformers architectures, in several datasets and
distribution shifts (namely, colored MNIST, CIFAR-10 and ImageNet).
| [
{
"version": "v1",
"created": "Fri, 17 Mar 2023 21:13:41 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Sarkar",
"Anirban",
""
],
[
"Groth",
"Matthew",
""
],
[
"Mason",
"Ian",
""
],
[
"Sasaki",
"Tomotake",
""
],
[
"Boix",
"Xavier",
""
]
] | TITLE: Deephys: Deep Electrophysiology, Debugging Neural Networks under
Distribution Shifts
ABSTRACT: Deep Neural Networks (DNNs) often fail in out-of-distribution scenarios. In
this paper, we introduce a tool to visualize and understand such failures. We
draw inspiration from concepts from neural electrophysiology, which are based
on inspecting the internal functioning of a neural networks by analyzing the
feature tuning and invariances of individual units. Deep Electrophysiology, in
short Deephys, provides insights of the DNN's failures in out-of-distribution
scenarios by comparative visualization of the neural activity in
in-distribution and out-of-distribution datasets. Deephys provides seamless
analyses of individual neurons, individual images, and a set of set of images
from a category, and it is capable of revealing failures due to the presence of
spurious features and novel features. We substantiate the validity of the
qualitative visualizations of Deephys thorough quantitative analyses using
convolutional and transformers architectures, in several datasets and
distribution shifts (namely, colored MNIST, CIFAR-10 and ImageNet).
| no_new_dataset | 0.947575 |
2303.11923 | Hancheng Ye | Hancheng Ye, Bo Zhang, Tao Chen, Jiayuan Fan, and Bin Wang | Performance-aware Approximation of Global Channel Pruning for Multitask
CNNs | Accepted for publication in T-PAMI, our code is available at
http://www.github.com/HankYe/PAGCP.git | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Global channel pruning (GCP) aims to remove a subset of channels (filters)
across different layers from a deep model without hurting the performance.
Previous works focus on either single task model pruning or simply adapting it
to multitask scenario, and still face the following problems when handling
multitask pruning: 1) Due to the task mismatch, a well-pruned backbone for
classification task focuses on preserving filters that can extract
category-sensitive information, causing filters that may be useful for other
tasks to be pruned during the backbone pruning stage; 2) For multitask
predictions, different filters within or between layers are more closely
related and interacted than that for single task prediction, making multitask
pruning more difficult. Therefore, aiming at multitask model compression, we
propose a Performance-Aware Global Channel Pruning (PAGCP) framework. We first
theoretically present the objective for achieving superior GCP, by considering
the joint saliency of filters from intra- and inter-layers. Then a sequentially
greedy pruning strategy is proposed to optimize the objective, where a
performance-aware oracle criterion is developed to evaluate sensitivity of
filters to each task and preserve the globally most task-related filters.
Experiments on several multitask datasets show that the proposed PAGCP can
reduce the FLOPs and parameters by over 60% with minor performance drop, and
achieves 1.2x$\sim$3.3x acceleration on both cloud and mobile platforms.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 15:15:21 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Ye",
"Hancheng",
""
],
[
"Zhang",
"Bo",
""
],
[
"Chen",
"Tao",
""
],
[
"Fan",
"Jiayuan",
""
],
[
"Wang",
"Bin",
""
]
] | TITLE: Performance-aware Approximation of Global Channel Pruning for Multitask
CNNs
ABSTRACT: Global channel pruning (GCP) aims to remove a subset of channels (filters)
across different layers from a deep model without hurting the performance.
Previous works focus on either single task model pruning or simply adapting it
to multitask scenario, and still face the following problems when handling
multitask pruning: 1) Due to the task mismatch, a well-pruned backbone for
classification task focuses on preserving filters that can extract
category-sensitive information, causing filters that may be useful for other
tasks to be pruned during the backbone pruning stage; 2) For multitask
predictions, different filters within or between layers are more closely
related and interacted than that for single task prediction, making multitask
pruning more difficult. Therefore, aiming at multitask model compression, we
propose a Performance-Aware Global Channel Pruning (PAGCP) framework. We first
theoretically present the objective for achieving superior GCP, by considering
the joint saliency of filters from intra- and inter-layers. Then a sequentially
greedy pruning strategy is proposed to optimize the objective, where a
performance-aware oracle criterion is developed to evaluate sensitivity of
filters to each task and preserve the globally most task-related filters.
Experiments on several multitask datasets show that the proposed PAGCP can
reduce the FLOPs and parameters by over 60% with minor performance drop, and
achieves 1.2x$\sim$3.3x acceleration on both cloud and mobile platforms.
| no_new_dataset | 0.943034 |
2303.11945 | Hongyan Ran | Hongyan Ran and Caiyan Jia | Unsupervised Cross-Domain Rumor Detection with Contrastive Learning and
Cross-Attention | null | null | null | null | cs.SI cs.AI cs.CL cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | Massive rumors usually appear along with breaking news or trending topics,
seriously hindering the truth. Existing rumor detection methods are mostly
focused on the same domain, and thus have poor performance in cross-domain
scenarios due to domain shift. In this work, we propose an end-to-end
instance-wise and prototype-wise contrastive learning model with a
cross-attention mechanism for cross-domain rumor detection. The model not only
performs cross-domain feature alignment but also enforces target samples to
align with the corresponding prototypes of a given source domain. Since target
labels in a target domain are unavailable, we use a clustering-based approach
with carefully initialized centers by a batch of source domain samples to
produce pseudo labels. Moreover, we use a cross-attention mechanism on a pair
of source data and target data with the same labels to learn domain-invariant
representations. Because the samples in a domain pair tend to express similar
semantic patterns, especially on the people's attitudes (e.g., supporting or
denying) towards the same category of rumors, the discrepancy between a pair of
the source domain and target domain will be decreased. We conduct experiments
on four groups of cross-domain datasets and show that our proposed model
achieves state-of-the-art performance.
| [
{
"version": "v1",
"created": "Mon, 20 Mar 2023 06:19:49 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Ran",
"Hongyan",
""
],
[
"Jia",
"Caiyan",
""
]
] | TITLE: Unsupervised Cross-Domain Rumor Detection with Contrastive Learning and
Cross-Attention
ABSTRACT: Massive rumors usually appear along with breaking news or trending topics,
seriously hindering the truth. Existing rumor detection methods are mostly
focused on the same domain, and thus have poor performance in cross-domain
scenarios due to domain shift. In this work, we propose an end-to-end
instance-wise and prototype-wise contrastive learning model with a
cross-attention mechanism for cross-domain rumor detection. The model not only
performs cross-domain feature alignment but also enforces target samples to
align with the corresponding prototypes of a given source domain. Since target
labels in a target domain are unavailable, we use a clustering-based approach
with carefully initialized centers by a batch of source domain samples to
produce pseudo labels. Moreover, we use a cross-attention mechanism on a pair
of source data and target data with the same labels to learn domain-invariant
representations. Because the samples in a domain pair tend to express similar
semantic patterns, especially on the people's attitudes (e.g., supporting or
denying) towards the same category of rumors, the discrepancy between a pair of
the source domain and target domain will be decreased. We conduct experiments
on four groups of cross-domain datasets and show that our proposed model
achieves state-of-the-art performance.
| no_new_dataset | 0.952706 |
2303.11950 | Jinshan Pan | Xiang Chen, Hao Li, Mingqiang Li, Jinshan Pan | Learning A Sparse Transformer Network for Effective Image Deraining | Accepted as a highlight paper in CVPR 2023 | CVPR 2023 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformers-based methods have achieved significant performance in image
deraining as they can model the non-local information which is vital for
high-quality image reconstruction. In this paper, we find that most existing
Transformers usually use all similarities of the tokens from the query-key
pairs for the feature aggregation. However, if the tokens from the query are
different from those of the key, the self-attention values estimated from these
tokens also involve in feature aggregation, which accordingly interferes with
the clear image restoration. To overcome this problem, we propose an effective
DeRaining network, Sparse Transformer (DRSformer) that can adaptively keep the
most useful self-attention values for feature aggregation so that the
aggregated features better facilitate high-quality image reconstruction.
Specifically, we develop a learnable top-k selection operator to adaptively
retain the most crucial attention scores from the keys for each query for
better feature aggregation. Simultaneously, as the naive feed-forward network
in Transformers does not model the multi-scale information that is important
for latent clear image restoration, we develop an effective mixed-scale
feed-forward network to generate better features for image deraining. To learn
an enriched set of hybrid features, which combines local context from CNN
operators, we equip our model with mixture of experts feature compensator to
present a cooperation refinement deraining scheme. Extensive experimental
results on the commonly used benchmarks demonstrate that the proposed method
achieves favorable performance against state-of-the-art approaches. The source
code and trained models are available at
https://github.com/cschenxiang/DRSformer.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 15:41:57 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Chen",
"Xiang",
""
],
[
"Li",
"Hao",
""
],
[
"Li",
"Mingqiang",
""
],
[
"Pan",
"Jinshan",
""
]
] | TITLE: Learning A Sparse Transformer Network for Effective Image Deraining
ABSTRACT: Transformers-based methods have achieved significant performance in image
deraining as they can model the non-local information which is vital for
high-quality image reconstruction. In this paper, we find that most existing
Transformers usually use all similarities of the tokens from the query-key
pairs for the feature aggregation. However, if the tokens from the query are
different from those of the key, the self-attention values estimated from these
tokens also involve in feature aggregation, which accordingly interferes with
the clear image restoration. To overcome this problem, we propose an effective
DeRaining network, Sparse Transformer (DRSformer) that can adaptively keep the
most useful self-attention values for feature aggregation so that the
aggregated features better facilitate high-quality image reconstruction.
Specifically, we develop a learnable top-k selection operator to adaptively
retain the most crucial attention scores from the keys for each query for
better feature aggregation. Simultaneously, as the naive feed-forward network
in Transformers does not model the multi-scale information that is important
for latent clear image restoration, we develop an effective mixed-scale
feed-forward network to generate better features for image deraining. To learn
an enriched set of hybrid features, which combines local context from CNN
operators, we equip our model with mixture of experts feature compensator to
present a cooperation refinement deraining scheme. Extensive experimental
results on the commonly used benchmarks demonstrate that the proposed method
achieves favorable performance against state-of-the-art approaches. The source
code and trained models are available at
https://github.com/cschenxiang/DRSformer.
| no_new_dataset | 0.946448 |
2303.12012 | Xiaoxu Meng | Xiaoxu Meng, Weikai Chen, Bo Yang | NeAT: Learning Neural Implicit Surfaces with Arbitrary Topologies from
Multi-view Images | null | null | null | null | cs.CV cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent progress in neural implicit functions has set new state-of-the-art in
reconstructing high-fidelity 3D shapes from a collection of images. However,
these approaches are limited to closed surfaces as they require the surface to
be represented by a signed distance field. In this paper, we propose NeAT, a
new neural rendering framework that can learn implicit surfaces with arbitrary
topologies from multi-view images. In particular, NeAT represents the 3D
surface as a level set of a signed distance function (SDF) with a validity
branch for estimating the surface existence probability at the query positions.
We also develop a novel neural volume rendering method, which uses SDF and
validity to calculate the volume opacity and avoids rendering points with low
validity. NeAT supports easy field-to-mesh conversion using the classic
Marching Cubes algorithm. Extensive experiments on DTU, MGN, and Deep Fashion
3D datasets indicate that our approach is able to faithfully reconstruct both
watertight and non-watertight surfaces. In particular, NeAT significantly
outperforms the state-of-the-art methods in the task of open surface
reconstruction both quantitatively and qualitatively.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 16:49:41 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Meng",
"Xiaoxu",
""
],
[
"Chen",
"Weikai",
""
],
[
"Yang",
"Bo",
""
]
] | TITLE: NeAT: Learning Neural Implicit Surfaces with Arbitrary Topologies from
Multi-view Images
ABSTRACT: Recent progress in neural implicit functions has set new state-of-the-art in
reconstructing high-fidelity 3D shapes from a collection of images. However,
these approaches are limited to closed surfaces as they require the surface to
be represented by a signed distance field. In this paper, we propose NeAT, a
new neural rendering framework that can learn implicit surfaces with arbitrary
topologies from multi-view images. In particular, NeAT represents the 3D
surface as a level set of a signed distance function (SDF) with a validity
branch for estimating the surface existence probability at the query positions.
We also develop a novel neural volume rendering method, which uses SDF and
validity to calculate the volume opacity and avoids rendering points with low
validity. NeAT supports easy field-to-mesh conversion using the classic
Marching Cubes algorithm. Extensive experiments on DTU, MGN, and Deep Fashion
3D datasets indicate that our approach is able to faithfully reconstruct both
watertight and non-watertight surfaces. In particular, NeAT significantly
outperforms the state-of-the-art methods in the task of open surface
reconstruction both quantitatively and qualitatively.
| no_new_dataset | 0.950503 |
2303.12038 | David Noever | Grant Rosario, David Noever | Grading Conversational Responses Of Chatbots | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by-sa/4.0/ | Chatbots have long been capable of answering basic questions and even
responding to obscure prompts, but recently their improvements have been far
more significant. Modern chatbots like Open AIs ChatGPT3 not only have the
ability to answer basic questions but can write code and movie scripts and
imitate well-known people. In this paper, we analyze ChatGPTs' responses to
various questions from a dataset of queries from the popular Quora forum. We
submitted sixty questions to ChatGPT and scored the answers based on three
industry-standard metrics for grading machine translation: BLEU, METEOR, and
ROUGE. These metrics allow us to compare the machine responses with the most
upvoted human answer to the same question to assess ChatGPT's ability to submit
a humanistic reply. The results showed that while the responses and translation
abilities of ChatGPT are remarkable, they still fall short of what a typical
human reaction would be.
| [
{
"version": "v1",
"created": "Wed, 1 Feb 2023 02:54:43 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Rosario",
"Grant",
""
],
[
"Noever",
"David",
""
]
] | TITLE: Grading Conversational Responses Of Chatbots
ABSTRACT: Chatbots have long been capable of answering basic questions and even
responding to obscure prompts, but recently their improvements have been far
more significant. Modern chatbots like Open AIs ChatGPT3 not only have the
ability to answer basic questions but can write code and movie scripts and
imitate well-known people. In this paper, we analyze ChatGPTs' responses to
various questions from a dataset of queries from the popular Quora forum. We
submitted sixty questions to ChatGPT and scored the answers based on three
industry-standard metrics for grading machine translation: BLEU, METEOR, and
ROUGE. These metrics allow us to compare the machine responses with the most
upvoted human answer to the same question to assess ChatGPT's ability to submit
a humanistic reply. The results showed that while the responses and translation
abilities of ChatGPT are remarkable, they still fall short of what a typical
human reaction would be.
| no_new_dataset | 0.934873 |
2303.12068 | Nicolas Dufour | Robin Courant, Maika Edberg, Nicolas Dufour and Vicky Kalogeiton | Machine Learning for Brain Disorders: Transformers and Visual
Transformers | To appear in O. Colliot (Ed.), Machine Learning for Brain Disorders,
Springer | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Transformers were initially introduced for natural language processing (NLP)
tasks, but fast they were adopted by most deep learning fields, including
computer vision. They measure the relationships between pairs of input tokens
(words in the case of text strings, parts of images for visual Transformers),
termed attention. The cost is exponential with the number of tokens. For image
classification, the most common Transformer Architecture uses only the
Transformer Encoder in order to transform the various input tokens. However,
there are also numerous other applications in which the decoder part of the
traditional Transformer Architecture is also used. Here, we first introduce the
Attention mechanism (Section 1), and then the Basic Transformer Block including
the Vision Transformer (Section 2). Next, we discuss some improvements of
visual Transformers to account for small datasets or less computation(Section
3). Finally, we introduce Visual Transformers applied to tasks other than image
classification, such as detection, segmentation, generation and training
without labels (Section 4) and other domains, such as video or multimodality
using text or audio data (Section 5).
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 17:57:33 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Courant",
"Robin",
""
],
[
"Edberg",
"Maika",
""
],
[
"Dufour",
"Nicolas",
""
],
[
"Kalogeiton",
"Vicky",
""
]
] | TITLE: Machine Learning for Brain Disorders: Transformers and Visual
Transformers
ABSTRACT: Transformers were initially introduced for natural language processing (NLP)
tasks, but fast they were adopted by most deep learning fields, including
computer vision. They measure the relationships between pairs of input tokens
(words in the case of text strings, parts of images for visual Transformers),
termed attention. The cost is exponential with the number of tokens. For image
classification, the most common Transformer Architecture uses only the
Transformer Encoder in order to transform the various input tokens. However,
there are also numerous other applications in which the decoder part of the
traditional Transformer Architecture is also used. Here, we first introduce the
Attention mechanism (Section 1), and then the Basic Transformer Block including
the Vision Transformer (Section 2). Next, we discuss some improvements of
visual Transformers to account for small datasets or less computation(Section
3). Finally, we introduce Visual Transformers applied to tasks other than image
classification, such as detection, segmentation, generation and training
without labels (Section 4) and other domains, such as video or multimodality
using text or audio data (Section 5).
| no_new_dataset | 0.951369 |
2303.12073 | Omkar Thawakar | Omkar Thawakar, Rao Muhammad Anwer, Jorma Laaksonen, Orly Reiner,
Mubarak Shah, Fahad Shahbaz Khan | 3D Mitochondria Instance Segmentation with Spatio-Temporal Transformers | 8 pages, 3 figures, 5 Tables, 2 page references | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Accurate 3D mitochondria instance segmentation in electron microscopy (EM) is
a challenging problem and serves as a prerequisite to empirically analyze their
distributions and morphology. Most existing approaches employ 3D convolutions
to obtain representative features. However, these convolution-based approaches
struggle to effectively capture long-range dependencies in the volume
mitochondria data, due to their limited local receptive field. To address this,
we propose a hybrid encoder-decoder framework based on a split spatio-temporal
attention module that efficiently computes spatial and temporal self-attentions
in parallel, which are later fused through a deformable convolution. Further,
we introduce a semantic foreground-background adversarial loss during training
that aids in delineating the region of mitochondria instances from the
background clutter. Our extensive experiments on three benchmarks, Lucchi,
MitoEM-R and MitoEM-H, reveal the benefits of the proposed contributions
achieving state-of-the-art results on all three datasets. Our code and models
are available at https://github.com/OmkarThawakar/STT-UNET.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 17:58:49 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Thawakar",
"Omkar",
""
],
[
"Anwer",
"Rao Muhammad",
""
],
[
"Laaksonen",
"Jorma",
""
],
[
"Reiner",
"Orly",
""
],
[
"Shah",
"Mubarak",
""
],
[
"Khan",
"Fahad Shahbaz",
""
]
] | TITLE: 3D Mitochondria Instance Segmentation with Spatio-Temporal Transformers
ABSTRACT: Accurate 3D mitochondria instance segmentation in electron microscopy (EM) is
a challenging problem and serves as a prerequisite to empirically analyze their
distributions and morphology. Most existing approaches employ 3D convolutions
to obtain representative features. However, these convolution-based approaches
struggle to effectively capture long-range dependencies in the volume
mitochondria data, due to their limited local receptive field. To address this,
we propose a hybrid encoder-decoder framework based on a split spatio-temporal
attention module that efficiently computes spatial and temporal self-attentions
in parallel, which are later fused through a deformable convolution. Further,
we introduce a semantic foreground-background adversarial loss during training
that aids in delineating the region of mitochondria instances from the
background clutter. Our extensive experiments on three benchmarks, Lucchi,
MitoEM-R and MitoEM-H, reveal the benefits of the proposed contributions
achieving state-of-the-art results on all three datasets. Our code and models
are available at https://github.com/OmkarThawakar/STT-UNET.
| no_new_dataset | 0.943919 |
2303.12078 | Fangyun Wei | Kun Yan, Xiao Li, Fangyun Wei, Jinglu Wang, Chenbin Zhang, Ping Wang,
Yan Lu | Two-shot Video Object Segmentation | Accepted by CVPR 2023. Code and models are available at
https://github.com/yk-pku/Two-shot-Video-Object-Segmentation | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Previous works on video object segmentation (VOS) are trained on densely
annotated videos. Nevertheless, acquiring annotations in pixel level is
expensive and time-consuming. In this work, we demonstrate the feasibility of
training a satisfactory VOS model on sparsely annotated videos-we merely
require two labeled frames per training video while the performance is
sustained. We term this novel training paradigm as two-shot video object
segmentation, or two-shot VOS for short. The underlying idea is to generate
pseudo labels for unlabeled frames during training and to optimize the model on
the combination of labeled and pseudo-labeled data. Our approach is extremely
simple and can be applied to a majority of existing frameworks. We first
pre-train a VOS model on sparsely annotated videos in a semi-supervised manner,
with the first frame always being a labeled one. Then, we adopt the pre-trained
VOS model to generate pseudo labels for all unlabeled frames, which are
subsequently stored in a pseudo-label bank. Finally, we retrain a VOS model on
both labeled and pseudo-labeled data without any restrictions on the first
frame. For the first time, we present a general way to train VOS models on
two-shot VOS datasets. By using 7.3% and 2.9% labeled data of YouTube-VOS and
DAVIS benchmarks, our approach achieves comparable results in contrast to the
counterparts trained on fully labeled set. Code and models are available at
https://github.com/yk-pku/Two-shot-Video-Object-Segmentation.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 17:59:56 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Yan",
"Kun",
""
],
[
"Li",
"Xiao",
""
],
[
"Wei",
"Fangyun",
""
],
[
"Wang",
"Jinglu",
""
],
[
"Zhang",
"Chenbin",
""
],
[
"Wang",
"Ping",
""
],
[
"Lu",
"Yan",
""
]
] | TITLE: Two-shot Video Object Segmentation
ABSTRACT: Previous works on video object segmentation (VOS) are trained on densely
annotated videos. Nevertheless, acquiring annotations in pixel level is
expensive and time-consuming. In this work, we demonstrate the feasibility of
training a satisfactory VOS model on sparsely annotated videos-we merely
require two labeled frames per training video while the performance is
sustained. We term this novel training paradigm as two-shot video object
segmentation, or two-shot VOS for short. The underlying idea is to generate
pseudo labels for unlabeled frames during training and to optimize the model on
the combination of labeled and pseudo-labeled data. Our approach is extremely
simple and can be applied to a majority of existing frameworks. We first
pre-train a VOS model on sparsely annotated videos in a semi-supervised manner,
with the first frame always being a labeled one. Then, we adopt the pre-trained
VOS model to generate pseudo labels for all unlabeled frames, which are
subsequently stored in a pseudo-label bank. Finally, we retrain a VOS model on
both labeled and pseudo-labeled data without any restrictions on the first
frame. For the first time, we present a general way to train VOS models on
two-shot VOS datasets. By using 7.3% and 2.9% labeled data of YouTube-VOS and
DAVIS benchmarks, our approach achieves comparable results in contrast to the
counterparts trained on fully labeled set. Code and models are available at
https://github.com/yk-pku/Two-shot-Video-Object-Segmentation.
| no_new_dataset | 0.95275 |
2303.12079 | Dongdong Chen | Junke Wang and Dongdong Chen and Zuxuan Wu and Chong Luo and Xiyang
Dai and Lu Yuan and Yu-Gang Jiang | OmniTracker: Unifying Object Tracking by Tracking-with-Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object tracking (OT) aims to estimate the positions of target objects in a
video sequence. Depending on whether the initial states of target objects are
specified by provided annotations in the first frame or the categories, OT
could be classified as instance tracking (e.g., SOT and VOS) and category
tracking (e.g., MOT, MOTS, and VIS) tasks. Combing the advantages of the best
practices developed in both communities, we propose a novel
tracking-with-detection paradigm, where tracking supplements appearance priors
for detection and detection provides tracking with candidate bounding boxes for
association. Equipped with such a design, a unified tracking model,
OmniTracker, is further presented to resolve all the tracking tasks with a
fully shared network architecture, model weights, and inference pipeline.
Extensive experiments on 7 tracking datasets, including LaSOT, TrackingNet,
DAVIS16-17, MOT17, MOTS20, and YTVIS19, demonstrate that OmniTracker achieves
on-par or even better results than both task-specific and unified tracking
models.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 17:59:57 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Wang",
"Junke",
""
],
[
"Chen",
"Dongdong",
""
],
[
"Wu",
"Zuxuan",
""
],
[
"Luo",
"Chong",
""
],
[
"Dai",
"Xiyang",
""
],
[
"Yuan",
"Lu",
""
],
[
"Jiang",
"Yu-Gang",
""
]
] | TITLE: OmniTracker: Unifying Object Tracking by Tracking-with-Detection
ABSTRACT: Object tracking (OT) aims to estimate the positions of target objects in a
video sequence. Depending on whether the initial states of target objects are
specified by provided annotations in the first frame or the categories, OT
could be classified as instance tracking (e.g., SOT and VOS) and category
tracking (e.g., MOT, MOTS, and VIS) tasks. Combing the advantages of the best
practices developed in both communities, we propose a novel
tracking-with-detection paradigm, where tracking supplements appearance priors
for detection and detection provides tracking with candidate bounding boxes for
association. Equipped with such a design, a unified tracking model,
OmniTracker, is further presented to resolve all the tracking tasks with a
fully shared network architecture, model weights, and inference pipeline.
Extensive experiments on 7 tracking datasets, including LaSOT, TrackingNet,
DAVIS16-17, MOT17, MOTS20, and YTVIS19, demonstrate that OmniTracker achieves
on-par or even better results than both task-specific and unified tracking
models.
| no_new_dataset | 0.950319 |
2303.12080 | Fangyun Wei | Ronglai Zuo, Fangyun Wei, Brian Mak | Natural Language-Assisted Sign Language Recognition | Accepted by CVPR 2023. Codes are available at
https://github.com/FangyunWei/SLRT | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sign languages are visual languages which convey information by signers'
handshape, facial expression, body movement, and so forth. Due to the inherent
restriction of combinations of these visual ingredients, there exist a
significant number of visually indistinguishable signs (VISigns) in sign
languages, which limits the recognition capacity of vision neural networks. To
mitigate the problem, we propose the Natural Language-Assisted Sign Language
Recognition (NLA-SLR) framework, which exploits semantic information contained
in glosses (sign labels). First, for VISigns with similar semantic meanings, we
propose language-aware label smoothing by generating soft labels for each
training sign whose smoothing weights are computed from the normalized semantic
similarities among the glosses to ease training. Second, for VISigns with
distinct semantic meanings, we present an inter-modality mixup technique which
blends vision and gloss features to further maximize the separability of
different signs under the supervision of blended labels. Besides, we also
introduce a novel backbone, video-keypoint network, which not only models both
RGB videos and human body keypoints but also derives knowledge from sign videos
of different temporal receptive fields. Empirically, our method achieves
state-of-the-art performance on three widely-adopted benchmarks: MSASL, WLASL,
and NMFs-CSL. Codes are available at https://github.com/FangyunWei/SLRT.
| [
{
"version": "v1",
"created": "Tue, 21 Mar 2023 17:59:57 GMT"
}
] | 2023-03-22T00:00:00 | [
[
"Zuo",
"Ronglai",
""
],
[
"Wei",
"Fangyun",
""
],
[
"Mak",
"Brian",
""
]
] | TITLE: Natural Language-Assisted Sign Language Recognition
ABSTRACT: Sign languages are visual languages which convey information by signers'
handshape, facial expression, body movement, and so forth. Due to the inherent
restriction of combinations of these visual ingredients, there exist a
significant number of visually indistinguishable signs (VISigns) in sign
languages, which limits the recognition capacity of vision neural networks. To
mitigate the problem, we propose the Natural Language-Assisted Sign Language
Recognition (NLA-SLR) framework, which exploits semantic information contained
in glosses (sign labels). First, for VISigns with similar semantic meanings, we
propose language-aware label smoothing by generating soft labels for each
training sign whose smoothing weights are computed from the normalized semantic
similarities among the glosses to ease training. Second, for VISigns with
distinct semantic meanings, we present an inter-modality mixup technique which
blends vision and gloss features to further maximize the separability of
different signs under the supervision of blended labels. Besides, we also
introduce a novel backbone, video-keypoint network, which not only models both
RGB videos and human body keypoints but also derives knowledge from sign videos
of different temporal receptive fields. Empirically, our method achieves
state-of-the-art performance on three widely-adopted benchmarks: MSASL, WLASL,
and NMFs-CSL. Codes are available at https://github.com/FangyunWei/SLRT.
| no_new_dataset | 0.951504 |
1910.13140 | Neo Christopher Chung < | Lennart Brocki, Neo Christopher Chung | Concept Saliency Maps to Visualize Relevant Features in Deep Generative
Models | 18th IEEE International Conference on Machine Learning and
Applications (ICMLA) | null | null | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evaluating, explaining, and visualizing high-level concepts in generative
models, such as variational autoencoders (VAEs), is challenging in part due to
a lack of known prediction classes that are required to generate saliency maps
in supervised learning. While saliency maps may help identify relevant features
(e.g., pixels) in the input for classification tasks of deep neural networks,
similar frameworks are understudied in unsupervised learning. Therefore, we
introduce a new method of obtaining saliency maps for latent representations of
known or novel high-level concepts, often called concept vectors in generative
models. Concept scores, analogous to class scores in classification tasks, are
defined as dot products between concept vectors and encoded input data, which
can be readily used to compute the gradients. The resulting concept saliency
maps are shown to highlight input features deemed important for high-level
concepts. Our method is applied to the VAE's latent space of CelebA dataset in
which known attributes such as "smiles" and "hats" are used to elucidate
relevant facial features. Furthermore, our application to spatial
transcriptomic (ST) data of a mouse olfactory bulb demonstrates the potential
of latent representations of morphological layers and molecular features in
advancing our understanding of complex biological systems. By extending the
popular method of saliency maps to generative models, the proposed concept
saliency maps help improve interpretability of latent variable models in deep
learning.
Codes to reproduce and to implement concept saliency maps:
https://github.com/lenbrocki/concept-saliency-maps
| [
{
"version": "v1",
"created": "Tue, 29 Oct 2019 09:15:25 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Brocki",
"Lennart",
""
],
[
"Chung",
"Neo Christopher",
""
]
] | TITLE: Concept Saliency Maps to Visualize Relevant Features in Deep Generative
Models
ABSTRACT: Evaluating, explaining, and visualizing high-level concepts in generative
models, such as variational autoencoders (VAEs), is challenging in part due to
a lack of known prediction classes that are required to generate saliency maps
in supervised learning. While saliency maps may help identify relevant features
(e.g., pixels) in the input for classification tasks of deep neural networks,
similar frameworks are understudied in unsupervised learning. Therefore, we
introduce a new method of obtaining saliency maps for latent representations of
known or novel high-level concepts, often called concept vectors in generative
models. Concept scores, analogous to class scores in classification tasks, are
defined as dot products between concept vectors and encoded input data, which
can be readily used to compute the gradients. The resulting concept saliency
maps are shown to highlight input features deemed important for high-level
concepts. Our method is applied to the VAE's latent space of CelebA dataset in
which known attributes such as "smiles" and "hats" are used to elucidate
relevant facial features. Furthermore, our application to spatial
transcriptomic (ST) data of a mouse olfactory bulb demonstrates the potential
of latent representations of morphological layers and molecular features in
advancing our understanding of complex biological systems. By extending the
popular method of saliency maps to generative models, the proposed concept
saliency maps help improve interpretability of latent variable models in deep
learning.
Codes to reproduce and to implement concept saliency maps:
https://github.com/lenbrocki/concept-saliency-maps
| no_new_dataset | 0.947527 |
2009.03107 | Tong Liu | Tong Liu, Roberto Amadini, Jacopo Mauro, Maurizio Gabbrielli | sunny-as2: Enhancing SUNNY for Algorithm Selection | null | Journal of Artificial Intelligence Research. 72 (2021) 329-376 | 10.1613/jair.1.13116 | null | cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | SUNNY is an Algorithm Selection (AS) technique originally tailored for
Constraint Programming (CP). SUNNY enables to schedule, from a portfolio of
solvers, a subset of solvers to be run on a given CP problem. This approach has
proved to be effective for CP problems, and its parallel version won many gold
medals in the Open category of the MiniZinc Challenge -- the yearly
international competition for CP solvers. In 2015, the ASlib benchmarks were
released for comparing AS systems coming from disparate fields (e.g., ASP, QBF,
and SAT) and SUNNY was extended to deal with generic AS problems. This led to
the development of sunny-as2, an algorithm selector based on SUNNY for ASlib
scenarios. A preliminary version of sunny-as2 was submitted to the Open
Algorithm Selection Challenge (OASC) in 2017, where it turned out to be the
best approach for the runtime minimization of decision problems. In this work,
we present the technical advancements of sunny-as2, including: (i)
wrapper-based feature selection; (ii) a training approach combining feature
selection and neighbourhood size configuration; (iii) the application of nested
cross-validation. We show how sunny-as2 performance varies depending on the
considered AS scenarios, and we discuss its strengths and weaknesses. Finally,
we also show how sunny-as2 improves on its preliminary version submitted to
OASC.
| [
{
"version": "v1",
"created": "Mon, 7 Sep 2020 13:55:45 GMT"
},
{
"version": "v2",
"created": "Sun, 26 Sep 2021 08:53:15 GMT"
},
{
"version": "v3",
"created": "Tue, 12 Oct 2021 09:49:31 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Liu",
"Tong",
""
],
[
"Amadini",
"Roberto",
""
],
[
"Mauro",
"Jacopo",
""
],
[
"Gabbrielli",
"Maurizio",
""
]
] | TITLE: sunny-as2: Enhancing SUNNY for Algorithm Selection
ABSTRACT: SUNNY is an Algorithm Selection (AS) technique originally tailored for
Constraint Programming (CP). SUNNY enables to schedule, from a portfolio of
solvers, a subset of solvers to be run on a given CP problem. This approach has
proved to be effective for CP problems, and its parallel version won many gold
medals in the Open category of the MiniZinc Challenge -- the yearly
international competition for CP solvers. In 2015, the ASlib benchmarks were
released for comparing AS systems coming from disparate fields (e.g., ASP, QBF,
and SAT) and SUNNY was extended to deal with generic AS problems. This led to
the development of sunny-as2, an algorithm selector based on SUNNY for ASlib
scenarios. A preliminary version of sunny-as2 was submitted to the Open
Algorithm Selection Challenge (OASC) in 2017, where it turned out to be the
best approach for the runtime minimization of decision problems. In this work,
we present the technical advancements of sunny-as2, including: (i)
wrapper-based feature selection; (ii) a training approach combining feature
selection and neighbourhood size configuration; (iii) the application of nested
cross-validation. We show how sunny-as2 performance varies depending on the
considered AS scenarios, and we discuss its strengths and weaknesses. Finally,
we also show how sunny-as2 improves on its preliminary version submitted to
OASC.
| no_new_dataset | 0.943712 |
2011.03213 | Jun Ma | Jun Ma, Zilong Cheng, Wenxin Wang, Abdullah Al Mamun, Clarence W. de
Silva, Tong Heng Lee | Data-Driven Predictive Control Towards Multi-Agent Motion Planning With
Non-Parametric Closed-Loop Behavior Learning | null | null | null | null | cs.MA cs.RO cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In many specific scenarios, accurate and effective system identification is a
commonly encountered challenge in the model predictive control (MPC)
formulation. As a consequence, the overall system performance could be
significantly weakened in outcome when the traditional MPC algorithm is adopted
under those circumstances when such accuracy is lacking. This paper
investigates a non-parametric closed-loop behavior learning method for
multi-agent motion planning, which underpins a data-driven predictive control
framework. Utilizing an innovative methodology with closed-loop input/output
measurements of the unknown system, the behavior of the system is learned based
on the collected dataset, and thus the constructed non-parametric predictive
model can be used to determine the optimal control actions. This non-parametric
predictive control framework alleviates the heavy computational burden commonly
encountered in the optimization procedures typically in alternate methodologies
requiring open-loop input/output measurement data collection and parametric
system identification. The proposed data-driven approach is also shown to
preserve good robustness properties. Finally, a multi-UAV system is used to
demonstrate the highly effective outcome of this promising development.
| [
{
"version": "v1",
"created": "Fri, 6 Nov 2020 07:16:18 GMT"
},
{
"version": "v2",
"created": "Mon, 26 Apr 2021 04:56:58 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 17:02:00 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Ma",
"Jun",
""
],
[
"Cheng",
"Zilong",
""
],
[
"Wang",
"Wenxin",
""
],
[
"Mamun",
"Abdullah Al",
""
],
[
"de Silva",
"Clarence W.",
""
],
[
"Lee",
"Tong Heng",
""
]
] | TITLE: Data-Driven Predictive Control Towards Multi-Agent Motion Planning With
Non-Parametric Closed-Loop Behavior Learning
ABSTRACT: In many specific scenarios, accurate and effective system identification is a
commonly encountered challenge in the model predictive control (MPC)
formulation. As a consequence, the overall system performance could be
significantly weakened in outcome when the traditional MPC algorithm is adopted
under those circumstances when such accuracy is lacking. This paper
investigates a non-parametric closed-loop behavior learning method for
multi-agent motion planning, which underpins a data-driven predictive control
framework. Utilizing an innovative methodology with closed-loop input/output
measurements of the unknown system, the behavior of the system is learned based
on the collected dataset, and thus the constructed non-parametric predictive
model can be used to determine the optimal control actions. This non-parametric
predictive control framework alleviates the heavy computational burden commonly
encountered in the optimization procedures typically in alternate methodologies
requiring open-loop input/output measurement data collection and parametric
system identification. The proposed data-driven approach is also shown to
preserve good robustness properties. Finally, a multi-UAV system is used to
demonstrate the highly effective outcome of this promising development.
| no_new_dataset | 0.944944 |
2106.03847 | Omri Avrahami | Omri Avrahami, Dani Lischinski, Ohad Fried | GAN Cocktail: mixing GANs without dataset access | ECCV 2022. Project page is available at:
https://omriavrahami.com/GAN-cocktail-page/ | null | 10.1007/978-3-031-20050-2_13 | null | cs.LG cs.CV cs.GR cs.NE | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Today's generative models are capable of synthesizing high-fidelity images,
but each model specializes on a specific target domain. This raises the need
for model merging: combining two or more pretrained generative models into a
single unified one. In this work we tackle the problem of model merging, given
two constraints that often come up in the real world: (1) no access to the
original training data, and (2) without increasing the size of the neural
network. To the best of our knowledge, model merging under these constraints
has not been studied thus far. We propose a novel, two-stage solution. In the
first stage, we transform the weights of all the models to the same parameter
space by a technique we term model rooting. In the second stage, we merge the
rooted models by averaging their weights and fine-tuning them for each specific
domain, using only data generated by the original trained models. We
demonstrate that our approach is superior to baseline methods and to existing
transfer learning techniques, and investigate several applications.
| [
{
"version": "v1",
"created": "Mon, 7 Jun 2021 17:59:04 GMT"
},
{
"version": "v2",
"created": "Mon, 11 Jul 2022 08:37:52 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Avrahami",
"Omri",
""
],
[
"Lischinski",
"Dani",
""
],
[
"Fried",
"Ohad",
""
]
] | TITLE: GAN Cocktail: mixing GANs without dataset access
ABSTRACT: Today's generative models are capable of synthesizing high-fidelity images,
but each model specializes on a specific target domain. This raises the need
for model merging: combining two or more pretrained generative models into a
single unified one. In this work we tackle the problem of model merging, given
two constraints that often come up in the real world: (1) no access to the
original training data, and (2) without increasing the size of the neural
network. To the best of our knowledge, model merging under these constraints
has not been studied thus far. We propose a novel, two-stage solution. In the
first stage, we transform the weights of all the models to the same parameter
space by a technique we term model rooting. In the second stage, we merge the
rooted models by averaging their weights and fine-tuning them for each specific
domain, using only data generated by the original trained models. We
demonstrate that our approach is superior to baseline methods and to existing
transfer learning techniques, and investigate several applications.
| no_new_dataset | 0.949576 |
2108.06600 | Binghao Liu | Qi Zhao, Binghao Liu, Shuchang Lyu and Huojin Chen | A Self-Distillation Embedded Supervised Affinity Attention Model for
Few-Shot Segmentation | 14 pages, 12 figures | null | 10.1109/TCDS.2023.3251371 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot segmentation focuses on the generalization of models to segment
unseen object with limited annotated samples. However, existing approaches
still face two main challenges. First, huge feature distinction between support
and query images causes knowledge transferring barrier, which harms the
segmentation performance. Second, limited support prototypes cannot adequately
represent features of support objects, hard to guide high-quality query
segmentation. To deal with the above two issues, we propose self-distillation
embedded supervised affinity attention model to improve the performance of
few-shot segmentation task. Specifically, the self-distillation guided
prototype module uses self-distillation to align the features of support and
query. The supervised affinity attention module generates high-quality query
attention map to provide sufficient object information. Extensive experiments
prove that our model significantly improves the performance compared to
existing methods. Comprehensive ablation experiments and visualization studies
also show the significant effect of our method on few-shot segmentation task.
On COCO-20i dataset, we achieve new state-of-the-art results. Training code and
pretrained models are available at https://github.com/cv516Buaa/SD-AANet.
| [
{
"version": "v1",
"created": "Sat, 14 Aug 2021 18:16:12 GMT"
},
{
"version": "v2",
"created": "Fri, 8 Apr 2022 18:06:31 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 14:53:07 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhao",
"Qi",
""
],
[
"Liu",
"Binghao",
""
],
[
"Lyu",
"Shuchang",
""
],
[
"Chen",
"Huojin",
""
]
] | TITLE: A Self-Distillation Embedded Supervised Affinity Attention Model for
Few-Shot Segmentation
ABSTRACT: Few-shot segmentation focuses on the generalization of models to segment
unseen object with limited annotated samples. However, existing approaches
still face two main challenges. First, huge feature distinction between support
and query images causes knowledge transferring barrier, which harms the
segmentation performance. Second, limited support prototypes cannot adequately
represent features of support objects, hard to guide high-quality query
segmentation. To deal with the above two issues, we propose self-distillation
embedded supervised affinity attention model to improve the performance of
few-shot segmentation task. Specifically, the self-distillation guided
prototype module uses self-distillation to align the features of support and
query. The supervised affinity attention module generates high-quality query
attention map to provide sufficient object information. Extensive experiments
prove that our model significantly improves the performance compared to
existing methods. Comprehensive ablation experiments and visualization studies
also show the significant effect of our method on few-shot segmentation task.
On COCO-20i dataset, we achieve new state-of-the-art results. Training code and
pretrained models are available at https://github.com/cv516Buaa/SD-AANet.
| no_new_dataset | 0.947478 |
2109.12218 | Jake Grigsby | Jake Grigsby, Zhe Wang, Nam Nguyen, Yanjun Qi | Long-Range Transformers for Dynamic Spatiotemporal Forecasting | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Multivariate time series forecasting focuses on predicting future values
based on historical context. State-of-the-art sequence-to-sequence models rely
on neural attention between timesteps, which allows for temporal learning but
fails to consider distinct spatial relationships between variables. In
contrast, methods based on graph neural networks explicitly model variable
relationships. However, these methods often rely on predefined graphs that
cannot change over time and perform separate spatial and temporal updates
without establishing direct connections between each variable at every
timestep. Our work addresses these problems by translating multivariate
forecasting into a "spatiotemporal sequence" formulation where each Transformer
input token represents the value of a single variable at a given time.
Long-Range Transformers can then learn interactions between space, time, and
value information jointly along this extended sequence. Our method, which we
call Spacetimeformer, achieves competitive results on benchmarks from traffic
forecasting to electricity demand and weather prediction while learning
spatiotemporal relationships purely from data.
| [
{
"version": "v1",
"created": "Fri, 24 Sep 2021 22:11:46 GMT"
},
{
"version": "v2",
"created": "Thu, 19 May 2022 18:46:35 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 03:03:25 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Grigsby",
"Jake",
""
],
[
"Wang",
"Zhe",
""
],
[
"Nguyen",
"Nam",
""
],
[
"Qi",
"Yanjun",
""
]
] | TITLE: Long-Range Transformers for Dynamic Spatiotemporal Forecasting
ABSTRACT: Multivariate time series forecasting focuses on predicting future values
based on historical context. State-of-the-art sequence-to-sequence models rely
on neural attention between timesteps, which allows for temporal learning but
fails to consider distinct spatial relationships between variables. In
contrast, methods based on graph neural networks explicitly model variable
relationships. However, these methods often rely on predefined graphs that
cannot change over time and perform separate spatial and temporal updates
without establishing direct connections between each variable at every
timestep. Our work addresses these problems by translating multivariate
forecasting into a "spatiotemporal sequence" formulation where each Transformer
input token represents the value of a single variable at a given time.
Long-Range Transformers can then learn interactions between space, time, and
value information jointly along this extended sequence. Our method, which we
call Spacetimeformer, achieves competitive results on benchmarks from traffic
forecasting to electricity demand and weather prediction while learning
spatiotemporal relationships purely from data.
| no_new_dataset | 0.945349 |
2110.04845 | Mohamed Abdalla | Mohamed Abdalla, Krishnapriya Vishnubhotla, Saif M. Mohammad | What Makes Sentences Semantically Related: A Textual Relatedness Dataset
and Empirical Study | Accepted to EACL 2023; Our dataset, data statement, and annotation
questionnaire can be found at: https://doi.org/10.5281/zenodo.7599667 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The degree of semantic relatedness of two units of language has long been
considered fundamental to understanding meaning. Additionally, automatically
determining relatedness has many applications such as question answering and
summarization. However, prior NLP work has largely focused on semantic
similarity, a subset of relatedness, because of a lack of relatedness datasets.
In this paper, we introduce a dataset for Semantic Textual Relatedness,
STR-2022, that has 5,500 English sentence pairs manually annotated using a
comparative annotation framework, resulting in fine-grained scores. We show
that human intuition regarding relatedness of sentence pairs is highly
reliable, with a repeat annotation correlation of 0.84. We use the dataset to
explore questions on what makes sentences semantically related. We also show
the utility of STR-2022 for evaluating automatic methods of sentence
representation and for various downstream NLP tasks.
Our dataset, data statement, and annotation questionnaire can be found at:
https://doi.org/10.5281/zenodo.7599667
| [
{
"version": "v1",
"created": "Sun, 10 Oct 2021 16:23:54 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Oct 2022 16:26:50 GMT"
},
{
"version": "v3",
"created": "Thu, 9 Feb 2023 11:39:35 GMT"
},
{
"version": "v4",
"created": "Mon, 20 Mar 2023 13:34:47 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Abdalla",
"Mohamed",
""
],
[
"Vishnubhotla",
"Krishnapriya",
""
],
[
"Mohammad",
"Saif M.",
""
]
] | TITLE: What Makes Sentences Semantically Related: A Textual Relatedness Dataset
and Empirical Study
ABSTRACT: The degree of semantic relatedness of two units of language has long been
considered fundamental to understanding meaning. Additionally, automatically
determining relatedness has many applications such as question answering and
summarization. However, prior NLP work has largely focused on semantic
similarity, a subset of relatedness, because of a lack of relatedness datasets.
In this paper, we introduce a dataset for Semantic Textual Relatedness,
STR-2022, that has 5,500 English sentence pairs manually annotated using a
comparative annotation framework, resulting in fine-grained scores. We show
that human intuition regarding relatedness of sentence pairs is highly
reliable, with a repeat annotation correlation of 0.84. We use the dataset to
explore questions on what makes sentences semantically related. We also show
the utility of STR-2022 for evaluating automatic methods of sentence
representation and for various downstream NLP tasks.
Our dataset, data statement, and annotation questionnaire can be found at:
https://doi.org/10.5281/zenodo.7599667
| new_dataset | 0.963161 |
2111.01460 | Alexander Terenin | No\'emie Jaquier, Viacheslav Borovitskiy, Andrei Smolensky, Alexander
Terenin, Tamim Asfour, Leonel Rozo | Geometry-aware Bayesian Optimization in Robotics using Riemannian
Mat\'ern Kernels | Source code: https://github.com/NoemieJaquier/MaternGaBO, Video:
https://youtu.be/6awfFRqP7wA | Conference on Robot Learning, 2021 | null | null | cs.RO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bayesian optimization is a data-efficient technique which can be used for
control parameter tuning, parametric policy adaptation, and structure design in
robotics. Many of these problems require optimization of functions defined on
non-Euclidean domains like spheres, rotation groups, or spaces of
positive-definite matrices. To do so, one must place a Gaussian process prior,
or equivalently define a kernel, on the space of interest. Effective kernels
typically reflect the geometry of the spaces they are defined on, but designing
them is generally non-trivial. Recent work on the Riemannian Mat\'ern kernels,
based on stochastic partial differential equations and spectral theory of the
Laplace-Beltrami operator, offers promising avenues towards constructing such
geometry-aware kernels. In this paper, we study techniques for implementing
these kernels on manifolds of interest in robotics, demonstrate their
performance on a set of artificial benchmark functions, and illustrate
geometry-aware Bayesian optimization for a variety of robotic applications,
covering orientation control, manipulability optimization, and motion planning,
while showing its improved performance.
| [
{
"version": "v1",
"created": "Tue, 2 Nov 2021 09:47:22 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 02:19:29 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Jaquier",
"Noémie",
""
],
[
"Borovitskiy",
"Viacheslav",
""
],
[
"Smolensky",
"Andrei",
""
],
[
"Terenin",
"Alexander",
""
],
[
"Asfour",
"Tamim",
""
],
[
"Rozo",
"Leonel",
""
]
] | TITLE: Geometry-aware Bayesian Optimization in Robotics using Riemannian
Mat\'ern Kernels
ABSTRACT: Bayesian optimization is a data-efficient technique which can be used for
control parameter tuning, parametric policy adaptation, and structure design in
robotics. Many of these problems require optimization of functions defined on
non-Euclidean domains like spheres, rotation groups, or spaces of
positive-definite matrices. To do so, one must place a Gaussian process prior,
or equivalently define a kernel, on the space of interest. Effective kernels
typically reflect the geometry of the spaces they are defined on, but designing
them is generally non-trivial. Recent work on the Riemannian Mat\'ern kernels,
based on stochastic partial differential equations and spectral theory of the
Laplace-Beltrami operator, offers promising avenues towards constructing such
geometry-aware kernels. In this paper, we study techniques for implementing
these kernels on manifolds of interest in robotics, demonstrate their
performance on a set of artificial benchmark functions, and illustrate
geometry-aware Bayesian optimization for a variety of robotic applications,
covering orientation control, manipulability optimization, and motion planning,
while showing its improved performance.
| no_new_dataset | 0.95222 |
2112.03073 | Shirong Shen | Shirong Shen and Zhen Li and Guilin Qi | Active Learning for Event Extraction with Memory-based Loss Prediction
Model | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Event extraction (EE) plays an important role in many industrial application
scenarios, and high-quality EE methods require a large amount of manual
annotation data to train supervised learning models. However, the cost of
obtaining annotation data is very high, especially for annotation of domain
events, which requires the participation of experts from corresponding domain.
So we introduce active learning (AL) technology to reduce the cost of event
annotation. But the existing AL methods have two main problems, which make them
not well used for event extraction. Firstly, the existing pool-based selection
strategies have limitations in terms of computational cost and sample validity.
Secondly, the existing evaluation of sample importance lacks the use of local
sample information. In this paper, we present a novel deep AL method for EE. We
propose a batch-based selection strategy and a Memory-Based Loss Prediction
model (MBLP) to select unlabeled samples efficiently. During the selection
process, we use an internal-external sample loss ranking method to evaluate the
sample importance by using local information. Finally, we propose a delayed
training strategy to train the MBLP model. Extensive experiments are performed
on three domain datasets, and our method outperforms other state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Fri, 26 Nov 2021 07:58:11 GMT"
},
{
"version": "v2",
"created": "Thu, 9 Mar 2023 09:22:29 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 03:55:06 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Shen",
"Shirong",
""
],
[
"Li",
"Zhen",
""
],
[
"Qi",
"Guilin",
""
]
] | TITLE: Active Learning for Event Extraction with Memory-based Loss Prediction
Model
ABSTRACT: Event extraction (EE) plays an important role in many industrial application
scenarios, and high-quality EE methods require a large amount of manual
annotation data to train supervised learning models. However, the cost of
obtaining annotation data is very high, especially for annotation of domain
events, which requires the participation of experts from corresponding domain.
So we introduce active learning (AL) technology to reduce the cost of event
annotation. But the existing AL methods have two main problems, which make them
not well used for event extraction. Firstly, the existing pool-based selection
strategies have limitations in terms of computational cost and sample validity.
Secondly, the existing evaluation of sample importance lacks the use of local
sample information. In this paper, we present a novel deep AL method for EE. We
propose a batch-based selection strategy and a Memory-Based Loss Prediction
model (MBLP) to select unlabeled samples efficiently. During the selection
process, we use an internal-external sample loss ranking method to evaluate the
sample importance by using local information. Finally, we propose a delayed
training strategy to train the MBLP model. Extensive experiments are performed
on three domain datasets, and our method outperforms other state-of-the-art
methods.
| no_new_dataset | 0.948394 |
2112.12925 | Jiaxiang Tang | Xiaokang Chen, Jiaxiang Tang, Jingbo Wang, Gang Zeng | Not All Voxels Are Equal: Semantic Scene Completion from the Point-Voxel
Perspective | Accepted to AAAI 2022 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We revisit Semantic Scene Completion (SSC), a useful task to predict the
semantic and occupancy representation of 3D scenes, in this paper. A number of
methods for this task are always based on voxelized scene representations for
keeping local scene structure. However, due to the existence of visible empty
voxels, these methods always suffer from heavy computation redundancy when the
network goes deeper, and thus limit the completion quality. To address this
dilemma, we propose our novel point-voxel aggregation network for this task.
Firstly, we transfer the voxelized scenes to point clouds by removing these
visible empty voxels and adopt a deep point stream to capture semantic
information from the scene efficiently. Meanwhile, a light-weight voxel stream
containing only two 3D convolution layers preserves local structures of the
voxelized scenes. Furthermore, we design an anisotropic voxel aggregation
operator to fuse the structure details from the voxel stream into the point
stream, and a semantic-aware propagation module to enhance the up-sampling
process in the point stream by semantic labels. We demonstrate that our model
surpasses state-of-the-arts on two benchmarks by a large margin, with only
depth images as the input.
| [
{
"version": "v1",
"created": "Fri, 24 Dec 2021 03:25:40 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 12:30:36 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Chen",
"Xiaokang",
""
],
[
"Tang",
"Jiaxiang",
""
],
[
"Wang",
"Jingbo",
""
],
[
"Zeng",
"Gang",
""
]
] | TITLE: Not All Voxels Are Equal: Semantic Scene Completion from the Point-Voxel
Perspective
ABSTRACT: We revisit Semantic Scene Completion (SSC), a useful task to predict the
semantic and occupancy representation of 3D scenes, in this paper. A number of
methods for this task are always based on voxelized scene representations for
keeping local scene structure. However, due to the existence of visible empty
voxels, these methods always suffer from heavy computation redundancy when the
network goes deeper, and thus limit the completion quality. To address this
dilemma, we propose our novel point-voxel aggregation network for this task.
Firstly, we transfer the voxelized scenes to point clouds by removing these
visible empty voxels and adopt a deep point stream to capture semantic
information from the scene efficiently. Meanwhile, a light-weight voxel stream
containing only two 3D convolution layers preserves local structures of the
voxelized scenes. Furthermore, we design an anisotropic voxel aggregation
operator to fuse the structure details from the voxel stream into the point
stream, and a semantic-aware propagation module to enhance the up-sampling
process in the point stream by semantic labels. We demonstrate that our model
surpasses state-of-the-arts on two benchmarks by a large margin, with only
depth images as the input.
| no_new_dataset | 0.947817 |
2201.01305 | Digvijay Wadekar | Digvijay Wadekar, Leander Thiele, Francisco Villaescusa-Navarro, J.
Colin Hill, Miles Cranmer, David N. Spergel, Nicholas Battaglia, Daniel
Angl\'es-Alc\'azar, Lars Hernquist, Shirley Ho | Augmenting astrophysical scaling relations with machine learning:
application to reducing the Sunyaev-Zeldovich flux-mass scatter | Version appearing in PNAS. Added additional tests but results
unchanged compared to previous version. The code and data associated with
this paper are available at https://github.com/JayWadekar/ScalingRelations_ML | PNAS March 17, 2023, 120 (12) e2202074120 | 10.1073/pnas.2202074120 | null | astro-ph.CO astro-ph.GA astro-ph.IM cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Complex astrophysical systems often exhibit low-scatter relations between
observable properties (e.g., luminosity, velocity dispersion, oscillation
period). These scaling relations illuminate the underlying physics, and can
provide observational tools for estimating masses and distances. Machine
learning can provide a fast and systematic way to search for new scaling
relations (or for simple extensions to existing relations) in abstract
high-dimensional parameter spaces. We use a machine learning tool called
symbolic regression (SR), which models patterns in a dataset in the form of
analytic equations. We focus on the Sunyaev-Zeldovich flux$-$cluster mass
relation ($Y_\mathrm{SZ}-M$), the scatter in which affects inference of
cosmological parameters from cluster abundance data. Using SR on the data from
the IllustrisTNG hydrodynamical simulation, we find a new proxy for cluster
mass which combines $Y_\mathrm{SZ}$ and concentration of ionized gas
($c_\mathrm{gas}$): $M \propto Y_\mathrm{conc}^{3/5} \equiv Y_\mathrm{SZ}^{3/5}
(1-A\, c_\mathrm{gas})$. $Y_\mathrm{conc}$ reduces the scatter in the predicted
$M$ by $\sim 20-30$\% for large clusters ($M\gtrsim 10^{14}\, h^{-1} \,
M_\odot$), as compared to using just $Y_\mathrm{SZ}$. We show that the
dependence on $c_\mathrm{gas}$ is linked to cores of clusters exhibiting larger
scatter than their outskirts. Finally, we test $Y_\mathrm{conc}$ on clusters
from CAMELS simulations and show that $Y_\mathrm{conc}$ is robust against
variations in cosmology, subgrid physics, and cosmic variance. Our results and
methodology can be useful for accurate multiwavelength cluster mass estimation
from upcoming CMB and X-ray surveys like ACT, SO, eROSITA and CMB-S4.
| [
{
"version": "v1",
"created": "Tue, 4 Jan 2022 19:00:01 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Jan 2022 08:07:27 GMT"
},
{
"version": "v3",
"created": "Fri, 17 Mar 2023 18:46:41 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Wadekar",
"Digvijay",
""
],
[
"Thiele",
"Leander",
""
],
[
"Villaescusa-Navarro",
"Francisco",
""
],
[
"Hill",
"J. Colin",
""
],
[
"Cranmer",
"Miles",
""
],
[
"Spergel",
"David N.",
""
],
[
"Battaglia",
"Nicholas",
""
],
[
"Anglés-Alcázar",
"Daniel",
""
],
[
"Hernquist",
"Lars",
""
],
[
"Ho",
"Shirley",
""
]
] | TITLE: Augmenting astrophysical scaling relations with machine learning:
application to reducing the Sunyaev-Zeldovich flux-mass scatter
ABSTRACT: Complex astrophysical systems often exhibit low-scatter relations between
observable properties (e.g., luminosity, velocity dispersion, oscillation
period). These scaling relations illuminate the underlying physics, and can
provide observational tools for estimating masses and distances. Machine
learning can provide a fast and systematic way to search for new scaling
relations (or for simple extensions to existing relations) in abstract
high-dimensional parameter spaces. We use a machine learning tool called
symbolic regression (SR), which models patterns in a dataset in the form of
analytic equations. We focus on the Sunyaev-Zeldovich flux$-$cluster mass
relation ($Y_\mathrm{SZ}-M$), the scatter in which affects inference of
cosmological parameters from cluster abundance data. Using SR on the data from
the IllustrisTNG hydrodynamical simulation, we find a new proxy for cluster
mass which combines $Y_\mathrm{SZ}$ and concentration of ionized gas
($c_\mathrm{gas}$): $M \propto Y_\mathrm{conc}^{3/5} \equiv Y_\mathrm{SZ}^{3/5}
(1-A\, c_\mathrm{gas})$. $Y_\mathrm{conc}$ reduces the scatter in the predicted
$M$ by $\sim 20-30$\% for large clusters ($M\gtrsim 10^{14}\, h^{-1} \,
M_\odot$), as compared to using just $Y_\mathrm{SZ}$. We show that the
dependence on $c_\mathrm{gas}$ is linked to cores of clusters exhibiting larger
scatter than their outskirts. Finally, we test $Y_\mathrm{conc}$ on clusters
from CAMELS simulations and show that $Y_\mathrm{conc}$ is robust against
variations in cosmology, subgrid physics, and cosmic variance. Our results and
methodology can be useful for accurate multiwavelength cluster mass estimation
from upcoming CMB and X-ray surveys like ACT, SO, eROSITA and CMB-S4.
| no_new_dataset | 0.942135 |
2201.03597 | Elisabeth Wetzer | Eva Breznik and Elisabeth Wetzer and Joakim Lindblad and Nata\v{s}a
Sladoje | Cross-Modality Sub-Image Retrieval using Contrastive Multimodal Image
Representations | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In tissue characterization and cancer diagnostics, multimodal imaging has
emerged as a powerful technique. Thanks to computational advances, large
datasets can be exploited to discover patterns in pathologies and improve
diagnosis. However, this requires efficient and scalable image retrieval
methods. Cross-modality image retrieval is particularly challenging, since
images of similar (or even the same) content captured by different modalities
might share few common structures. We propose a new application-independent
content-based image retrieval (CBIR) system for reverse (sub-)image search
across modalities, which combines deep learning to generate representations
(embedding the different modalities in a common space) with classical feature
extraction and bag-of-words models for efficient and reliable retrieval. We
illustrate its advantages through a replacement study, exploring a number of
feature extractors and learned representations, as well as through comparison
to recent (cross-modality) CBIR methods. For the task of (sub-)image retrieval
on a (publicly available) dataset of brightfield and second harmonic generation
microscopy images, the results show that our approach is superior to all tested
alternatives. We discuss the shortcomings of the compared methods and observe
the importance of equivariance and invariance properties of the learned
representations and feature extractors in the CBIR pipeline. Code is available
at: \url{https://github.com/MIDA-group/CrossModal_ImgRetrieval}.
| [
{
"version": "v1",
"created": "Mon, 10 Jan 2022 19:04:28 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 08:58:19 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Breznik",
"Eva",
""
],
[
"Wetzer",
"Elisabeth",
""
],
[
"Lindblad",
"Joakim",
""
],
[
"Sladoje",
"Nataša",
""
]
] | TITLE: Cross-Modality Sub-Image Retrieval using Contrastive Multimodal Image
Representations
ABSTRACT: In tissue characterization and cancer diagnostics, multimodal imaging has
emerged as a powerful technique. Thanks to computational advances, large
datasets can be exploited to discover patterns in pathologies and improve
diagnosis. However, this requires efficient and scalable image retrieval
methods. Cross-modality image retrieval is particularly challenging, since
images of similar (or even the same) content captured by different modalities
might share few common structures. We propose a new application-independent
content-based image retrieval (CBIR) system for reverse (sub-)image search
across modalities, which combines deep learning to generate representations
(embedding the different modalities in a common space) with classical feature
extraction and bag-of-words models for efficient and reliable retrieval. We
illustrate its advantages through a replacement study, exploring a number of
feature extractors and learned representations, as well as through comparison
to recent (cross-modality) CBIR methods. For the task of (sub-)image retrieval
on a (publicly available) dataset of brightfield and second harmonic generation
microscopy images, the results show that our approach is superior to all tested
alternatives. We discuss the shortcomings of the compared methods and observe
the importance of equivariance and invariance properties of the learned
representations and feature extractors in the CBIR pipeline. Code is available
at: \url{https://github.com/MIDA-group/CrossModal_ImgRetrieval}.
| no_new_dataset | 0.94743 |
2201.09548 | Zhisheng Huang | Zhigang Tu, Zhisheng Huang, Yujin Chen, Di Kang, Linchao Bao, Bisheng
Yang, and Junsong Yuan | Consistent 3D Hand Reconstruction in Video via self-supervised Learning | arXiv admin note: substantial text overlap with arXiv:2103.11703 | IEEE Transactions on Pattern Analysis and Machine Intelligence.
2023 | 10.1109/TPAMI.2023.3247907 | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a method for reconstructing accurate and consistent 3D hands from
a monocular video. We observe that detected 2D hand keypoints and the image
texture provide important cues about the geometry and texture of the 3D hand,
which can reduce or even eliminate the requirement on 3D hand annotation. Thus
we propose ${\rm {S}^{2}HAND}$, a self-supervised 3D hand reconstruction model,
that can jointly estimate pose, shape, texture, and the camera viewpoint from a
single RGB input through the supervision of easily accessible 2D detected
keypoints. We leverage the continuous hand motion information contained in the
unlabeled video data and propose ${\rm {S}^{2}HAND(V)}$, which uses a set of
weights shared ${\rm {S}^{2}HAND}$ to process each frame and exploits
additional motion, texture, and shape consistency constrains to promote more
accurate hand poses and more consistent shapes and textures. Experiments on
benchmark datasets demonstrate that our self-supervised approach produces
comparable hand reconstruction performance compared with the recent
full-supervised methods in single-frame as input setup, and notably improves
the reconstruction accuracy and consistency when using video training data.
| [
{
"version": "v1",
"created": "Mon, 24 Jan 2022 09:44:11 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 04:28:10 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Tu",
"Zhigang",
""
],
[
"Huang",
"Zhisheng",
""
],
[
"Chen",
"Yujin",
""
],
[
"Kang",
"Di",
""
],
[
"Bao",
"Linchao",
""
],
[
"Yang",
"Bisheng",
""
],
[
"Yuan",
"Junsong",
""
]
] | TITLE: Consistent 3D Hand Reconstruction in Video via self-supervised Learning
ABSTRACT: We present a method for reconstructing accurate and consistent 3D hands from
a monocular video. We observe that detected 2D hand keypoints and the image
texture provide important cues about the geometry and texture of the 3D hand,
which can reduce or even eliminate the requirement on 3D hand annotation. Thus
we propose ${\rm {S}^{2}HAND}$, a self-supervised 3D hand reconstruction model,
that can jointly estimate pose, shape, texture, and the camera viewpoint from a
single RGB input through the supervision of easily accessible 2D detected
keypoints. We leverage the continuous hand motion information contained in the
unlabeled video data and propose ${\rm {S}^{2}HAND(V)}$, which uses a set of
weights shared ${\rm {S}^{2}HAND}$ to process each frame and exploits
additional motion, texture, and shape consistency constrains to promote more
accurate hand poses and more consistent shapes and textures. Experiments on
benchmark datasets demonstrate that our self-supervised approach produces
comparable hand reconstruction performance compared with the recent
full-supervised methods in single-frame as input setup, and notably improves
the reconstruction accuracy and consistency when using video training data.
| no_new_dataset | 0.94887 |
2202.03104 | Jun Xia | Jun Xia, Lirong Wu, Jintao Chen, Bozhen Hu, Stan Z.Li | SimGRACE: A Simple Framework for Graph Contrastive Learning without Data
Augmentation | Accepted by The Web Conference 2022 (WWW 2022) | The Web Conference 2022 | 10.1145/3485447.3512156 | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph contrastive learning (GCL) has emerged as a dominant technique for
graph representation learning which maximizes the mutual information between
paired graph augmentations that share the same semantics. Unfortunately, it is
difficult to preserve semantics well during augmentations in view of the
diverse nature of graph data. Currently, data augmentations in GCL that are
designed to preserve semantics broadly fall into three unsatisfactory ways.
First, the augmentations can be manually picked per dataset by
trial-and-errors. Second, the augmentations can be selected via cumbersome
search. Third, the augmentations can be obtained by introducing expensive
domain-specific knowledge as guidance. All of these limit the efficiency and
more general applicability of existing GCL methods. To circumvent these crucial
issues, we propose a \underline{Sim}ple framework for \underline{GRA}ph
\underline{C}ontrastive l\underline{E}arning, \textbf{SimGRACE} for brevity,
which does not require data augmentations. Specifically, we take original graph
as input and GNN model with its perturbed version as two encoders to obtain two
correlated views for contrast. SimGRACE is inspired by the observation that
graph data can preserve their semantics well during encoder perturbations while
not requiring manual trial-and-errors, cumbersome search or expensive domain
knowledge for augmentations selection. Also, we explain why SimGRACE can
succeed. Furthermore, we devise adversarial training scheme, dubbed
\textbf{AT-SimGRACE}, to enhance the robustness of graph contrastive learning
and theoretically explain the reasons. Albeit simple, we show that SimGRACE can
yield competitive or better performance compared with state-of-the-art methods
in terms of generalizability, transferability and robustness, while enjoying
unprecedented degree of flexibility and efficiency.
| [
{
"version": "v1",
"created": "Mon, 7 Feb 2022 12:33:14 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Feb 2022 15:26:05 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 14:51:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Xia",
"Jun",
""
],
[
"Wu",
"Lirong",
""
],
[
"Chen",
"Jintao",
""
],
[
"Hu",
"Bozhen",
""
],
[
"Li",
"Stan Z.",
""
]
] | TITLE: SimGRACE: A Simple Framework for Graph Contrastive Learning without Data
Augmentation
ABSTRACT: Graph contrastive learning (GCL) has emerged as a dominant technique for
graph representation learning which maximizes the mutual information between
paired graph augmentations that share the same semantics. Unfortunately, it is
difficult to preserve semantics well during augmentations in view of the
diverse nature of graph data. Currently, data augmentations in GCL that are
designed to preserve semantics broadly fall into three unsatisfactory ways.
First, the augmentations can be manually picked per dataset by
trial-and-errors. Second, the augmentations can be selected via cumbersome
search. Third, the augmentations can be obtained by introducing expensive
domain-specific knowledge as guidance. All of these limit the efficiency and
more general applicability of existing GCL methods. To circumvent these crucial
issues, we propose a \underline{Sim}ple framework for \underline{GRA}ph
\underline{C}ontrastive l\underline{E}arning, \textbf{SimGRACE} for brevity,
which does not require data augmentations. Specifically, we take original graph
as input and GNN model with its perturbed version as two encoders to obtain two
correlated views for contrast. SimGRACE is inspired by the observation that
graph data can preserve their semantics well during encoder perturbations while
not requiring manual trial-and-errors, cumbersome search or expensive domain
knowledge for augmentations selection. Also, we explain why SimGRACE can
succeed. Furthermore, we devise adversarial training scheme, dubbed
\textbf{AT-SimGRACE}, to enhance the robustness of graph contrastive learning
and theoretically explain the reasons. Albeit simple, we show that SimGRACE can
yield competitive or better performance compared with state-of-the-art methods
in terms of generalizability, transferability and robustness, while enjoying
unprecedented degree of flexibility and efficiency.
| no_new_dataset | 0.944485 |
2202.05126 | Jiang Hao | Hao Jiang, Yanning Zhou, Yi Lin, Ronald CK Chan, Jiang Liu, Hao Chen | Deep Learning for Computational Cytology: A Survey | null | Medical Image Analysis, Nov 2022 | 10.1016/j.media.2022.102691 | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Computational cytology is a critical, rapid-developing, yet challenging topic
in the field of medical image computing which analyzes the digitized cytology
image by computer-aided technologies for cancer screening. Recently, an
increasing number of deep learning (DL) algorithms have made significant
progress in medical image analysis, leading to the boosting publications of
cytological studies. To investigate the advanced methods and comprehensive
applications, we survey more than 120 publications of DL-based cytology image
analysis in this article. We first introduce various deep learning methods,
including fully supervised, weakly supervised, unsupervised, and transfer
learning. Then, we systematically summarize the public datasets, evaluation
metrics, versatile cytology image analysis applications including
classification, detection, segmentation, and other related tasks. Finally, we
discuss current challenges and potential research directions of computational
cytology.
| [
{
"version": "v1",
"created": "Thu, 10 Feb 2022 16:22:10 GMT"
},
{
"version": "v2",
"created": "Wed, 16 Feb 2022 05:19:50 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Jiang",
"Hao",
""
],
[
"Zhou",
"Yanning",
""
],
[
"Lin",
"Yi",
""
],
[
"Chan",
"Ronald CK",
""
],
[
"Liu",
"Jiang",
""
],
[
"Chen",
"Hao",
""
]
] | TITLE: Deep Learning for Computational Cytology: A Survey
ABSTRACT: Computational cytology is a critical, rapid-developing, yet challenging topic
in the field of medical image computing which analyzes the digitized cytology
image by computer-aided technologies for cancer screening. Recently, an
increasing number of deep learning (DL) algorithms have made significant
progress in medical image analysis, leading to the boosting publications of
cytological studies. To investigate the advanced methods and comprehensive
applications, we survey more than 120 publications of DL-based cytology image
analysis in this article. We first introduce various deep learning methods,
including fully supervised, weakly supervised, unsupervised, and transfer
learning. Then, we systematically summarize the public datasets, evaluation
metrics, versatile cytology image analysis applications including
classification, detection, segmentation, and other related tasks. Finally, we
discuss current challenges and potential research directions of computational
cytology.
| no_new_dataset | 0.946843 |
2202.07728 | Thomas Fel | Thomas Fel, Melanie Ducoffe, David Vigouroux, Remi Cadene, Mikael
Capelle, Claire Nicodeme, Thomas Serre | Don't Lie to Me! Robust and Efficient Explainability with Verified
Perturbation Analysis | null | Proceedings of the IEEE / CVF Computer Vision and Pattern
Recognition Conference (CVPR), 2023 | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | A variety of methods have been proposed to try to explain how deep neural
networks make their decisions. Key to those approaches is the need to sample
the pixel space efficiently in order to derive importance maps. However, it has
been shown that the sampling methods used to date introduce biases and other
artifacts, leading to inaccurate estimates of the importance of individual
pixels and severely limit the reliability of current explainability methods.
Unfortunately, the alternative -- to exhaustively sample the image space is
computationally prohibitive. In this paper, we introduce EVA (Explaining using
Verified perturbation Analysis) -- the first explainability method guarantee to
have an exhaustive exploration of a perturbation space. Specifically, we
leverage the beneficial properties of verified perturbation analysis -- time
efficiency, tractability and guaranteed complete coverage of a manifold -- to
efficiently characterize the input variables that are most likely to drive the
model decision. We evaluate the approach systematically and demonstrate
state-of-the-art results on multiple benchmarks.
| [
{
"version": "v1",
"created": "Tue, 15 Feb 2022 21:13:55 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 10:45:10 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Fel",
"Thomas",
""
],
[
"Ducoffe",
"Melanie",
""
],
[
"Vigouroux",
"David",
""
],
[
"Cadene",
"Remi",
""
],
[
"Capelle",
"Mikael",
""
],
[
"Nicodeme",
"Claire",
""
],
[
"Serre",
"Thomas",
""
]
] | TITLE: Don't Lie to Me! Robust and Efficient Explainability with Verified
Perturbation Analysis
ABSTRACT: A variety of methods have been proposed to try to explain how deep neural
networks make their decisions. Key to those approaches is the need to sample
the pixel space efficiently in order to derive importance maps. However, it has
been shown that the sampling methods used to date introduce biases and other
artifacts, leading to inaccurate estimates of the importance of individual
pixels and severely limit the reliability of current explainability methods.
Unfortunately, the alternative -- to exhaustively sample the image space is
computationally prohibitive. In this paper, we introduce EVA (Explaining using
Verified perturbation Analysis) -- the first explainability method guarantee to
have an exhaustive exploration of a perturbation space. Specifically, we
leverage the beneficial properties of verified perturbation analysis -- time
efficiency, tractability and guaranteed complete coverage of a manifold -- to
efficiently characterize the input variables that are most likely to drive the
model decision. We evaluate the approach systematically and demonstrate
state-of-the-art results on multiple benchmarks.
| no_new_dataset | 0.943086 |
2203.05659 | Benjamin Horne | Maur\'icio Gruppi, Benjamin D. Horne, Sibel Adal{\i} | NELA-GT-2022: A Large Multi-Labelled News Dataset for The Study of
Misinformation in News Articles | Technical report documenting the NELA-GT recent update
(NELA-GT-2022). arXiv admin note: substantial text overlap with
arXiv:2102.04567 | null | null | null | cs.CL cs.CY cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we present the fifth installment of the NELA-GT datasets,
NELA-GT-2022. The dataset contains 1,778,361 articles from 361 outlets between
January 1st, 2022 and December 31st, 2022. Just as in past releases of the
dataset, NELA-GT-2022 includes outlet-level veracity labels from Media
Bias/Fact Check and tweets embedded in collected news articles. The
NELA-GT-2022 dataset can be found at: https://doi.org/10.7910/DVN/AMCV2H
| [
{
"version": "v1",
"created": "Thu, 10 Mar 2022 21:58:33 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Mar 2023 22:21:50 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Gruppi",
"Maurício",
""
],
[
"Horne",
"Benjamin D.",
""
],
[
"Adalı",
"Sibel",
""
]
] | TITLE: NELA-GT-2022: A Large Multi-Labelled News Dataset for The Study of
Misinformation in News Articles
ABSTRACT: In this paper, we present the fifth installment of the NELA-GT datasets,
NELA-GT-2022. The dataset contains 1,778,361 articles from 361 outlets between
January 1st, 2022 and December 31st, 2022. Just as in past releases of the
dataset, NELA-GT-2022 includes outlet-level veracity labels from Media
Bias/Fact Check and tweets embedded in collected news articles. The
NELA-GT-2022 dataset can be found at: https://doi.org/10.7910/DVN/AMCV2H
| new_dataset | 0.922062 |
2203.11854 | Jakob Hoydis | Jakob Hoydis, Sebastian Cammerer, Fay\c{c}al Ait Aoudia, Avinash Vem,
Nikolaus Binder, Guillermo Marcus, Alexander Keller | Sionna: An Open-Source Library for Next-Generation Physical Layer
Research | 5 pages, 1 figure, 4 code listings | null | null | null | cs.IT cs.AI cs.LG math.IT | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Sionna is a GPU-accelerated open-source library for link-level simulations
based on TensorFlow. It enables the rapid prototyping of complex communication
system architectures and provides native support for the integration of neural
networks. Sionna implements a wide breadth of carefully tested state-of-the-art
algorithms that can be used for benchmarking and end-to-end performance
evaluation. This allows researchers to focus on their research, making it more
impactful and reproducible, while saving time implementing components outside
their area of expertise. This white paper provides a brief introduction to
Sionna, explains its design principles and features, as well as future
extensions, such as integrated ray tracing and custom CUDA kernels. We believe
that Sionna is a valuable tool for research on next-generation communication
systems, such as 6G, and we welcome contributions from our community.
| [
{
"version": "v1",
"created": "Tue, 22 Mar 2022 16:31:44 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 13:51:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Hoydis",
"Jakob",
""
],
[
"Cammerer",
"Sebastian",
""
],
[
"Aoudia",
"Fayçal Ait",
""
],
[
"Vem",
"Avinash",
""
],
[
"Binder",
"Nikolaus",
""
],
[
"Marcus",
"Guillermo",
""
],
[
"Keller",
"Alexander",
""
]
] | TITLE: Sionna: An Open-Source Library for Next-Generation Physical Layer
Research
ABSTRACT: Sionna is a GPU-accelerated open-source library for link-level simulations
based on TensorFlow. It enables the rapid prototyping of complex communication
system architectures and provides native support for the integration of neural
networks. Sionna implements a wide breadth of carefully tested state-of-the-art
algorithms that can be used for benchmarking and end-to-end performance
evaluation. This allows researchers to focus on their research, making it more
impactful and reproducible, while saving time implementing components outside
their area of expertise. This white paper provides a brief introduction to
Sionna, explains its design principles and features, as well as future
extensions, such as integrated ray tracing and custom CUDA kernels. We believe
that Sionna is a valuable tool for research on next-generation communication
systems, such as 6G, and we welcome contributions from our community.
| no_new_dataset | 0.941922 |
2204.03652 | Tu\u{g}berk Erol | Tugberk Erol and Duygu Sarikaya | PlutoNet: An Efficient Polyp Segmentation Network with Modified Partial
Decoder and Decoder Consistency Training | 10 pages, 2 figures, 2 tables | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Deep learning models are used to minimize the number of polyps that goes
unnoticed by the experts and to accurately segment the detected polyps during
interventions. Although state-of-the-art models are proposed, it remains a
challenge to define representations that are able to generalize well and that
mediate between capturing low-level features and higher-level semantic details
without being redundant. Another challenge with these models is that they
require too many parameters, which can pose a problem with real-time
applications. To address these problems, we propose PlutoNet for polyp
segmentation which requires only 2,626,537 parameters, less than 10\% of the
parameters required by its counterparts. With PlutoNet, we propose a novel
\emph{decoder consistency training} approach that consists of a shared encoder,
the modified partial decoder which is a combination of the partial decoder and
full-scale connections that capture salient features at different scales
without being redundant, and the auxiliary decoder which focuses on
higher-level relevant semantic features. We train the modified partial decoder
and the auxiliary decoder with a combined loss to enforce consistency, which
helps improve the encoders representations. This way we are able to reduce
uncertainty and false positive rates. We perform ablation studies and extensive
experiments which show that PlutoNet performs significantly better than the
state-of-the-art models, particularly on unseen datasets and datasets across
different domains.
| [
{
"version": "v1",
"created": "Wed, 6 Apr 2022 20:29:00 GMT"
},
{
"version": "v2",
"created": "Tue, 12 Apr 2022 19:26:59 GMT"
},
{
"version": "v3",
"created": "Sat, 12 Nov 2022 23:14:38 GMT"
},
{
"version": "v4",
"created": "Sat, 18 Mar 2023 16:25:15 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Erol",
"Tugberk",
""
],
[
"Sarikaya",
"Duygu",
""
]
] | TITLE: PlutoNet: An Efficient Polyp Segmentation Network with Modified Partial
Decoder and Decoder Consistency Training
ABSTRACT: Deep learning models are used to minimize the number of polyps that goes
unnoticed by the experts and to accurately segment the detected polyps during
interventions. Although state-of-the-art models are proposed, it remains a
challenge to define representations that are able to generalize well and that
mediate between capturing low-level features and higher-level semantic details
without being redundant. Another challenge with these models is that they
require too many parameters, which can pose a problem with real-time
applications. To address these problems, we propose PlutoNet for polyp
segmentation which requires only 2,626,537 parameters, less than 10\% of the
parameters required by its counterparts. With PlutoNet, we propose a novel
\emph{decoder consistency training} approach that consists of a shared encoder,
the modified partial decoder which is a combination of the partial decoder and
full-scale connections that capture salient features at different scales
without being redundant, and the auxiliary decoder which focuses on
higher-level relevant semantic features. We train the modified partial decoder
and the auxiliary decoder with a combined loss to enforce consistency, which
helps improve the encoders representations. This way we are able to reduce
uncertainty and false positive rates. We perform ablation studies and extensive
experiments which show that PlutoNet performs significantly better than the
state-of-the-art models, particularly on unseen datasets and datasets across
different domains.
| no_new_dataset | 0.947962 |
2204.06601 | Jeremy Tien | Jeremy Tien, Jerry Zhi-Yang He, Zackory Erickson, Anca D. Dragan,
Daniel S. Brown | Causal Confusion and Reward Misidentification in Preference-Based Reward
Learning | In the proceedings of the Eleventh International Conference on
Learning Representations (ICLR 2023).
https://iclr.cc/virtual/2023/poster/10822 | null | null | null | cs.LG cs.RO | http://creativecommons.org/licenses/by/4.0/ | Learning policies via preference-based reward learning is an increasingly
popular method for customizing agent behavior, but has been shown anecdotally
to be prone to spurious correlations and reward hacking behaviors. While much
prior work focuses on causal confusion in reinforcement learning and behavioral
cloning, we focus on a systematic study of causal confusion and reward
misidentification when learning from preferences. In particular, we perform a
series of sensitivity and ablation analyses on several benchmark domains where
rewards learned from preferences achieve minimal test error but fail to
generalize to out-of-distribution states -- resulting in poor policy
performance when optimized. We find that the presence of non-causal distractor
features, noise in the stated preferences, and partial state observability can
all exacerbate reward misidentification. We also identify a set of methods with
which to interpret misidentified learned rewards. In general, we observe that
optimizing misidentified rewards drives the policy off the reward's training
distribution, resulting in high predicted (learned) rewards but low true
rewards. These findings illuminate the susceptibility of preference learning to
reward misidentification and causal confusion -- failure to consider even one
of many factors can result in unexpected, undesirable behavior.
| [
{
"version": "v1",
"created": "Wed, 13 Apr 2022 18:41:41 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Oct 2022 01:52:35 GMT"
},
{
"version": "v3",
"created": "Thu, 9 Mar 2023 02:45:48 GMT"
},
{
"version": "v4",
"created": "Sat, 18 Mar 2023 20:44:45 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Tien",
"Jeremy",
""
],
[
"He",
"Jerry Zhi-Yang",
""
],
[
"Erickson",
"Zackory",
""
],
[
"Dragan",
"Anca D.",
""
],
[
"Brown",
"Daniel S.",
""
]
] | TITLE: Causal Confusion and Reward Misidentification in Preference-Based Reward
Learning
ABSTRACT: Learning policies via preference-based reward learning is an increasingly
popular method for customizing agent behavior, but has been shown anecdotally
to be prone to spurious correlations and reward hacking behaviors. While much
prior work focuses on causal confusion in reinforcement learning and behavioral
cloning, we focus on a systematic study of causal confusion and reward
misidentification when learning from preferences. In particular, we perform a
series of sensitivity and ablation analyses on several benchmark domains where
rewards learned from preferences achieve minimal test error but fail to
generalize to out-of-distribution states -- resulting in poor policy
performance when optimized. We find that the presence of non-causal distractor
features, noise in the stated preferences, and partial state observability can
all exacerbate reward misidentification. We also identify a set of methods with
which to interpret misidentified learned rewards. In general, we observe that
optimizing misidentified rewards drives the policy off the reward's training
distribution, resulting in high predicted (learned) rewards but low true
rewards. These findings illuminate the susceptibility of preference learning to
reward misidentification and causal confusion -- failure to consider even one
of many factors can result in unexpected, undesirable behavior.
| no_new_dataset | 0.948394 |
2204.13386 | Yang Liu | Yang Liu, Ying Tan, Haoyuan Lan | Self-supervised Contrastive Learning for Audio-Visual Action Recognition | 6 pages, 2 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The underlying correlation between audio and visual modalities can be
utilized to learn supervised information for unlabeled videos. In this paper,
we propose an end-to-end self-supervised framework named Audio-Visual
Contrastive Learning (AVCL), to learn discriminative audio-visual
representations for action recognition. Specifically, we design an attention
based multi-modal fusion module (AMFM) to fuse audio and visual modalities. To
align heterogeneous audio-visual modalities, we construct a novel
co-correlation guided representation alignment module (CGRA). To learn
supervised information from unlabeled videos, we propose a novel
self-supervised contrastive learning module (SelfCL). Furthermore, we build a
new audio-visual action recognition dataset named Kinetics-Sounds100.
Experimental results on Kinetics-Sounds32 and Kinetics-Sounds100 datasets
demonstrate the superiority of our AVCL over the state-of-the-art methods on
large-scale action recognition benchmark.
| [
{
"version": "v1",
"created": "Thu, 28 Apr 2022 10:01:36 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 07:33:09 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Liu",
"Yang",
""
],
[
"Tan",
"Ying",
""
],
[
"Lan",
"Haoyuan",
""
]
] | TITLE: Self-supervised Contrastive Learning for Audio-Visual Action Recognition
ABSTRACT: The underlying correlation between audio and visual modalities can be
utilized to learn supervised information for unlabeled videos. In this paper,
we propose an end-to-end self-supervised framework named Audio-Visual
Contrastive Learning (AVCL), to learn discriminative audio-visual
representations for action recognition. Specifically, we design an attention
based multi-modal fusion module (AMFM) to fuse audio and visual modalities. To
align heterogeneous audio-visual modalities, we construct a novel
co-correlation guided representation alignment module (CGRA). To learn
supervised information from unlabeled videos, we propose a novel
self-supervised contrastive learning module (SelfCL). Furthermore, we build a
new audio-visual action recognition dataset named Kinetics-Sounds100.
Experimental results on Kinetics-Sounds32 and Kinetics-Sounds100 datasets
demonstrate the superiority of our AVCL over the state-of-the-art methods on
large-scale action recognition benchmark.
| new_dataset | 0.944485 |
2205.06090 | Uisub Shin | Uisub Shin, Cong Ding, Bingzhao Zhu, Yashwanth Vyza, Alix Trouillet,
Emilie C. M. Revol, St\'ephanie P. Lacour, and Mahsa Shoaran | NeuralTree: A 256-Channel 0.227-$\mu$J/Class Versatile Neural Activity
Classification and Closed-Loop Neuromodulation SoC | null | IEEE Journal of Solid-State Circuits, vol. 57, no. 11, pp.
3243-3257, Nov. 2022 | 10.1109/JSSC.2022.3204508 | null | eess.SP cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Closed-loop neural interfaces with on-chip machine learning can detect and
suppress disease symptoms in neurological disorders or restore lost functions
in paralyzed patients. While high-density neural recording can provide rich
neural activity information for accurate disease-state detection, existing
systems have low channel counts and poor scalability, which could limit their
therapeutic efficacy. This work presents a highly scalable and versatile
closed-loop neural interface SoC that can overcome these limitations. A
256-channel time-division multiplexed (TDM) front-end with a two-step
fast-settling mixed-signal DC servo loop (DSL) is proposed to record
high-spatial-resolution neural activity and perform channel-selective
brain-state inference. A tree-structured neural network (NeuralTree)
classification processor extracts a rich set of neural biomarkers in a patient-
and disease-specific manner. Trained with an energy-aware learning algorithm,
the NeuralTree classifier detects the symptoms of underlying disorders (e.g.,
epilepsy and movement disorders) at an optimal energy-accuracy tradeoff. A
16-channel high-voltage (HV) compliant neurostimulator closes the therapeutic
loop by delivering charge-balanced biphasic current pulses to the brain. The
proposed SoC was fabricated in 65-nm CMOS and achieved a 0.227-$\mu$J/class
energy efficiency in a compact area of 0.014mm$^2$/channel. The SoC was
extensively verified on human electroencephalography (EEG) and intracranial EEG
(iEEG) epilepsy datasets, obtaining 95.6%/94% sensitivity and 96.8%/96.9%
specificity, respectively. In vivo neural recordings using soft $\mu$ECoG
arrays and multi-domain biomarker extraction were further performed on a rat
model of epilepsy. In addition, for the first time in literature, on-chip
classification of rest-state tremor in Parkinson's disease (PD) from human
local field potentials (LFPs) was demonstrated.
| [
{
"version": "v1",
"created": "Thu, 12 May 2022 13:56:01 GMT"
},
{
"version": "v2",
"created": "Sat, 21 May 2022 12:06:16 GMT"
},
{
"version": "v3",
"created": "Thu, 8 Dec 2022 17:03:11 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Shin",
"Uisub",
""
],
[
"Ding",
"Cong",
""
],
[
"Zhu",
"Bingzhao",
""
],
[
"Vyza",
"Yashwanth",
""
],
[
"Trouillet",
"Alix",
""
],
[
"Revol",
"Emilie C. M.",
""
],
[
"Lacour",
"Stéphanie P.",
""
],
[
"Shoaran",
"Mahsa",
""
]
] | TITLE: NeuralTree: A 256-Channel 0.227-$\mu$J/Class Versatile Neural Activity
Classification and Closed-Loop Neuromodulation SoC
ABSTRACT: Closed-loop neural interfaces with on-chip machine learning can detect and
suppress disease symptoms in neurological disorders or restore lost functions
in paralyzed patients. While high-density neural recording can provide rich
neural activity information for accurate disease-state detection, existing
systems have low channel counts and poor scalability, which could limit their
therapeutic efficacy. This work presents a highly scalable and versatile
closed-loop neural interface SoC that can overcome these limitations. A
256-channel time-division multiplexed (TDM) front-end with a two-step
fast-settling mixed-signal DC servo loop (DSL) is proposed to record
high-spatial-resolution neural activity and perform channel-selective
brain-state inference. A tree-structured neural network (NeuralTree)
classification processor extracts a rich set of neural biomarkers in a patient-
and disease-specific manner. Trained with an energy-aware learning algorithm,
the NeuralTree classifier detects the symptoms of underlying disorders (e.g.,
epilepsy and movement disorders) at an optimal energy-accuracy tradeoff. A
16-channel high-voltage (HV) compliant neurostimulator closes the therapeutic
loop by delivering charge-balanced biphasic current pulses to the brain. The
proposed SoC was fabricated in 65-nm CMOS and achieved a 0.227-$\mu$J/class
energy efficiency in a compact area of 0.014mm$^2$/channel. The SoC was
extensively verified on human electroencephalography (EEG) and intracranial EEG
(iEEG) epilepsy datasets, obtaining 95.6%/94% sensitivity and 96.8%/96.9%
specificity, respectively. In vivo neural recordings using soft $\mu$ECoG
arrays and multi-domain biomarker extraction were further performed on a rat
model of epilepsy. In addition, for the first time in literature, on-chip
classification of rest-state tremor in Parkinson's disease (PD) from human
local field potentials (LFPs) was demonstrated.
| no_new_dataset | 0.947137 |
2205.11083 | Jinwoo Bae | Jinwoo Bae, Sungho Moon, Sunghoon Im | Deep Digging into the Generalization of Self-Supervised Monocular Depth
Estimation | Accepted to AAAI 2023 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-supervised monocular depth estimation has been widely studied recently.
Most of the work has focused on improving performance on benchmark datasets,
such as KITTI, but has offered a few experiments on generalization performance.
In this paper, we investigate the backbone networks (e.g. CNNs, Transformers,
and CNN-Transformer hybrid models) toward the generalization of monocular depth
estimation. We first evaluate state-of-the-art models on diverse public
datasets, which have never been seen during the network training. Next, we
investigate the effects of texture-biased and shape-biased representations
using the various texture-shifted datasets that we generated. We observe that
Transformers exhibit a strong shape bias and CNNs do a strong texture-bias. We
also find that shape-biased models show better generalization performance for
monocular depth estimation compared to texture-biased models. Based on these
observations, we newly design a CNN-Transformer hybrid network with a
multi-level adaptive feature fusion module, called MonoFormer. The design
intuition behind MonoFormer is to increase shape bias by employing Transformers
while compensating for the weak locality bias of Transformers by adaptively
fusing multi-level representations. Extensive experiments show that the
proposed method achieves state-of-the-art performance with various public
datasets. Our method also shows the best generalization ability among the
competitive methods.
| [
{
"version": "v1",
"created": "Mon, 23 May 2022 06:56:25 GMT"
},
{
"version": "v2",
"created": "Sat, 19 Nov 2022 16:28:05 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 03:52:42 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Bae",
"Jinwoo",
""
],
[
"Moon",
"Sungho",
""
],
[
"Im",
"Sunghoon",
""
]
] | TITLE: Deep Digging into the Generalization of Self-Supervised Monocular Depth
Estimation
ABSTRACT: Self-supervised monocular depth estimation has been widely studied recently.
Most of the work has focused on improving performance on benchmark datasets,
such as KITTI, but has offered a few experiments on generalization performance.
In this paper, we investigate the backbone networks (e.g. CNNs, Transformers,
and CNN-Transformer hybrid models) toward the generalization of monocular depth
estimation. We first evaluate state-of-the-art models on diverse public
datasets, which have never been seen during the network training. Next, we
investigate the effects of texture-biased and shape-biased representations
using the various texture-shifted datasets that we generated. We observe that
Transformers exhibit a strong shape bias and CNNs do a strong texture-bias. We
also find that shape-biased models show better generalization performance for
monocular depth estimation compared to texture-biased models. Based on these
observations, we newly design a CNN-Transformer hybrid network with a
multi-level adaptive feature fusion module, called MonoFormer. The design
intuition behind MonoFormer is to increase shape bias by employing Transformers
while compensating for the weak locality bias of Transformers by adaptively
fusing multi-level representations. Extensive experiments show that the
proposed method achieves state-of-the-art performance with various public
datasets. Our method also shows the best generalization ability among the
competitive methods.
| no_new_dataset | 0.94868 |
2205.11346 | Xin Wang | Xin Wang, Sheng Wang, Honglin Xiong, Kai Xuan, Zixu Zhuang, Mengjun
Liu, Zhenrong Shen, Xiangyu Zhao, Lichi Zhang, Qian Wang | Spatial Attention-based Implicit Neural Representation for Arbitrary
Reduction of MRI Slice Spacing | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Magnetic resonance (MR) images collected in 2D clinical protocols typically
have large inter-slice spacing, resulting in high in-plane resolution and
reduced through-plane resolution. Super-resolution technique can enhance the
through-plane resolution of MR images to facilitate downstream visualization
and computer-aided diagnosis. However, most existing works train the
super-resolution network at a fixed scaling factor, which is not friendly to
clinical scenes of varying inter-slice spacing in MR scanning. Inspired by the
recent progress in implicit neural representation, we propose a Spatial
Attention-based Implicit Neural Representation (SA-INR) network for arbitrary
reduction of MR inter-slice spacing. The SA-INR aims to represent an MR image
as a continuous implicit function of 3D coordinates. In this way, the SA-INR
can reconstruct the MR image with arbitrary inter-slice spacing by continuously
sampling the coordinates in 3D space. In particular, a local-aware spatial
attention operation is introduced to model nearby voxels and their affinity
more accurately in a larger receptive field. Meanwhile, to improve the
computational efficiency, a gradient-guided gating mask is proposed for
applying the local-aware spatial attention to selected areas only. We evaluate
our method on the public HCP-1200 dataset and the clinical knee MR dataset to
demonstrate its superiority over other existing methods.
| [
{
"version": "v1",
"created": "Mon, 23 May 2022 14:36:59 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 01:59:51 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Wang",
"Xin",
""
],
[
"Wang",
"Sheng",
""
],
[
"Xiong",
"Honglin",
""
],
[
"Xuan",
"Kai",
""
],
[
"Zhuang",
"Zixu",
""
],
[
"Liu",
"Mengjun",
""
],
[
"Shen",
"Zhenrong",
""
],
[
"Zhao",
"Xiangyu",
""
],
[
"Zhang",
"Lichi",
""
],
[
"Wang",
"Qian",
""
]
] | TITLE: Spatial Attention-based Implicit Neural Representation for Arbitrary
Reduction of MRI Slice Spacing
ABSTRACT: Magnetic resonance (MR) images collected in 2D clinical protocols typically
have large inter-slice spacing, resulting in high in-plane resolution and
reduced through-plane resolution. Super-resolution technique can enhance the
through-plane resolution of MR images to facilitate downstream visualization
and computer-aided diagnosis. However, most existing works train the
super-resolution network at a fixed scaling factor, which is not friendly to
clinical scenes of varying inter-slice spacing in MR scanning. Inspired by the
recent progress in implicit neural representation, we propose a Spatial
Attention-based Implicit Neural Representation (SA-INR) network for arbitrary
reduction of MR inter-slice spacing. The SA-INR aims to represent an MR image
as a continuous implicit function of 3D coordinates. In this way, the SA-INR
can reconstruct the MR image with arbitrary inter-slice spacing by continuously
sampling the coordinates in 3D space. In particular, a local-aware spatial
attention operation is introduced to model nearby voxels and their affinity
more accurately in a larger receptive field. Meanwhile, to improve the
computational efficiency, a gradient-guided gating mask is proposed for
applying the local-aware spatial attention to selected areas only. We evaluate
our method on the public HCP-1200 dataset and the clinical knee MR dataset to
demonstrate its superiority over other existing methods.
| no_new_dataset | 0.949435 |
2205.11765 | Banghua Zhu | Banghua Zhu, Lun Wang, Qi Pang, Shuai Wang, Jiantao Jiao, Dawn Song,
Michael I. Jordan | Byzantine-Robust Federated Learning with Optimal Statistical Rates and
Privacy Guarantees | null | null | null | null | cs.LG cs.AI cs.CR cs.DC stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose Byzantine-robust federated learning protocols with nearly optimal
statistical rates. In contrast to prior work, our proposed protocols improve
the dimension dependence and achieve a tight statistical rate in terms of all
the parameters for strongly convex losses. We benchmark against competing
protocols and show the empirical superiority of the proposed protocols.
Finally, we remark that our protocols with bucketing can be naturally combined
with privacy-guaranteeing procedures to introduce security against a
semi-honest server. The code for evaluation is provided in
https://github.com/wanglun1996/secure-robust-federated-learning.
| [
{
"version": "v1",
"created": "Tue, 24 May 2022 04:03:07 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 18:33:56 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhu",
"Banghua",
""
],
[
"Wang",
"Lun",
""
],
[
"Pang",
"Qi",
""
],
[
"Wang",
"Shuai",
""
],
[
"Jiao",
"Jiantao",
""
],
[
"Song",
"Dawn",
""
],
[
"Jordan",
"Michael I.",
""
]
] | TITLE: Byzantine-Robust Federated Learning with Optimal Statistical Rates and
Privacy Guarantees
ABSTRACT: We propose Byzantine-robust federated learning protocols with nearly optimal
statistical rates. In contrast to prior work, our proposed protocols improve
the dimension dependence and achieve a tight statistical rate in terms of all
the parameters for strongly convex losses. We benchmark against competing
protocols and show the empirical superiority of the proposed protocols.
Finally, we remark that our protocols with bucketing can be naturally combined
with privacy-guaranteeing procedures to introduce security against a
semi-honest server. The code for evaluation is provided in
https://github.com/wanglun1996/secure-robust-federated-learning.
| no_new_dataset | 0.950319 |
2205.11857 | Shijie Zhang | Shijie Zhang, Wei Yuan and Hongzhi Yin | Comprehensive Privacy Analysis on Federated Recommender System against
Attribute Inference Attacks | null | null | null | null | cs.IR | http://creativecommons.org/publicdomain/zero/1.0/ | In recent years, recommender systems are crucially important for the delivery
of personalized services that satisfy users' preferences. With personalized
recommendation services, users can enjoy a variety of recommendations such as
movies, books, ads, restaurants, and more. Despite the great benefits,
personalized recommendations typically require the collection of personal data
for user modelling and analysis, which can make users susceptible to attribute
inference attacks. Specifically, the vulnerability of existing centralized
recommenders under attribute inference attacks leaves malicious attackers a
backdoor to infer users' private attributes, as the systems remember
information of their training data (i.e., interaction data and side
information). An emerging practice is to implement recommender systems in the
federated setting, which enables all user devices to collaboratively learn a
shared global recommender while keeping all the training data on device.
However, the privacy issues in federated recommender systems have been rarely
explored. In this paper, we first design a novel attribute inference attacker
to perform a comprehensive privacy analysis of the state-of-the-art federated
recommender models. The experimental results show that the vulnerability of
each model component against attribute inference attack is varied, highlighting
the need for new defense approaches. Therefore, we propose a novel adaptive
privacy-preserving approach to protect users' sensitive data in the presence of
attribute inference attacks and meanwhile maximize the recommendation accuracy.
Extensive experimental results on two real-world datasets validate the superior
performance of our model on both recommendation effectiveness and resistance to
inference attacks.
| [
{
"version": "v1",
"created": "Tue, 24 May 2022 07:33:28 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 10:19:31 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhang",
"Shijie",
""
],
[
"Yuan",
"Wei",
""
],
[
"Yin",
"Hongzhi",
""
]
] | TITLE: Comprehensive Privacy Analysis on Federated Recommender System against
Attribute Inference Attacks
ABSTRACT: In recent years, recommender systems are crucially important for the delivery
of personalized services that satisfy users' preferences. With personalized
recommendation services, users can enjoy a variety of recommendations such as
movies, books, ads, restaurants, and more. Despite the great benefits,
personalized recommendations typically require the collection of personal data
for user modelling and analysis, which can make users susceptible to attribute
inference attacks. Specifically, the vulnerability of existing centralized
recommenders under attribute inference attacks leaves malicious attackers a
backdoor to infer users' private attributes, as the systems remember
information of their training data (i.e., interaction data and side
information). An emerging practice is to implement recommender systems in the
federated setting, which enables all user devices to collaboratively learn a
shared global recommender while keeping all the training data on device.
However, the privacy issues in federated recommender systems have been rarely
explored. In this paper, we first design a novel attribute inference attacker
to perform a comprehensive privacy analysis of the state-of-the-art federated
recommender models. The experimental results show that the vulnerability of
each model component against attribute inference attack is varied, highlighting
the need for new defense approaches. Therefore, we propose a novel adaptive
privacy-preserving approach to protect users' sensitive data in the presence of
attribute inference attacks and meanwhile maximize the recommendation accuracy.
Extensive experimental results on two real-world datasets validate the superior
performance of our model on both recommendation effectiveness and resistance to
inference attacks.
| no_new_dataset | 0.944587 |
2206.01685 | Charlotte Caucheteux | Juliette Millet, Charlotte Caucheteux, Pierre Orhan, Yves Boubenec,
Alexandre Gramfort, Ewan Dunbar, Christophe Pallier, Jean-Remi King | Toward a realistic model of speech processing in the brain with
self-supervised learning | Accepted to NeurIPS 2022 | Neural Information Processing Systems (NeurIPS), 2022 | null | null | q-bio.NC cs.AI cs.CL | http://creativecommons.org/licenses/by/4.0/ | Several deep neural networks have recently been shown to generate activations
similar to those of the brain in response to the same input. These algorithms,
however, remain largely implausible: they require (1) extraordinarily large
amounts of data, (2) unobtainable supervised labels, (3) textual rather than
raw sensory input, and / or (4) implausibly large memory (e.g. thousands of
contextual words). These elements highlight the need to identify algorithms
that, under these limitations, would suffice to account for both behavioral and
brain responses. Focusing on the issue of speech processing, we here
hypothesize that self-supervised algorithms trained on the raw waveform
constitute a promising candidate. Specifically, we compare a recent
self-supervised architecture, Wav2Vec 2.0, to the brain activity of 412
English, French, and Mandarin individuals recorded with functional Magnetic
Resonance Imaging (fMRI), while they listened to ~1h of audio books. Our
results are four-fold. First, we show that this algorithm learns brain-like
representations with as little as 600 hours of unlabelled speech -- a quantity
comparable to what infants can be exposed to during language acquisition.
Second, its functional hierarchy aligns with the cortical hierarchy of speech
processing. Third, different training regimes reveal a functional
specialization akin to the cortex: Wav2Vec 2.0 learns sound-generic,
speech-specific and language-specific representations similar to those of the
prefrontal and temporal cortices. Fourth, we confirm the similarity of this
specialization with the behavior of 386 additional participants. These
elements, resulting from the largest neuroimaging benchmark to date, show how
self-supervised learning can account for a rich organization of speech
processing in the brain, and thus delineate a path to identify the laws of
language acquisition which shape the human brain.
| [
{
"version": "v1",
"created": "Fri, 3 Jun 2022 17:01:46 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 10:11:41 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Millet",
"Juliette",
""
],
[
"Caucheteux",
"Charlotte",
""
],
[
"Orhan",
"Pierre",
""
],
[
"Boubenec",
"Yves",
""
],
[
"Gramfort",
"Alexandre",
""
],
[
"Dunbar",
"Ewan",
""
],
[
"Pallier",
"Christophe",
""
],
[
"King",
"Jean-Remi",
""
]
] | TITLE: Toward a realistic model of speech processing in the brain with
self-supervised learning
ABSTRACT: Several deep neural networks have recently been shown to generate activations
similar to those of the brain in response to the same input. These algorithms,
however, remain largely implausible: they require (1) extraordinarily large
amounts of data, (2) unobtainable supervised labels, (3) textual rather than
raw sensory input, and / or (4) implausibly large memory (e.g. thousands of
contextual words). These elements highlight the need to identify algorithms
that, under these limitations, would suffice to account for both behavioral and
brain responses. Focusing on the issue of speech processing, we here
hypothesize that self-supervised algorithms trained on the raw waveform
constitute a promising candidate. Specifically, we compare a recent
self-supervised architecture, Wav2Vec 2.0, to the brain activity of 412
English, French, and Mandarin individuals recorded with functional Magnetic
Resonance Imaging (fMRI), while they listened to ~1h of audio books. Our
results are four-fold. First, we show that this algorithm learns brain-like
representations with as little as 600 hours of unlabelled speech -- a quantity
comparable to what infants can be exposed to during language acquisition.
Second, its functional hierarchy aligns with the cortical hierarchy of speech
processing. Third, different training regimes reveal a functional
specialization akin to the cortex: Wav2Vec 2.0 learns sound-generic,
speech-specific and language-specific representations similar to those of the
prefrontal and temporal cortices. Fourth, we confirm the similarity of this
specialization with the behavior of 386 additional participants. These
elements, resulting from the largest neuroimaging benchmark to date, show how
self-supervised learning can account for a rich organization of speech
processing in the brain, and thus delineate a path to identify the laws of
language acquisition which shape the human brain.
| no_new_dataset | 0.94079 |
2206.04803 | Mirko Zichichi | Nadia Pocher, Mirko Zichichi, Fabio Merizzi, Muhammad Zohaib Shafiq
and Stefano Ferretti | Detecting Anomalous Cryptocurrency Transactions: an AML/CFT Application
of Machine Learning-based Forensics | null | null | null | null | cs.CR cs.DC cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In shaping the Internet of Money, the application of blockchain and
distributed ledger technologies (DLTs) to the financial sector triggered
regulatory concerns. Notably, while the user anonymity enabled in this field
may safeguard privacy and data protection, the lack of identifiability hinders
accountability and challenges the fight against money laundering and the
financing of terrorism and proliferation (AML/CFT). As law enforcement agencies
and the private sector apply forensics to track crypto transfers across
ecosystems that are socio-technical in nature, this paper focuses on the
growing relevance of these techniques in a domain where their deployment
impacts the traits and evolution of the sphere. In particular, this work offers
contextualized insights into the application of methods of machine learning and
transaction graph analysis. Namely, it analyzes a real-world dataset of Bitcoin
transactions represented as a directed graph network through various
techniques. The modeling of blockchain transactions as a complex network
suggests that the use of graph-based data analysis methods can help classify
transactions and identify illicit ones. Indeed, this work shows that the neural
network types known as Graph Convolutional Networks (GCN) and Graph Attention
Networks (GAT) are a promising AML/CFT solution. Notably, in this scenario GCN
outperform other classic approaches and GAT are applied for the first time to
detect anomalies in Bitcoin. Ultimately, the paper upholds the value of
public-private synergies to devise forensic strategies conscious of the spirit
of explainability and data openness.
| [
{
"version": "v1",
"created": "Tue, 7 Jun 2022 16:22:55 GMT"
},
{
"version": "v2",
"created": "Wed, 26 Oct 2022 14:20:55 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 14:41:31 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Pocher",
"Nadia",
""
],
[
"Zichichi",
"Mirko",
""
],
[
"Merizzi",
"Fabio",
""
],
[
"Shafiq",
"Muhammad Zohaib",
""
],
[
"Ferretti",
"Stefano",
""
]
] | TITLE: Detecting Anomalous Cryptocurrency Transactions: an AML/CFT Application
of Machine Learning-based Forensics
ABSTRACT: In shaping the Internet of Money, the application of blockchain and
distributed ledger technologies (DLTs) to the financial sector triggered
regulatory concerns. Notably, while the user anonymity enabled in this field
may safeguard privacy and data protection, the lack of identifiability hinders
accountability and challenges the fight against money laundering and the
financing of terrorism and proliferation (AML/CFT). As law enforcement agencies
and the private sector apply forensics to track crypto transfers across
ecosystems that are socio-technical in nature, this paper focuses on the
growing relevance of these techniques in a domain where their deployment
impacts the traits and evolution of the sphere. In particular, this work offers
contextualized insights into the application of methods of machine learning and
transaction graph analysis. Namely, it analyzes a real-world dataset of Bitcoin
transactions represented as a directed graph network through various
techniques. The modeling of blockchain transactions as a complex network
suggests that the use of graph-based data analysis methods can help classify
transactions and identify illicit ones. Indeed, this work shows that the neural
network types known as Graph Convolutional Networks (GCN) and Graph Attention
Networks (GAT) are a promising AML/CFT solution. Notably, in this scenario GCN
outperform other classic approaches and GAT are applied for the first time to
detect anomalies in Bitcoin. Ultimately, the paper upholds the value of
public-private synergies to devise forensic strategies conscious of the spirit
of explainability and data openness.
| no_new_dataset | 0.950365 |
2207.01331 | Wenyu Liu | Wenyu Liu, Wentong Li, Jianke Zhu, Miaomiao Cui, Xuansong Xie, Lei
Zhang | Improving Nighttime Driving-Scene Segmentation via Dual Image-adaptive
Learnable Filters | Accepted by IEEE TCSVT(2023) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semantic segmentation on driving-scene images is vital for autonomous
driving. Although encouraging performance has been achieved on daytime images,
the performance on nighttime images are less satisfactory due to the
insufficient exposure and the lack of labeled data. To address these issues, we
present an add-on module called dual image-adaptive learnable filters
(DIAL-Filters) to improve the semantic segmentation in nighttime driving
conditions, aiming at exploiting the intrinsic features of driving-scene images
under different illuminations. DIAL-Filters consist of two parts, including an
image-adaptive processing module (IAPM) and a learnable guided filter (LGF).
With DIAL-Filters, we design both unsupervised and supervised frameworks for
nighttime driving-scene segmentation, which can be trained in an end-to-end
manner. Specifically, the IAPM module consists of a small convolutional neural
network with a set of differentiable image filters, where each image can be
adaptively enhanced for better segmentation with respect to the different
illuminations. The LGF is employed to enhance the output of segmentation
network to get the final segmentation result. The DIAL-Filters are light-weight
and efficient and they can be readily applied for both daytime and nighttime
images. Our experiments show that DAIL-Filters can significantly improve the
supervised segmentation performance on ACDC_Night and NightCity datasets, while
it demonstrates the state-of-the-art performance on unsupervised nighttime
semantic segmentation on Dark Zurich and Nighttime Driving testbeds.
| [
{
"version": "v1",
"created": "Mon, 4 Jul 2022 11:24:10 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 07:50:13 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Liu",
"Wenyu",
""
],
[
"Li",
"Wentong",
""
],
[
"Zhu",
"Jianke",
""
],
[
"Cui",
"Miaomiao",
""
],
[
"Xie",
"Xuansong",
""
],
[
"Zhang",
"Lei",
""
]
] | TITLE: Improving Nighttime Driving-Scene Segmentation via Dual Image-adaptive
Learnable Filters
ABSTRACT: Semantic segmentation on driving-scene images is vital for autonomous
driving. Although encouraging performance has been achieved on daytime images,
the performance on nighttime images are less satisfactory due to the
insufficient exposure and the lack of labeled data. To address these issues, we
present an add-on module called dual image-adaptive learnable filters
(DIAL-Filters) to improve the semantic segmentation in nighttime driving
conditions, aiming at exploiting the intrinsic features of driving-scene images
under different illuminations. DIAL-Filters consist of two parts, including an
image-adaptive processing module (IAPM) and a learnable guided filter (LGF).
With DIAL-Filters, we design both unsupervised and supervised frameworks for
nighttime driving-scene segmentation, which can be trained in an end-to-end
manner. Specifically, the IAPM module consists of a small convolutional neural
network with a set of differentiable image filters, where each image can be
adaptively enhanced for better segmentation with respect to the different
illuminations. The LGF is employed to enhance the output of segmentation
network to get the final segmentation result. The DIAL-Filters are light-weight
and efficient and they can be readily applied for both daytime and nighttime
images. Our experiments show that DAIL-Filters can significantly improve the
supervised segmentation performance on ACDC_Night and NightCity datasets, while
it demonstrates the state-of-the-art performance on unsupervised nighttime
semantic segmentation on Dark Zurich and Nighttime Driving testbeds.
| no_new_dataset | 0.942823 |
2207.08536 | Xi Li | Zequn Qin, Jingyu Chen, Chao Chen, Xiaozhi Chen, Xi Li | UniFusion: Unified Multi-view Fusion Transformer for Spatial-Temporal
Representation in Bird's-Eye-View | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bird's eye view (BEV) representation is a new perception formulation for
autonomous driving, which is based on spatial fusion. Further, temporal fusion
is also introduced in BEV representation and gains great success. In this work,
we propose a new method that unifies both spatial and temporal fusion and
merges them into a unified mathematical formulation. The unified fusion could
not only provide a new perspective on BEV fusion but also brings new
capabilities. With the proposed unified spatial-temporal fusion, our method
could support long-range fusion, which is hard to achieve in conventional BEV
methods. Moreover, the BEV fusion in our work is temporal-adaptive and the
weights of temporal fusion are learnable. In contrast, conventional methods
mainly use fixed and equal weights for temporal fusion. Besides, the proposed
unified fusion could avoid information lost in conventional BEV fusion methods
and make full use of features. Extensive experiments and ablation studies on
the NuScenes dataset show the effectiveness of the proposed method and our
method gains the state-of-the-art performance in the map segmentation task.
| [
{
"version": "v1",
"created": "Mon, 18 Jul 2022 11:59:10 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 03:12:09 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Qin",
"Zequn",
""
],
[
"Chen",
"Jingyu",
""
],
[
"Chen",
"Chao",
""
],
[
"Chen",
"Xiaozhi",
""
],
[
"Li",
"Xi",
""
]
] | TITLE: UniFusion: Unified Multi-view Fusion Transformer for Spatial-Temporal
Representation in Bird's-Eye-View
ABSTRACT: Bird's eye view (BEV) representation is a new perception formulation for
autonomous driving, which is based on spatial fusion. Further, temporal fusion
is also introduced in BEV representation and gains great success. In this work,
we propose a new method that unifies both spatial and temporal fusion and
merges them into a unified mathematical formulation. The unified fusion could
not only provide a new perspective on BEV fusion but also brings new
capabilities. With the proposed unified spatial-temporal fusion, our method
could support long-range fusion, which is hard to achieve in conventional BEV
methods. Moreover, the BEV fusion in our work is temporal-adaptive and the
weights of temporal fusion are learnable. In contrast, conventional methods
mainly use fixed and equal weights for temporal fusion. Besides, the proposed
unified fusion could avoid information lost in conventional BEV fusion methods
and make full use of features. Extensive experiments and ablation studies on
the NuScenes dataset show the effectiveness of the proposed method and our
method gains the state-of-the-art performance in the map segmentation task.
| no_new_dataset | 0.949529 |
2209.08543 | Tianyue Wu | Tianyue Wu and Fei Gao | A Decoupled and Linear Framework for Global Outlier Rejection over
Planar Pose Graph | 7 pages, 4 figures. To appear in ICRA 2023 | null | null | null | cs.RO math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a robust framework for the planar pose graph optimization
contaminated by loop closure outliers. Our framework rejects outliers by first
decoupling the robust PGO problem wrapped by a Truncated Least Squares kernel
into two subproblems. Then, the framework introduces a linear angle
representation to rewrite the first subproblem that is originally formulated
with rotation matrices. The framework is configured with the Graduated
Non-Convexity (GNC) algorithm to solve the two non-convex subproblems in
succession without initial guesses. Thanks to the linearity properties of both
the subproblems, our framework requires only linear solvers to optimally solve
the optimization problems encountered in GNC. We extensively validate the
proposed framework, named DEGNC-LAF (DEcoupled Graduated Non-Convexity with
Linear Angle Formulation) in planar PGO benchmarks. It turns out that it runs
significantly (sometimes up to over 30 times) faster than the standard and
general-purpose GNC while resulting in high-quality estimates.
| [
{
"version": "v1",
"created": "Sun, 18 Sep 2022 11:49:44 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 04:31:35 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Wu",
"Tianyue",
""
],
[
"Gao",
"Fei",
""
]
] | TITLE: A Decoupled and Linear Framework for Global Outlier Rejection over
Planar Pose Graph
ABSTRACT: We propose a robust framework for the planar pose graph optimization
contaminated by loop closure outliers. Our framework rejects outliers by first
decoupling the robust PGO problem wrapped by a Truncated Least Squares kernel
into two subproblems. Then, the framework introduces a linear angle
representation to rewrite the first subproblem that is originally formulated
with rotation matrices. The framework is configured with the Graduated
Non-Convexity (GNC) algorithm to solve the two non-convex subproblems in
succession without initial guesses. Thanks to the linearity properties of both
the subproblems, our framework requires only linear solvers to optimally solve
the optimization problems encountered in GNC. We extensively validate the
proposed framework, named DEGNC-LAF (DEcoupled Graduated Non-Convexity with
Linear Angle Formulation) in planar PGO benchmarks. It turns out that it runs
significantly (sometimes up to over 30 times) faster than the standard and
general-purpose GNC while resulting in high-quality estimates.
| no_new_dataset | 0.945096 |
2209.09419 | Tianpeng Zhang | Tianpeng Zhang (1), Kasper Johansson (2), Na Li (1)((1) Harvard
University, (2) Stanford University) | Multi-armed Bandit Learning on a Graph | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The multi-armed bandit(MAB) problem is a simple yet powerful framework that
has been extensively studied in the context of decision-making under
uncertainty. In many real-world applications, such as robotic applications,
selecting an arm corresponds to a physical action that constrains the choices
of the next available arms (actions). Motivated by this, we study an extension
of MAB called the graph bandit, where an agent travels over a graph to maximize
the reward collected from different nodes. The graph defines the agent's
freedom in selecting the next available nodes at each step. We assume the graph
structure is fully available, but the reward distributions are unknown. Built
upon an offline graph-based planning algorithm and the principle of optimism,
we design a learning algorithm, G-UCB, that balances long-term
exploration-exploitation using the principle of optimism. We show that our
proposed algorithm achieves $O(\sqrt{|S|T\log(T)}+D|S|\log T)$ learning regret,
where $|S|$ is the number of nodes and $D$ is the diameter of the graph, which
matches the theoretical lower bound $\Omega(\sqrt{|S|T})$ up to logarithmic
factors. To our knowledge, this result is among the first tight regret bounds
in non-episodic, un-discounted learning problems with known deterministic
transitions. Numerical experiments confirm that our algorithm outperforms
several benchmarks.
| [
{
"version": "v1",
"created": "Tue, 20 Sep 2022 02:31:42 GMT"
},
{
"version": "v2",
"created": "Tue, 13 Dec 2022 18:21:31 GMT"
},
{
"version": "v3",
"created": "Wed, 14 Dec 2022 01:57:07 GMT"
},
{
"version": "v4",
"created": "Mon, 20 Mar 2023 17:06:12 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhang",
"Tianpeng",
""
],
[
"Johansson",
"Kasper",
""
],
[
"Li",
"Na",
""
]
] | TITLE: Multi-armed Bandit Learning on a Graph
ABSTRACT: The multi-armed bandit(MAB) problem is a simple yet powerful framework that
has been extensively studied in the context of decision-making under
uncertainty. In many real-world applications, such as robotic applications,
selecting an arm corresponds to a physical action that constrains the choices
of the next available arms (actions). Motivated by this, we study an extension
of MAB called the graph bandit, where an agent travels over a graph to maximize
the reward collected from different nodes. The graph defines the agent's
freedom in selecting the next available nodes at each step. We assume the graph
structure is fully available, but the reward distributions are unknown. Built
upon an offline graph-based planning algorithm and the principle of optimism,
we design a learning algorithm, G-UCB, that balances long-term
exploration-exploitation using the principle of optimism. We show that our
proposed algorithm achieves $O(\sqrt{|S|T\log(T)}+D|S|\log T)$ learning regret,
where $|S|$ is the number of nodes and $D$ is the diameter of the graph, which
matches the theoretical lower bound $\Omega(\sqrt{|S|T})$ up to logarithmic
factors. To our knowledge, this result is among the first tight regret bounds
in non-episodic, un-discounted learning problems with known deterministic
transitions. Numerical experiments confirm that our algorithm outperforms
several benchmarks.
| no_new_dataset | 0.942454 |
2209.09659 | Thorbj{\o}rn Mosekj{\ae}r Iversen | Thorbj{\o}rn Mosekj{\ae}r Iversen, Rasmus Laurvig Haugaard, Anders
Glent Buch | Ki-Pode: Keypoint-based Implicit Pose Distribution Estimation of Rigid
Objects | 11 pages, 2 figures | The 33rd British Machine Vision Conference Proceedings: BMVC 2022 | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The estimation of 6D poses of rigid objects is a fundamental problem in
computer vision. Traditionally pose estimation is concerned with the
determination of a single best estimate. However, a single estimate is unable
to express visual ambiguity, which in many cases is unavoidable due to object
symmetries or occlusion of identifying features. Inability to account for
ambiguities in pose can lead to failure in subsequent methods, which is
unacceptable when the cost of failure is high. Estimates of full pose
distributions are, contrary to single estimates, well suited for expressing
uncertainty on pose. Motivated by this, we propose a novel pose distribution
estimation method. An implicit formulation of the probability distribution over
object pose is derived from an intermediary representation of an object as a
set of keypoints. This ensures that the pose distribution estimates have a high
level of interpretability. Furthermore, our method is based on conservative
approximations, which leads to reliable estimates. The method has been
evaluated on the task of rotation distribution estimation on the YCB-V and
T-LESS datasets and performs reliably on all objects.
| [
{
"version": "v1",
"created": "Tue, 20 Sep 2022 11:59:05 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Iversen",
"Thorbjørn Mosekjær",
""
],
[
"Haugaard",
"Rasmus Laurvig",
""
],
[
"Buch",
"Anders Glent",
""
]
] | TITLE: Ki-Pode: Keypoint-based Implicit Pose Distribution Estimation of Rigid
Objects
ABSTRACT: The estimation of 6D poses of rigid objects is a fundamental problem in
computer vision. Traditionally pose estimation is concerned with the
determination of a single best estimate. However, a single estimate is unable
to express visual ambiguity, which in many cases is unavoidable due to object
symmetries or occlusion of identifying features. Inability to account for
ambiguities in pose can lead to failure in subsequent methods, which is
unacceptable when the cost of failure is high. Estimates of full pose
distributions are, contrary to single estimates, well suited for expressing
uncertainty on pose. Motivated by this, we propose a novel pose distribution
estimation method. An implicit formulation of the probability distribution over
object pose is derived from an intermediary representation of an object as a
set of keypoints. This ensures that the pose distribution estimates have a high
level of interpretability. Furthermore, our method is based on conservative
approximations, which leads to reliable estimates. The method has been
evaluated on the task of rotation distribution estimation on the YCB-V and
T-LESS datasets and performs reliably on all objects.
| no_new_dataset | 0.946051 |
2209.15214 | Shumin Deng | Shumin Deng, Chengming Wang, Zhoubo Li, Ningyu Zhang, Zelin Dai,
Hehong Chen, Feiyu Xiong, Ming Yan, Qiang Chen, Mosha Chen, Jiaoyan Chen,
Jeff Z. Pan, Bryan Hooi, Huajun Chen | Construction and Applications of Billion-Scale Pre-Trained Multimodal
Business Knowledge Graph | OpenBG. Accepted by ICDE 2023. The project is released at
https://github.com/OpenBGBenchmark/OpenBG . Website: https://kg.alibaba.com/
, Leaderboard: https://tianchi.aliyun.com/dataset/dataDetail?dataId=122271 | null | null | null | cs.AI cs.CL cs.IR cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Business Knowledge Graphs (KGs) are important to many enterprises today,
providing factual knowledge and structured data that steer many products and
make them more intelligent. Despite their promising benefits, building business
KG necessitates solving prohibitive issues of deficient structure and multiple
modalities. In this paper, we advance the understanding of the practical
challenges related to building KG in non-trivial real-world systems. We
introduce the process of building an open business knowledge graph (OpenBG)
derived from a well-known enterprise, Alibaba Group. Specifically, we define a
core ontology to cover various abstract products and consumption demands, with
fine-grained taxonomy and multimodal facts in deployed applications. OpenBG is
an open business KG of unprecedented scale: 2.6 billion triples with more than
88 million entities covering over 1 million core classes/concepts and 2,681
types of relations. We release all the open resources (OpenBG benchmarks)
derived from it for the community and report experimental results of KG-centric
tasks. We also run up an online competition based on OpenBG benchmarks, and has
attracted thousands of teams. We further pre-train OpenBG and apply it to many
KG- enhanced downstream tasks in business scenarios, demonstrating the
effectiveness of billion-scale multimodal knowledge for e-commerce. All the
resources with codes have been released at
\url{https://github.com/OpenBGBenchmark/OpenBG}.
| [
{
"version": "v1",
"created": "Fri, 30 Sep 2022 04:03:26 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Oct 2022 12:24:19 GMT"
},
{
"version": "v3",
"created": "Thu, 27 Oct 2022 09:14:02 GMT"
},
{
"version": "v4",
"created": "Wed, 15 Feb 2023 04:14:39 GMT"
},
{
"version": "v5",
"created": "Sat, 25 Feb 2023 06:55:52 GMT"
},
{
"version": "v6",
"created": "Sun, 19 Mar 2023 11:28:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Deng",
"Shumin",
""
],
[
"Wang",
"Chengming",
""
],
[
"Li",
"Zhoubo",
""
],
[
"Zhang",
"Ningyu",
""
],
[
"Dai",
"Zelin",
""
],
[
"Chen",
"Hehong",
""
],
[
"Xiong",
"Feiyu",
""
],
[
"Yan",
"Ming",
""
],
[
"Chen",
"Qiang",
""
],
[
"Chen",
"Mosha",
""
],
[
"Chen",
"Jiaoyan",
""
],
[
"Pan",
"Jeff Z.",
""
],
[
"Hooi",
"Bryan",
""
],
[
"Chen",
"Huajun",
""
]
] | TITLE: Construction and Applications of Billion-Scale Pre-Trained Multimodal
Business Knowledge Graph
ABSTRACT: Business Knowledge Graphs (KGs) are important to many enterprises today,
providing factual knowledge and structured data that steer many products and
make them more intelligent. Despite their promising benefits, building business
KG necessitates solving prohibitive issues of deficient structure and multiple
modalities. In this paper, we advance the understanding of the practical
challenges related to building KG in non-trivial real-world systems. We
introduce the process of building an open business knowledge graph (OpenBG)
derived from a well-known enterprise, Alibaba Group. Specifically, we define a
core ontology to cover various abstract products and consumption demands, with
fine-grained taxonomy and multimodal facts in deployed applications. OpenBG is
an open business KG of unprecedented scale: 2.6 billion triples with more than
88 million entities covering over 1 million core classes/concepts and 2,681
types of relations. We release all the open resources (OpenBG benchmarks)
derived from it for the community and report experimental results of KG-centric
tasks. We also run up an online competition based on OpenBG benchmarks, and has
attracted thousands of teams. We further pre-train OpenBG and apply it to many
KG- enhanced downstream tasks in business scenarios, demonstrating the
effectiveness of billion-scale multimodal knowledge for e-commerce. All the
resources with codes have been released at
\url{https://github.com/OpenBGBenchmark/OpenBG}.
| no_new_dataset | 0.940953 |
2210.04227 | Yu Cai | Yu Cai, Hao Chen, Xin Yang, Yu Zhou, Kwang-Ting Cheng | Dual-distribution discrepancy with self-supervised refinement for
anomaly detection in medical images | Accepted to Medical Image Analysis, 2023 | null | 10.1016/j.media.2023.102794 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Medical anomaly detection is a crucial yet challenging task aimed at
recognizing abnormal images to assist in diagnosis. Due to the high-cost
annotations of abnormal images, most methods utilize only known normal images
during training and identify samples deviating from the normal profile as
anomalies in the testing phase. Many readily available unlabeled images
containing anomalies are thus ignored in the training phase, restricting the
performance. To solve this problem, we introduce one-class semi-supervised
learning (OC-SSL) to utilize known normal and unlabeled images for training,
and propose Dual-distribution Discrepancy for Anomaly Detection (DDAD) based on
this setting. Ensembles of reconstruction networks are designed to model the
distribution of normal images and the distribution of both normal and unlabeled
images, deriving the normative distribution module (NDM) and unknown
distribution module (UDM). Subsequently, the intra-discrepancy of NDM and
inter-discrepancy between the two modules are designed as anomaly scores.
Furthermore, we propose a new perspective on self-supervised learning, which is
designed to refine the anomaly scores rather than detect anomalies directly.
Five medical datasets, including chest X-rays, brain MRIs and retinal fundus
images, are organized as benchmarks for evaluation. Experiments on these
benchmarks comprehensively compare a wide range of anomaly detection methods
and demonstrate that our method achieves significant gains and outperforms the
state-of-the-art. Code and organized benchmarks are available at
https://github.com/caiyu6666/DDAD-ASR.
| [
{
"version": "v1",
"created": "Sun, 9 Oct 2022 11:18:45 GMT"
},
{
"version": "v2",
"created": "Tue, 21 Feb 2023 08:15:23 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 09:13:09 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Cai",
"Yu",
""
],
[
"Chen",
"Hao",
""
],
[
"Yang",
"Xin",
""
],
[
"Zhou",
"Yu",
""
],
[
"Cheng",
"Kwang-Ting",
""
]
] | TITLE: Dual-distribution discrepancy with self-supervised refinement for
anomaly detection in medical images
ABSTRACT: Medical anomaly detection is a crucial yet challenging task aimed at
recognizing abnormal images to assist in diagnosis. Due to the high-cost
annotations of abnormal images, most methods utilize only known normal images
during training and identify samples deviating from the normal profile as
anomalies in the testing phase. Many readily available unlabeled images
containing anomalies are thus ignored in the training phase, restricting the
performance. To solve this problem, we introduce one-class semi-supervised
learning (OC-SSL) to utilize known normal and unlabeled images for training,
and propose Dual-distribution Discrepancy for Anomaly Detection (DDAD) based on
this setting. Ensembles of reconstruction networks are designed to model the
distribution of normal images and the distribution of both normal and unlabeled
images, deriving the normative distribution module (NDM) and unknown
distribution module (UDM). Subsequently, the intra-discrepancy of NDM and
inter-discrepancy between the two modules are designed as anomaly scores.
Furthermore, we propose a new perspective on self-supervised learning, which is
designed to refine the anomaly scores rather than detect anomalies directly.
Five medical datasets, including chest X-rays, brain MRIs and retinal fundus
images, are organized as benchmarks for evaluation. Experiments on these
benchmarks comprehensively compare a wide range of anomaly detection methods
and demonstrate that our method achieves significant gains and outperforms the
state-of-the-art. Code and organized benchmarks are available at
https://github.com/caiyu6666/DDAD-ASR.
| no_new_dataset | 0.950595 |
2210.07316 | Niklas Muennighoff | Niklas Muennighoff, Nouamane Tazi, Lo\"ic Magne, Nils Reimers | MTEB: Massive Text Embedding Benchmark | 24 pages, 14 tables, 6 figures | null | null | null | cs.CL cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text embeddings are commonly evaluated on a small set of datasets from a
single task not covering their possible applications to other tasks. It is
unclear whether state-of-the-art embeddings on semantic textual similarity
(STS) can be equally well applied to other tasks like clustering or reranking.
This makes progress in the field difficult to track, as various models are
constantly being proposed without proper evaluation. To solve this problem, we
introduce the Massive Text Embedding Benchmark (MTEB). MTEB spans 8 embedding
tasks covering a total of 58 datasets and 112 languages. Through the
benchmarking of 33 models on MTEB, we establish the most comprehensive
benchmark of text embeddings to date. We find that no particular text embedding
method dominates across all tasks. This suggests that the field has yet to
converge on a universal text embedding method and scale it up sufficiently to
provide state-of-the-art results on all embedding tasks. MTEB comes with
open-source code and a public leaderboard at
https://github.com/embeddings-benchmark/mteb.
| [
{
"version": "v1",
"created": "Thu, 13 Oct 2022 19:42:08 GMT"
},
{
"version": "v2",
"created": "Sun, 5 Feb 2023 15:59:49 GMT"
},
{
"version": "v3",
"created": "Sun, 19 Mar 2023 13:37:01 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Muennighoff",
"Niklas",
""
],
[
"Tazi",
"Nouamane",
""
],
[
"Magne",
"Loïc",
""
],
[
"Reimers",
"Nils",
""
]
] | TITLE: MTEB: Massive Text Embedding Benchmark
ABSTRACT: Text embeddings are commonly evaluated on a small set of datasets from a
single task not covering their possible applications to other tasks. It is
unclear whether state-of-the-art embeddings on semantic textual similarity
(STS) can be equally well applied to other tasks like clustering or reranking.
This makes progress in the field difficult to track, as various models are
constantly being proposed without proper evaluation. To solve this problem, we
introduce the Massive Text Embedding Benchmark (MTEB). MTEB spans 8 embedding
tasks covering a total of 58 datasets and 112 languages. Through the
benchmarking of 33 models on MTEB, we establish the most comprehensive
benchmark of text embeddings to date. We find that no particular text embedding
method dominates across all tasks. This suggests that the field has yet to
converge on a universal text embedding method and scale it up sufficiently to
provide state-of-the-art results on all embedding tasks. MTEB comes with
open-source code and a public leaderboard at
https://github.com/embeddings-benchmark/mteb.
| no_new_dataset | 0.746278 |
2210.08817 | Yang Deng | Yang Deng, Wenqiang Lei, Wenxuan Zhang, Wai Lam, Tat-Seng Chua | PACIFIC: Towards Proactive Conversational Question Answering over
Tabular and Textual Data in Finance | Accepted by EMNLP 2022 (main conference) | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | To facilitate conversational question answering (CQA) over hybrid contexts in
finance, we present a new dataset, named PACIFIC. Compared with existing CQA
datasets, PACIFIC exhibits three key features: (i) proactivity, (ii) numerical
reasoning, and (iii) hybrid context of tables and text. A new task is defined
accordingly to study Proactive Conversational Question Answering (PCQA), which
combines clarification question generation and CQA. In addition, we propose a
novel method, namely UniPCQA, to adapt a hybrid format of input and output
content in PCQA into the Seq2Seq problem, including the reformulation of the
numerical reasoning process as code generation. UniPCQA performs multi-task
learning over all sub-tasks in PCQA and incorporates a simple ensemble strategy
to alleviate the error propagation issue in the multi-task learning by
cross-validating top-$k$ sampled Seq2Seq outputs. We benchmark the PACIFIC
dataset with extensive baselines and provide comprehensive evaluations on each
sub-task of PCQA.
| [
{
"version": "v1",
"created": "Mon, 17 Oct 2022 08:06:56 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 03:39:16 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Deng",
"Yang",
""
],
[
"Lei",
"Wenqiang",
""
],
[
"Zhang",
"Wenxuan",
""
],
[
"Lam",
"Wai",
""
],
[
"Chua",
"Tat-Seng",
""
]
] | TITLE: PACIFIC: Towards Proactive Conversational Question Answering over
Tabular and Textual Data in Finance
ABSTRACT: To facilitate conversational question answering (CQA) over hybrid contexts in
finance, we present a new dataset, named PACIFIC. Compared with existing CQA
datasets, PACIFIC exhibits three key features: (i) proactivity, (ii) numerical
reasoning, and (iii) hybrid context of tables and text. A new task is defined
accordingly to study Proactive Conversational Question Answering (PCQA), which
combines clarification question generation and CQA. In addition, we propose a
novel method, namely UniPCQA, to adapt a hybrid format of input and output
content in PCQA into the Seq2Seq problem, including the reformulation of the
numerical reasoning process as code generation. UniPCQA performs multi-task
learning over all sub-tasks in PCQA and incorporates a simple ensemble strategy
to alleviate the error propagation issue in the multi-task learning by
cross-validating top-$k$ sampled Seq2Seq outputs. We benchmark the PACIFIC
dataset with extensive baselines and provide comprehensive evaluations on each
sub-task of PCQA.
| new_dataset | 0.948965 |
2211.00692 | Sadegh Mahdavi | Sadegh Mahdavi, Kevin Swersky, Thomas Kipf, Milad Hashemi, Christos
Thrampoulidis, Renjie Liao | Towards Better Out-of-Distribution Generalization of Neural Algorithmic
Reasoning Tasks | Transactions on Machine Learning Research (TMLR), 2023 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In this paper, we study the OOD generalization of neural algorithmic
reasoning tasks, where the goal is to learn an algorithm (e.g., sorting,
breadth-first search, and depth-first search) from input-output pairs using
deep neural networks. First, we argue that OOD generalization in this setting
is significantly different than common OOD settings. For example, some
phenomena in OOD generalization of image classifications such as \emph{accuracy
on the line} are not observed here, and techniques such as data augmentation
methods do not help as assumptions underlying many augmentation techniques are
often violated. Second, we analyze the main challenges (e.g., input
distribution shift, non-representative data generation, and uninformative
validation metrics) of the current leading benchmark, i.e., CLRS
\citep{deepmind2021clrs}, which contains 30 algorithmic reasoning tasks. We
propose several solutions, including a simple-yet-effective fix to the input
distribution shift and improved data generation. Finally, we propose an
attention-based 2WL-graph neural network (GNN) processor which complements
message-passing GNNs so their combination outperforms the state-of-the-art
model by a 3% margin averaged over all algorithms. Our code is available at:
\url{https://github.com/smahdavi4/clrs}.
| [
{
"version": "v1",
"created": "Tue, 1 Nov 2022 18:33:20 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 08:23:33 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Mahdavi",
"Sadegh",
""
],
[
"Swersky",
"Kevin",
""
],
[
"Kipf",
"Thomas",
""
],
[
"Hashemi",
"Milad",
""
],
[
"Thrampoulidis",
"Christos",
""
],
[
"Liao",
"Renjie",
""
]
] | TITLE: Towards Better Out-of-Distribution Generalization of Neural Algorithmic
Reasoning Tasks
ABSTRACT: In this paper, we study the OOD generalization of neural algorithmic
reasoning tasks, where the goal is to learn an algorithm (e.g., sorting,
breadth-first search, and depth-first search) from input-output pairs using
deep neural networks. First, we argue that OOD generalization in this setting
is significantly different than common OOD settings. For example, some
phenomena in OOD generalization of image classifications such as \emph{accuracy
on the line} are not observed here, and techniques such as data augmentation
methods do not help as assumptions underlying many augmentation techniques are
often violated. Second, we analyze the main challenges (e.g., input
distribution shift, non-representative data generation, and uninformative
validation metrics) of the current leading benchmark, i.e., CLRS
\citep{deepmind2021clrs}, which contains 30 algorithmic reasoning tasks. We
propose several solutions, including a simple-yet-effective fix to the input
distribution shift and improved data generation. Finally, we propose an
attention-based 2WL-graph neural network (GNN) processor which complements
message-passing GNNs so their combination outperforms the state-of-the-art
model by a 3% margin averaged over all algorithms. Our code is available at:
\url{https://github.com/smahdavi4/clrs}.
| no_new_dataset | 0.948346 |
2211.02642 | Abdellah Rahmani | Abdellah Rahmani, Arun Venkitaraman, Pascal Frossard | A Meta-GNN approach to personalized seizure detection and classification | null | null | null | null | eess.SP cs.LG | http://creativecommons.org/licenses/by/4.0/ | In this paper, we propose a personalized seizure detection and classification
framework that quickly adapts to a specific patient from limited seizure
samples. We achieve this by combining two novel paradigms that have recently
seen much success in a wide variety of real-world applications: graph neural
networks (GNN), and meta-learning. We train a Meta-GNN based classifier that
learns a global model from a set of training patients such that this global
model can eventually be adapted to a new unseen patient using very limited
samples. We apply our approach on the TUSZ-dataset, one of the largest and
publicly available benchmark datasets for epilepsy. We show that our method
outperforms the baselines by reaching 82.7% on accuracy and 82.08% on F1 score
after only 20 iterations on new unseen patients.
| [
{
"version": "v1",
"created": "Tue, 1 Nov 2022 14:12:58 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 16:08:30 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Rahmani",
"Abdellah",
""
],
[
"Venkitaraman",
"Arun",
""
],
[
"Frossard",
"Pascal",
""
]
] | TITLE: A Meta-GNN approach to personalized seizure detection and classification
ABSTRACT: In this paper, we propose a personalized seizure detection and classification
framework that quickly adapts to a specific patient from limited seizure
samples. We achieve this by combining two novel paradigms that have recently
seen much success in a wide variety of real-world applications: graph neural
networks (GNN), and meta-learning. We train a Meta-GNN based classifier that
learns a global model from a set of training patients such that this global
model can eventually be adapted to a new unseen patient using very limited
samples. We apply our approach on the TUSZ-dataset, one of the largest and
publicly available benchmark datasets for epilepsy. We show that our method
outperforms the baselines by reaching 82.7% on accuracy and 82.08% on F1 score
after only 20 iterations on new unseen patients.
| no_new_dataset | 0.952618 |
2211.04473 | Anton Ratnarajah Mr | Anton Ratnarajah, Ishwarya Ananthabhotla, Vamsi Krishna Ithapu, Pablo
Hoffmann, Dinesh Manocha, Paul Calamia | Towards Improved Room Impulse Response Estimation for Speech Recognition | Accepted at ICASSP 2023. More results are available at
https://anton-jeran.github.io/S2IR/ | null | null | null | cs.SD cs.AI eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a novel approach for blind room impulse response (RIR) estimation
systems in the context of a downstream application scenario, far-field
automatic speech recognition (ASR). We first draw the connection between
improved RIR estimation and improved ASR performance, as a means of evaluating
neural RIR estimators. We then propose a generative adversarial network (GAN)
based architecture that encodes RIR features from reverberant speech and
constructs an RIR from the encoded features, and uses a novel energy decay
relief loss to optimize for capturing energy-based properties of the input
reverberant speech. We show that our model outperforms the state-of-the-art
baselines on acoustic benchmarks (by 17\% on the energy decay relief and 22\%
on an early-reflection energy metric), as well as in an ASR evaluation task (by
6.9\% in word error rate).
| [
{
"version": "v1",
"created": "Tue, 8 Nov 2022 00:40:27 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 20:23:19 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Ratnarajah",
"Anton",
""
],
[
"Ananthabhotla",
"Ishwarya",
""
],
[
"Ithapu",
"Vamsi Krishna",
""
],
[
"Hoffmann",
"Pablo",
""
],
[
"Manocha",
"Dinesh",
""
],
[
"Calamia",
"Paul",
""
]
] | TITLE: Towards Improved Room Impulse Response Estimation for Speech Recognition
ABSTRACT: We propose a novel approach for blind room impulse response (RIR) estimation
systems in the context of a downstream application scenario, far-field
automatic speech recognition (ASR). We first draw the connection between
improved RIR estimation and improved ASR performance, as a means of evaluating
neural RIR estimators. We then propose a generative adversarial network (GAN)
based architecture that encodes RIR features from reverberant speech and
constructs an RIR from the encoded features, and uses a novel energy decay
relief loss to optimize for capturing energy-based properties of the input
reverberant speech. We show that our model outperforms the state-of-the-art
baselines on acoustic benchmarks (by 17\% on the energy decay relief and 22\%
on an early-reflection energy metric), as well as in an ASR evaluation task (by
6.9\% in word error rate).
| no_new_dataset | 0.94801 |
2211.05405 | Nghia Hieu Nguyen | Nghia Hieu Nguyen, Duong T.D. Vo, Minh-Quan Ha | VieCap4H-VLSP 2021: ObjectAoA-Enhancing performance of Object Relation
Transformer with Attention on Attention for Vietnamese image captioning | Accepted for publishing at the VNU Journal of Science: Computer
Science and Communication Engineering | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Image captioning is currently a challenging task that requires the ability to
both understand visual information and use human language to describe this
visual information in the image. In this paper, we propose an efficient way to
improve the image understanding ability of transformer-based method by
extending Object Relation Transformer architecture with Attention on Attention
mechanism. Experiments on the VieCap4H dataset show that our proposed method
significantly outperforms its original structure on both the public test and
private test of the Image Captioning shared task held by VLSP.
| [
{
"version": "v1",
"created": "Thu, 10 Nov 2022 08:19:44 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Nov 2022 08:10:13 GMT"
},
{
"version": "v3",
"created": "Mon, 27 Feb 2023 03:35:55 GMT"
},
{
"version": "v4",
"created": "Mon, 20 Mar 2023 08:29:29 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Nguyen",
"Nghia Hieu",
""
],
[
"Vo",
"Duong T. D.",
""
],
[
"Ha",
"Minh-Quan",
""
]
] | TITLE: VieCap4H-VLSP 2021: ObjectAoA-Enhancing performance of Object Relation
Transformer with Attention on Attention for Vietnamese image captioning
ABSTRACT: Image captioning is currently a challenging task that requires the ability to
both understand visual information and use human language to describe this
visual information in the image. In this paper, we propose an efficient way to
improve the image understanding ability of transformer-based method by
extending Object Relation Transformer architecture with Attention on Attention
mechanism. Experiments on the VieCap4H dataset show that our proposed method
significantly outperforms its original structure on both the public test and
private test of the Image Captioning shared task held by VLSP.
| no_new_dataset | 0.947137 |
2211.11432 | Muhammad Jehanzeb Mirza | M. Jehanzeb Mirza, Inkyu Shin, Wei Lin, Andreas Schriebl, Kunyang Sun,
Jaesung Choe, Horst Possegger, Mateusz Kozinski, In So Kweon, Kun-Jin Yoon,
Horst Bischof | MATE: Masked Autoencoders are Online 3D Test-Time Learners | Code is available at this repository:
https://github.com/jmiemirza/MATE | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Our MATE is the first Test-Time-Training (TTT) method designed for 3D data,
which makes deep networks trained for point cloud classification robust to
distribution shifts occurring in test data. Like existing TTT methods from the
2D image domain, MATE also leverages test data for adaptation. Its test-time
objective is that of a Masked Autoencoder: a large portion of each test point
cloud is removed before it is fed to the network, tasked with reconstructing
the full point cloud. Once the network is updated, it is used to classify the
point cloud. We test MATE on several 3D object classification datasets and show
that it significantly improves robustness of deep networks to several types of
corruptions commonly occurring in 3D point clouds. We show that MATE is very
efficient in terms of the fraction of points it needs for the adaptation. It
can effectively adapt given as few as 5% of tokens of each test sample, making
it extremely lightweight. Our experiments show that MATE also achieves
competitive performance by adapting sparsely on the test data, which further
reduces its computational overhead, making it ideal for real-time applications.
| [
{
"version": "v1",
"created": "Mon, 21 Nov 2022 13:19:08 GMT"
},
{
"version": "v2",
"created": "Thu, 24 Nov 2022 10:52:59 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 09:44:58 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Mirza",
"M. Jehanzeb",
""
],
[
"Shin",
"Inkyu",
""
],
[
"Lin",
"Wei",
""
],
[
"Schriebl",
"Andreas",
""
],
[
"Sun",
"Kunyang",
""
],
[
"Choe",
"Jaesung",
""
],
[
"Possegger",
"Horst",
""
],
[
"Kozinski",
"Mateusz",
""
],
[
"Kweon",
"In So",
""
],
[
"Yoon",
"Kun-Jin",
""
],
[
"Bischof",
"Horst",
""
]
] | TITLE: MATE: Masked Autoencoders are Online 3D Test-Time Learners
ABSTRACT: Our MATE is the first Test-Time-Training (TTT) method designed for 3D data,
which makes deep networks trained for point cloud classification robust to
distribution shifts occurring in test data. Like existing TTT methods from the
2D image domain, MATE also leverages test data for adaptation. Its test-time
objective is that of a Masked Autoencoder: a large portion of each test point
cloud is removed before it is fed to the network, tasked with reconstructing
the full point cloud. Once the network is updated, it is used to classify the
point cloud. We test MATE on several 3D object classification datasets and show
that it significantly improves robustness of deep networks to several types of
corruptions commonly occurring in 3D point clouds. We show that MATE is very
efficient in terms of the fraction of points it needs for the adaptation. It
can effectively adapt given as few as 5% of tokens of each test sample, making
it extremely lightweight. Our experiments show that MATE also achieves
competitive performance by adapting sparsely on the test data, which further
reduces its computational overhead, making it ideal for real-time applications.
| no_new_dataset | 0.94743 |
2211.11674 | Dario Pavllo | Dario Pavllo, David Joseph Tan, Marie-Julie Rakotosaona, Federico
Tombari | Shape, Pose, and Appearance from a Single Image via Bootstrapped
Radiance Field Inversion | CVPR 2023. Code and models are available at
https://github.com/google-research/nerf-from-image | null | null | null | cs.CV cs.AI cs.GR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural Radiance Fields (NeRF) coupled with GANs represent a promising
direction in the area of 3D reconstruction from a single view, owing to their
ability to efficiently model arbitrary topologies. Recent work in this area,
however, has mostly focused on synthetic datasets where exact ground-truth
poses are known, and has overlooked pose estimation, which is important for
certain downstream applications such as augmented reality (AR) and robotics. We
introduce a principled end-to-end reconstruction framework for natural images,
where accurate ground-truth poses are not available. Our approach recovers an
SDF-parameterized 3D shape, pose, and appearance from a single image of an
object, without exploiting multiple views during training. More specifically,
we leverage an unconditional 3D-aware generator, to which we apply a hybrid
inversion scheme where a model produces a first guess of the solution which is
then refined via optimization. Our framework can de-render an image in as few
as 10 steps, enabling its use in practical scenarios. We demonstrate
state-of-the-art results on a variety of real and synthetic benchmarks.
| [
{
"version": "v1",
"created": "Mon, 21 Nov 2022 17:42:42 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 11:33:18 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Pavllo",
"Dario",
""
],
[
"Tan",
"David Joseph",
""
],
[
"Rakotosaona",
"Marie-Julie",
""
],
[
"Tombari",
"Federico",
""
]
] | TITLE: Shape, Pose, and Appearance from a Single Image via Bootstrapped
Radiance Field Inversion
ABSTRACT: Neural Radiance Fields (NeRF) coupled with GANs represent a promising
direction in the area of 3D reconstruction from a single view, owing to their
ability to efficiently model arbitrary topologies. Recent work in this area,
however, has mostly focused on synthetic datasets where exact ground-truth
poses are known, and has overlooked pose estimation, which is important for
certain downstream applications such as augmented reality (AR) and robotics. We
introduce a principled end-to-end reconstruction framework for natural images,
where accurate ground-truth poses are not available. Our approach recovers an
SDF-parameterized 3D shape, pose, and appearance from a single image of an
object, without exploiting multiple views during training. More specifically,
we leverage an unconditional 3D-aware generator, to which we apply a hybrid
inversion scheme where a model produces a first guess of the solution which is
then refined via optimization. Our framework can de-render an image in as few
as 10 steps, enabling its use in practical scenarios. We demonstrate
state-of-the-art results on a variety of real and synthetic benchmarks.
| no_new_dataset | 0.947332 |
2211.13221 | Yingqing He | Yingqing He, Tianyu Yang, Yong Zhang, Ying Shan, Qifeng Chen | Latent Video Diffusion Models for High-Fidelity Long Video Generation | Project Page: https://yingqinghe.github.io/LVDM/ Github:
https://github.com/YingqingHe/LVDM | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | AI-generated content has attracted lots of attention recently, but
photo-realistic video synthesis is still challenging. Although many attempts
using GANs and autoregressive models have been made in this area, the visual
quality and length of generated videos are far from satisfactory. Diffusion
models have shown remarkable results recently but require significant
computational resources. To address this, we introduce lightweight video
diffusion models by leveraging a low-dimensional 3D latent space, significantly
outperforming previous pixel-space video diffusion models under a limited
computational budget. In addition, we propose hierarchical diffusion in the
latent space such that longer videos with more than one thousand frames can be
produced. To further overcome the performance degradation issue for long video
generation, we propose conditional latent perturbation and unconditional
guidance that effectively mitigate the accumulated errors during the extension
of video length. Extensive experiments on small domain datasets of different
categories suggest that our framework generates more realistic and longer
videos than previous strong baselines. We additionally provide an extension to
large-scale text-to-video generation to demonstrate the superiority of our
work. Our code and models will be made publicly available.
| [
{
"version": "v1",
"created": "Wed, 23 Nov 2022 18:58:39 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 17:29:45 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"He",
"Yingqing",
""
],
[
"Yang",
"Tianyu",
""
],
[
"Zhang",
"Yong",
""
],
[
"Shan",
"Ying",
""
],
[
"Chen",
"Qifeng",
""
]
] | TITLE: Latent Video Diffusion Models for High-Fidelity Long Video Generation
ABSTRACT: AI-generated content has attracted lots of attention recently, but
photo-realistic video synthesis is still challenging. Although many attempts
using GANs and autoregressive models have been made in this area, the visual
quality and length of generated videos are far from satisfactory. Diffusion
models have shown remarkable results recently but require significant
computational resources. To address this, we introduce lightweight video
diffusion models by leveraging a low-dimensional 3D latent space, significantly
outperforming previous pixel-space video diffusion models under a limited
computational budget. In addition, we propose hierarchical diffusion in the
latent space such that longer videos with more than one thousand frames can be
produced. To further overcome the performance degradation issue for long video
generation, we propose conditional latent perturbation and unconditional
guidance that effectively mitigate the accumulated errors during the extension
of video length. Extensive experiments on small domain datasets of different
categories suggest that our framework generates more realistic and longer
videos than previous strong baselines. We additionally provide an extension to
large-scale text-to-video generation to demonstrate the superiority of our
work. Our code and models will be made publicly available.
| no_new_dataset | 0.951142 |
2211.14638 | Zuhui Wang | Zuhui Wang | Cross-domain Microscopy Cell Counting by Disentangled Transfer Learning | Accepted by ICLR 2023 Workshop on TML4H | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Microscopy images from different imaging conditions, organs, and tissues
often have numerous cells with various shapes on a range of backgrounds. As a
result, designing a deep learning model to count cells in a source domain
becomes precarious when transferring them to a new target domain. To address
this issue, manual annotation costs are typically the norm when training deep
learning-based cell counting models across different domains. In this paper, we
propose a cross-domain cell counting approach that requires only weak human
annotation efforts. Initially, we implement a cell counting network that
disentangles domain-specific knowledge from domain-agnostic knowledge in cell
images, where they pertain to the creation of domain style images and cell
density maps, respectively. We then devise an image synthesis technique capable
of generating massive synthetic images founded on a few target-domain images
that have been labeled. Finally, we use a public dataset consisting of
synthetic cells as the source domain, where no manual annotation cost is
present, to train our cell counting network; subsequently, we transfer only the
domain-agnostic knowledge to a new target domain of real cell images. By
progressively refining the trained model using synthesized target-domain images
and several real annotated ones, our proposed cross-domain cell counting method
achieves good performance compared to state-of-the-art techniques that rely on
fully annotated training images in the target domain. We evaluated the efficacy
of our cross-domain approach on two target domain datasets of actual microscopy
cells, demonstrating the feasibility of requiring annotations on only a few
images in a new domain.
| [
{
"version": "v1",
"created": "Sat, 26 Nov 2022 18:41:37 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 03:08:09 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Wang",
"Zuhui",
""
]
] | TITLE: Cross-domain Microscopy Cell Counting by Disentangled Transfer Learning
ABSTRACT: Microscopy images from different imaging conditions, organs, and tissues
often have numerous cells with various shapes on a range of backgrounds. As a
result, designing a deep learning model to count cells in a source domain
becomes precarious when transferring them to a new target domain. To address
this issue, manual annotation costs are typically the norm when training deep
learning-based cell counting models across different domains. In this paper, we
propose a cross-domain cell counting approach that requires only weak human
annotation efforts. Initially, we implement a cell counting network that
disentangles domain-specific knowledge from domain-agnostic knowledge in cell
images, where they pertain to the creation of domain style images and cell
density maps, respectively. We then devise an image synthesis technique capable
of generating massive synthetic images founded on a few target-domain images
that have been labeled. Finally, we use a public dataset consisting of
synthetic cells as the source domain, where no manual annotation cost is
present, to train our cell counting network; subsequently, we transfer only the
domain-agnostic knowledge to a new target domain of real cell images. By
progressively refining the trained model using synthesized target-domain images
and several real annotated ones, our proposed cross-domain cell counting method
achieves good performance compared to state-of-the-art techniques that rely on
fully annotated training images in the target domain. We evaluated the efficacy
of our cross-domain approach on two target domain datasets of actual microscopy
cells, demonstrating the feasibility of requiring annotations on only a few
images in a new domain.
| no_new_dataset | 0.953751 |
2212.02870 | Ayush Tripathi | Ayush Tripathi, Prathosh AP, Suriya Prakash Muthukrishnan, Lalan Kumar | TripCEAiR: A Multi-Loss minimization approach for surface EMG based
Airwriting Recognition | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Airwriting Recognition refers to the problem of identification of letters
written in space with movement of the finger. It can be seen as a special case
of dynamic gesture recognition wherein the set of gestures are letters in a
particular language. Surface Electromyography (sEMG) is a non-invasive approach
used to capture electrical signals generated as a result of contraction and
relaxation of the muscles. sEMG has been widely adopted for gesture recognition
applications. Unlike static gestures, dynamic gestures are user-friendly and
can be used as a method for input with applications in Human Computer
Interaction. There has been limited work in recognition of dynamic gestures
such as airwriting, using sEMG signals and forms the core of the current work.
In this work, a multi-loss minimization framework for sEMG based airwriting
recognition is proposed. The proposed framework aims at learning a feature
embedding vector that minimizes the triplet loss, while simultaneously learning
the parameters of a classifier head to recognize corresponding alphabets. The
proposed method is validated on a dataset recorded in the lab comprising of
sEMG signals from 50 participants writing English uppercase alphabets. The
effect of different variations of triplet loss, triplet mining strategies and
feature embedding dimension is also presented. The best-achieved accuracy was
81.26% and 65.62% in user-dependent and independent scenarios respectively by
using semihard positive and hard negative triplet mining. The code for our
implementation will be made available at https://github.com/ayushayt/TripCEAiR.
| [
{
"version": "v1",
"created": "Tue, 6 Dec 2022 10:20:19 GMT"
},
{
"version": "v2",
"created": "Mon, 2 Jan 2023 17:03:38 GMT"
},
{
"version": "v3",
"created": "Sun, 19 Mar 2023 16:39:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Tripathi",
"Ayush",
""
],
[
"AP",
"Prathosh",
""
],
[
"Muthukrishnan",
"Suriya Prakash",
""
],
[
"Kumar",
"Lalan",
""
]
] | TITLE: TripCEAiR: A Multi-Loss minimization approach for surface EMG based
Airwriting Recognition
ABSTRACT: Airwriting Recognition refers to the problem of identification of letters
written in space with movement of the finger. It can be seen as a special case
of dynamic gesture recognition wherein the set of gestures are letters in a
particular language. Surface Electromyography (sEMG) is a non-invasive approach
used to capture electrical signals generated as a result of contraction and
relaxation of the muscles. sEMG has been widely adopted for gesture recognition
applications. Unlike static gestures, dynamic gestures are user-friendly and
can be used as a method for input with applications in Human Computer
Interaction. There has been limited work in recognition of dynamic gestures
such as airwriting, using sEMG signals and forms the core of the current work.
In this work, a multi-loss minimization framework for sEMG based airwriting
recognition is proposed. The proposed framework aims at learning a feature
embedding vector that minimizes the triplet loss, while simultaneously learning
the parameters of a classifier head to recognize corresponding alphabets. The
proposed method is validated on a dataset recorded in the lab comprising of
sEMG signals from 50 participants writing English uppercase alphabets. The
effect of different variations of triplet loss, triplet mining strategies and
feature embedding dimension is also presented. The best-achieved accuracy was
81.26% and 65.62% in user-dependent and independent scenarios respectively by
using semihard positive and hard negative triplet mining. The code for our
implementation will be made available at https://github.com/ayushayt/TripCEAiR.
| no_new_dataset | 0.922343 |
2212.04689 | Haitao Lin | Haitao Lin, Lirong Wu, Yongjie Xu, Yufei Huang, Siyuan Li, Guojiang
Zhao, Stan Z. Li | Non-equispaced Fourier Neural Solvers for PDEs | 27 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Solving partial differential equations is difficult. Recently proposed neural
resolution-invariant models, despite their effectiveness and efficiency,
usually require equispaced spatial points of data. However, sampling in spatial
domain is sometimes inevitably non-equispaced in real-world systems, limiting
their applicability. In this paper, we propose a Non-equispaced Fourier PDE
Solver (\textsc{NFS}) with adaptive interpolation on resampled equispaced
points and a variant of Fourier Neural Operators as its components.
Experimental results on complex PDEs demonstrate its advantages in accuracy and
efficiency. Compared with the spatially-equispaced benchmark methods, it
achieves superior performance with $42.85\%$ improvements on MAE, and is able
to handle non-equispaced data with a tiny loss of accuracy. Besides, to our
best knowledge, \textsc{NFS} is the first ML-based method with mesh invariant
inference ability to successfully model turbulent flows in non-equispaced
scenarios, with a minor deviation of the error on unseen spatial points.
| [
{
"version": "v1",
"created": "Fri, 9 Dec 2022 06:36:54 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 08:52:33 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Lin",
"Haitao",
""
],
[
"Wu",
"Lirong",
""
],
[
"Xu",
"Yongjie",
""
],
[
"Huang",
"Yufei",
""
],
[
"Li",
"Siyuan",
""
],
[
"Zhao",
"Guojiang",
""
],
[
"Li",
"Stan Z.",
""
]
] | TITLE: Non-equispaced Fourier Neural Solvers for PDEs
ABSTRACT: Solving partial differential equations is difficult. Recently proposed neural
resolution-invariant models, despite their effectiveness and efficiency,
usually require equispaced spatial points of data. However, sampling in spatial
domain is sometimes inevitably non-equispaced in real-world systems, limiting
their applicability. In this paper, we propose a Non-equispaced Fourier PDE
Solver (\textsc{NFS}) with adaptive interpolation on resampled equispaced
points and a variant of Fourier Neural Operators as its components.
Experimental results on complex PDEs demonstrate its advantages in accuracy and
efficiency. Compared with the spatially-equispaced benchmark methods, it
achieves superior performance with $42.85\%$ improvements on MAE, and is able
to handle non-equispaced data with a tiny loss of accuracy. Besides, to our
best knowledge, \textsc{NFS} is the first ML-based method with mesh invariant
inference ability to successfully model turbulent flows in non-equispaced
scenarios, with a minor deviation of the error on unseen spatial points.
| no_new_dataset | 0.944022 |
2212.08830 | Tsung-Ming Tai | Tsung-Ming Tai, Giuseppe Fiameni, Cheng-Kuang Lee, Simon See, Oswald
Lanz | Inductive Attention for Video Action Anticipation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anticipating future actions based on spatiotemporal observations is essential
in video understanding and predictive computer vision. Moreover, a model
capable of anticipating the future has important applications, it can benefit
precautionary systems to react before an event occurs. However, unlike in the
action recognition task, future information is inaccessible at observation time
-- a model cannot directly map the video frames to the target action to solve
the anticipation task. Instead, the temporal inference is required to associate
the relevant evidence with possible future actions. Consequently, existing
solutions based on the action recognition models are only suboptimal. Recently,
researchers proposed extending the observation window to capture longer
pre-action profiles from past moments and leveraging attention to retrieve the
subtle evidence to improve the anticipation predictions. However, existing
attention designs typically use frame inputs as the query which is suboptimal,
as a video frame only weakly connects to the future action. To this end, we
propose an inductive attention model, dubbed IAM, which leverages the current
prediction priors as the query to infer future action and can efficiently
process the long video content. Furthermore, our method considers the
uncertainty of the future via the many-to-many association in the attention
design. As a result, IAM consistently outperforms the state-of-the-art
anticipation models on multiple large-scale egocentric video datasets while
using significantly fewer model parameters.
| [
{
"version": "v1",
"created": "Sat, 17 Dec 2022 09:51:17 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 04:48:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Tai",
"Tsung-Ming",
""
],
[
"Fiameni",
"Giuseppe",
""
],
[
"Lee",
"Cheng-Kuang",
""
],
[
"See",
"Simon",
""
],
[
"Lanz",
"Oswald",
""
]
] | TITLE: Inductive Attention for Video Action Anticipation
ABSTRACT: Anticipating future actions based on spatiotemporal observations is essential
in video understanding and predictive computer vision. Moreover, a model
capable of anticipating the future has important applications, it can benefit
precautionary systems to react before an event occurs. However, unlike in the
action recognition task, future information is inaccessible at observation time
-- a model cannot directly map the video frames to the target action to solve
the anticipation task. Instead, the temporal inference is required to associate
the relevant evidence with possible future actions. Consequently, existing
solutions based on the action recognition models are only suboptimal. Recently,
researchers proposed extending the observation window to capture longer
pre-action profiles from past moments and leveraging attention to retrieve the
subtle evidence to improve the anticipation predictions. However, existing
attention designs typically use frame inputs as the query which is suboptimal,
as a video frame only weakly connects to the future action. To this end, we
propose an inductive attention model, dubbed IAM, which leverages the current
prediction priors as the query to infer future action and can efficiently
process the long video content. Furthermore, our method considers the
uncertainty of the future via the many-to-many association in the attention
design. As a result, IAM consistently outperforms the state-of-the-art
anticipation models on multiple large-scale egocentric video datasets while
using significantly fewer model parameters.
| no_new_dataset | 0.946399 |
2212.09581 | Yuming Jiang | Yuming Jiang, Kelvin C.K. Chan, Xintao Wang, Chen Change Loy, Ziwei
Liu | Reference-based Image and Video Super-Resolution via C2-Matching | To appear in IEEE Transactions on Pattern Analysis and Machine
Intelligence (TPAMI). arXiv admin note: substantial text overlap with
arXiv:2106.01863 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reference-based Super-Resolution (Ref-SR) has recently emerged as a promising
paradigm to enhance a low-resolution (LR) input image or video by introducing
an additional high-resolution (HR) reference image. Existing Ref-SR methods
mostly rely on implicit correspondence matching to borrow HR textures from
reference images to compensate for the information loss in input images.
However, performing local transfer is difficult because of two gaps between
input and reference images: the transformation gap (e.g., scale and rotation)
and the resolution gap (e.g., HR and LR). To tackle these challenges, we
propose C2-Matching in this work, which performs explicit robust matching
crossing transformation and resolution. 1) To bridge the transformation gap, we
propose a contrastive correspondence network, which learns
transformation-robust correspondences using augmented views of the input image.
2) To address the resolution gap, we adopt teacher-student correlation
distillation, which distills knowledge from the easier HR-HR matching to guide
the more ambiguous LR-HR matching. 3) Finally, we design a dynamic aggregation
module to address the potential misalignment issue between input images and
reference images. In addition, to faithfully evaluate the performance of
Reference-based Image Super-Resolution under a realistic setting, we contribute
the Webly-Referenced SR (WR-SR) dataset, mimicking the practical usage
scenario. We also extend C2-Matching to Reference-based Video Super-Resolution
task, where an image taken in a similar scene serves as the HR reference image.
Extensive experiments demonstrate that our proposed C2-Matching significantly
outperforms state of the arts on the standard CUFED5 benchmark and also boosts
the performance of video SR by incorporating the C2-Matching component into
Video SR pipelines.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 16:15:02 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 13:54:29 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Jiang",
"Yuming",
""
],
[
"Chan",
"Kelvin C. K.",
""
],
[
"Wang",
"Xintao",
""
],
[
"Loy",
"Chen Change",
""
],
[
"Liu",
"Ziwei",
""
]
] | TITLE: Reference-based Image and Video Super-Resolution via C2-Matching
ABSTRACT: Reference-based Super-Resolution (Ref-SR) has recently emerged as a promising
paradigm to enhance a low-resolution (LR) input image or video by introducing
an additional high-resolution (HR) reference image. Existing Ref-SR methods
mostly rely on implicit correspondence matching to borrow HR textures from
reference images to compensate for the information loss in input images.
However, performing local transfer is difficult because of two gaps between
input and reference images: the transformation gap (e.g., scale and rotation)
and the resolution gap (e.g., HR and LR). To tackle these challenges, we
propose C2-Matching in this work, which performs explicit robust matching
crossing transformation and resolution. 1) To bridge the transformation gap, we
propose a contrastive correspondence network, which learns
transformation-robust correspondences using augmented views of the input image.
2) To address the resolution gap, we adopt teacher-student correlation
distillation, which distills knowledge from the easier HR-HR matching to guide
the more ambiguous LR-HR matching. 3) Finally, we design a dynamic aggregation
module to address the potential misalignment issue between input images and
reference images. In addition, to faithfully evaluate the performance of
Reference-based Image Super-Resolution under a realistic setting, we contribute
the Webly-Referenced SR (WR-SR) dataset, mimicking the practical usage
scenario. We also extend C2-Matching to Reference-based Video Super-Resolution
task, where an image taken in a similar scene serves as the HR reference image.
Extensive experiments demonstrate that our proposed C2-Matching significantly
outperforms state of the arts on the standard CUFED5 benchmark and also boosts
the performance of video SR by incorporating the C2-Matching component into
Video SR pipelines.
| no_new_dataset | 0.95018 |
2212.13407 | Yi Song | Yi Song, Chuanzong Zhang, Xinhua Lu, Fabio Saggese, Zhongyong Wang | Hybrid Message Passing Algorithm for Downlink FDD Massive MIMO-OFDM
Channel Estimation | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The design of message passing (MP) algorithms on factor graphs is an
effective manner to implement channel estimation (CE) in wireless communication
systems, which performance can be further improved by exploiting prior
probability models that accurately match the channel characteristics. In this
work, we study the CE problem in a downlink massive multiple-input
multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM)
system. As the prior probability, we propose the Markov chain two-state
Gaussian mixture with large variance differences (TSGM-LVD) model to exploit
the structured sparsity in the angle-frequency domain of the channel. Existing
single and combined MP rules cannot deal with the message computation of the
proposed probability model. To overcome this issue, we present a general method
to derive the hybrid message passing (HMP) rule, which allows the calculation
of messages described by mixed linear and non-linear functions. Accordingly, we
design the HMP-TSGM-LVD algorithm under the structured turbo framework (STF).
Simulation results demonstrate that the proposed algorithm converges faster and
obtains better and more stable performance than its counterparts. In
particular, the gain of the proposed approach is maximum (3 dB) in the high
signal-to-noise ratio regime, while benchmark approaches experience oscillating
behavior due to the improper prior model characterization.
| [
{
"version": "v1",
"created": "Tue, 27 Dec 2022 08:37:32 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 15:53:18 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Song",
"Yi",
""
],
[
"Zhang",
"Chuanzong",
""
],
[
"Lu",
"Xinhua",
""
],
[
"Saggese",
"Fabio",
""
],
[
"Wang",
"Zhongyong",
""
]
] | TITLE: Hybrid Message Passing Algorithm for Downlink FDD Massive MIMO-OFDM
Channel Estimation
ABSTRACT: The design of message passing (MP) algorithms on factor graphs is an
effective manner to implement channel estimation (CE) in wireless communication
systems, which performance can be further improved by exploiting prior
probability models that accurately match the channel characteristics. In this
work, we study the CE problem in a downlink massive multiple-input
multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM)
system. As the prior probability, we propose the Markov chain two-state
Gaussian mixture with large variance differences (TSGM-LVD) model to exploit
the structured sparsity in the angle-frequency domain of the channel. Existing
single and combined MP rules cannot deal with the message computation of the
proposed probability model. To overcome this issue, we present a general method
to derive the hybrid message passing (HMP) rule, which allows the calculation
of messages described by mixed linear and non-linear functions. Accordingly, we
design the HMP-TSGM-LVD algorithm under the structured turbo framework (STF).
Simulation results demonstrate that the proposed algorithm converges faster and
obtains better and more stable performance than its counterparts. In
particular, the gain of the proposed approach is maximum (3 dB) in the high
signal-to-noise ratio regime, while benchmark approaches experience oscillating
behavior due to the improper prior model characterization.
| no_new_dataset | 0.945349 |
2301.04631 | Nazmul Shahadat | Nazmul Shahadat, Anthony S. Maida | Deep Residual Axial Networks | null | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While convolutional neural networks (CNNs) demonstrate outstanding
performance on computer vision tasks, their computational costs remain high.
Several techniques are used to reduce these costs, like reducing channel count,
and using separable and depthwise separable convolutions. This paper reduces
computational costs by introducing a novel architecture, axial CNNs, which
replaces spatial 2D convolution operations with two consecutive depthwise
separable 1D operations. The axial CNNs are predicated on the assumption that
the dataset supports approximately separable convolution operations with little
or no loss of training accuracy. Deep axial separable CNNs still suffer from
gradient problems when training deep networks. We modify the construction of
axial separable CNNs with residual connections to improve the performance of
deep axial architectures and introduce our final novel architecture namely
residual axial networks (RANs). Extensive benchmark evaluation shows that RANs
achieve at least 1% higher performance with about 77%, 86%, 75%, and 34% fewer
parameters and about 75%, 80%, 67%, and 26% fewer flops than ResNets, wide
ResNets, MobileNets, and SqueezeNexts on CIFAR benchmarks, SVHN, and Tiny
ImageNet image classification datasets. Moreover, our proposed RANs improve
deep recursive residual networks performance with 94% fewer parameters on the
image super-resolution dataset.
| [
{
"version": "v1",
"created": "Wed, 11 Jan 2023 18:36:54 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 01:48:48 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Shahadat",
"Nazmul",
""
],
[
"Maida",
"Anthony S.",
""
]
] | TITLE: Deep Residual Axial Networks
ABSTRACT: While convolutional neural networks (CNNs) demonstrate outstanding
performance on computer vision tasks, their computational costs remain high.
Several techniques are used to reduce these costs, like reducing channel count,
and using separable and depthwise separable convolutions. This paper reduces
computational costs by introducing a novel architecture, axial CNNs, which
replaces spatial 2D convolution operations with two consecutive depthwise
separable 1D operations. The axial CNNs are predicated on the assumption that
the dataset supports approximately separable convolution operations with little
or no loss of training accuracy. Deep axial separable CNNs still suffer from
gradient problems when training deep networks. We modify the construction of
axial separable CNNs with residual connections to improve the performance of
deep axial architectures and introduce our final novel architecture namely
residual axial networks (RANs). Extensive benchmark evaluation shows that RANs
achieve at least 1% higher performance with about 77%, 86%, 75%, and 34% fewer
parameters and about 75%, 80%, 67%, and 26% fewer flops than ResNets, wide
ResNets, MobileNets, and SqueezeNexts on CIFAR benchmarks, SVHN, and Tiny
ImageNet image classification datasets. Moreover, our proposed RANs improve
deep recursive residual networks performance with 94% fewer parameters on the
image super-resolution dataset.
| no_new_dataset | 0.95418 |
2301.05871 | Zhang Songchun | Songchun Zhang and Chunhui Zhao | Dyna-DepthFormer: Multi-frame Transformer for Self-Supervised Depth
Estimation in Dynamic Scenes | ICRA 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-supervised methods have showed promising results on depth estimation
task. However, previous methods estimate the target depth map and camera
ego-motion simultaneously, underusing multi-frame correlation information and
ignoring the motion of dynamic objects. In this paper, we propose a novel
Dyna-Depthformer framework, which predicts scene depth and 3D motion field
jointly and aggregates multi-frame information with transformer. Our
contributions are two-fold. First, we leverage multi-view correlation through a
series of self- and cross-attention layers in order to obtain enhanced depth
feature representation. Specifically, we use the perspective transformation to
acquire the initial reference point, and use deformable attention to reduce the
computational cost. Second, we propose a warping-based Motion Network to
estimate the motion field of dynamic objects without using semantic prior. To
improve the motion field predictions, we propose an iterative optimization
strategy, together with a sparsity-regularized loss. The entire pipeline
achieves end-to-end self-supervised training by constructing a minimum
reprojection loss. Extensive experiments on the KITTI and Cityscapes benchmarks
demonstrate the effectiveness of our method and show that our method
outperforms state-of-the-art algorithms.
| [
{
"version": "v1",
"created": "Sat, 14 Jan 2023 09:43:23 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 09:12:21 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhang",
"Songchun",
""
],
[
"Zhao",
"Chunhui",
""
]
] | TITLE: Dyna-DepthFormer: Multi-frame Transformer for Self-Supervised Depth
Estimation in Dynamic Scenes
ABSTRACT: Self-supervised methods have showed promising results on depth estimation
task. However, previous methods estimate the target depth map and camera
ego-motion simultaneously, underusing multi-frame correlation information and
ignoring the motion of dynamic objects. In this paper, we propose a novel
Dyna-Depthformer framework, which predicts scene depth and 3D motion field
jointly and aggregates multi-frame information with transformer. Our
contributions are two-fold. First, we leverage multi-view correlation through a
series of self- and cross-attention layers in order to obtain enhanced depth
feature representation. Specifically, we use the perspective transformation to
acquire the initial reference point, and use deformable attention to reduce the
computational cost. Second, we propose a warping-based Motion Network to
estimate the motion field of dynamic objects without using semantic prior. To
improve the motion field predictions, we propose an iterative optimization
strategy, together with a sparsity-regularized loss. The entire pipeline
achieves end-to-end self-supervised training by constructing a minimum
reprojection loss. Extensive experiments on the KITTI and Cityscapes benchmarks
demonstrate the effectiveness of our method and show that our method
outperforms state-of-the-art algorithms.
| no_new_dataset | 0.946941 |
2301.06323 | Rui Sun | Rui Sun, Xiuyu Wu, Yunfang Wu | An Error-Guided Correction Model for Chinese Spelling Error Correction | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Although existing neural network approaches have achieved great success on
Chinese spelling correction, there is still room to improve. The model is
required to avoid over-correction and to distinguish a correct token from its
phonological and visually similar ones. In this paper, we propose an
error-guided correction model (EGCM) to improve Chinese spelling correction. By
borrowing the powerful ability of BERT, we propose a novel zero-shot error
detection method to do a preliminary detection, which guides our model to
attend more on the probably wrong tokens in encoding and to avoid modifying the
correct tokens in generating. Furthermore, we introduce a new loss function to
integrate the error confusion set, which enables our model to distinguish
easily misused tokens. Moreover, our model supports highly parallel decoding to
meet real application requirements. Experiments are conducted on widely used
benchmarks. Our model achieves superior performance against state-of-the-art
approaches by a remarkable margin, on both the correction quality and
computation speed.
| [
{
"version": "v1",
"created": "Mon, 16 Jan 2023 09:27:45 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 08:37:45 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Sun",
"Rui",
""
],
[
"Wu",
"Xiuyu",
""
],
[
"Wu",
"Yunfang",
""
]
] | TITLE: An Error-Guided Correction Model for Chinese Spelling Error Correction
ABSTRACT: Although existing neural network approaches have achieved great success on
Chinese spelling correction, there is still room to improve. The model is
required to avoid over-correction and to distinguish a correct token from its
phonological and visually similar ones. In this paper, we propose an
error-guided correction model (EGCM) to improve Chinese spelling correction. By
borrowing the powerful ability of BERT, we propose a novel zero-shot error
detection method to do a preliminary detection, which guides our model to
attend more on the probably wrong tokens in encoding and to avoid modifying the
correct tokens in generating. Furthermore, we introduce a new loss function to
integrate the error confusion set, which enables our model to distinguish
easily misused tokens. Moreover, our model supports highly parallel decoding to
meet real application requirements. Experiments are conducted on widely used
benchmarks. Our model achieves superior performance against state-of-the-art
approaches by a remarkable margin, on both the correction quality and
computation speed.
| no_new_dataset | 0.946843 |
2302.00706 | Aurore Loisy | Aurore Loisy, Robin A. Heinonen | Deep reinforcement learning for the olfactory search POMDP: a
quantitative benchmark | null | Eur. Phys. J. E 46, 17 (2023) | 10.1140/epje/s10189-023-00277-8 | null | cs.RO physics.bio-ph physics.flu-dyn | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The olfactory search POMDP (partially observable Markov decision process) is
a sequential decision-making problem designed to mimic the task faced by
insects searching for a source of odor in turbulence, and its solutions have
applications to sniffer robots. As exact solutions are out of reach, the
challenge consists in finding the best possible approximate solutions while
keeping the computational cost reasonable. We provide a quantitative
benchmarking of a solver based on deep reinforcement learning against
traditional POMDP approximate solvers. We show that deep reinforcement learning
is a competitive alternative to standard methods, in particular to generate
lightweight policies suitable for robots.
| [
{
"version": "v1",
"created": "Wed, 1 Feb 2023 19:01:47 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Mar 2023 11:55:24 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Loisy",
"Aurore",
""
],
[
"Heinonen",
"Robin A.",
""
]
] | TITLE: Deep reinforcement learning for the olfactory search POMDP: a
quantitative benchmark
ABSTRACT: The olfactory search POMDP (partially observable Markov decision process) is
a sequential decision-making problem designed to mimic the task faced by
insects searching for a source of odor in turbulence, and its solutions have
applications to sniffer robots. As exact solutions are out of reach, the
challenge consists in finding the best possible approximate solutions while
keeping the computational cost reasonable. We provide a quantitative
benchmarking of a solver based on deep reinforcement learning against
traditional POMDP approximate solvers. We show that deep reinforcement learning
is a competitive alternative to standard methods, in particular to generate
lightweight policies suitable for robots.
| no_new_dataset | 0.946101 |
2302.04545 | Bosong Huang | Bosong Huang, Weihao Yu, Ruzhong Xie, Jing Xiao, Jin Huang | Lorentz Equivariant Model for Knowledge-Enhanced Hyperbolic
Collaborative Filtering | 11 pages, 6 figures | null | null | null | cs.IR cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Introducing prior auxiliary information from the knowledge graph (KG) to
assist the user-item graph can improve the comprehensive performance of the
recommender system. Many recent studies show that the ensemble properties of
hyperbolic spaces fit the scale-free and hierarchical characteristics exhibited
in the above two types of graphs well. However, existing hyperbolic methods
ignore the consideration of equivariance, thus they cannot generalize symmetric
features under given transformations, which seriously limits the capability of
the model. Moreover, they cannot balance preserving the heterogeneity and
mining the high-order entity information to users across two graphs. To fill
these gaps, we propose a rigorously Lorentz group equivariant
knowledge-enhanced collaborative filtering model (LECF). Innovatively, we
jointly update the attribute embeddings (containing the high-order entity
signals from the KG) and hyperbolic embeddings (the distance between hyperbolic
embeddings reveals the recommendation tendency) by the LECF layer with Lorentz
Equivariant Transformation. Moreover, we propose Hyperbolic Sparse Attention
Mechanism to sample the most informative neighbor nodes. Lorentz equivariance
is strictly maintained throughout the entire model, and enforcing equivariance
is proven necessary experimentally. Extensive experiments on three real-world
benchmarks demonstrate that LECF remarkably outperforms state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Thu, 9 Feb 2023 10:20:23 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 02:10:10 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Huang",
"Bosong",
""
],
[
"Yu",
"Weihao",
""
],
[
"Xie",
"Ruzhong",
""
],
[
"Xiao",
"Jing",
""
],
[
"Huang",
"Jin",
""
]
] | TITLE: Lorentz Equivariant Model for Knowledge-Enhanced Hyperbolic
Collaborative Filtering
ABSTRACT: Introducing prior auxiliary information from the knowledge graph (KG) to
assist the user-item graph can improve the comprehensive performance of the
recommender system. Many recent studies show that the ensemble properties of
hyperbolic spaces fit the scale-free and hierarchical characteristics exhibited
in the above two types of graphs well. However, existing hyperbolic methods
ignore the consideration of equivariance, thus they cannot generalize symmetric
features under given transformations, which seriously limits the capability of
the model. Moreover, they cannot balance preserving the heterogeneity and
mining the high-order entity information to users across two graphs. To fill
these gaps, we propose a rigorously Lorentz group equivariant
knowledge-enhanced collaborative filtering model (LECF). Innovatively, we
jointly update the attribute embeddings (containing the high-order entity
signals from the KG) and hyperbolic embeddings (the distance between hyperbolic
embeddings reveals the recommendation tendency) by the LECF layer with Lorentz
Equivariant Transformation. Moreover, we propose Hyperbolic Sparse Attention
Mechanism to sample the most informative neighbor nodes. Lorentz equivariance
is strictly maintained throughout the entire model, and enforcing equivariance
is proven necessary experimentally. Extensive experiments on three real-world
benchmarks demonstrate that LECF remarkably outperforms state-of-the-art
methods.
| no_new_dataset | 0.94868 |
2302.05738 | Junhong Shen | Junhong Shen, Liam Li, Lucio M. Dery, Corey Staten, Mikhail Khodak,
Graham Neubig, Ameet Talwalkar | Cross-Modal Fine-Tuning: Align then Refine | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Fine-tuning large-scale pretrained models has led to tremendous progress in
well-studied modalities such as vision and NLP. However, similar gains have not
been observed in many other modalities due to a lack of relevant pretrained
models. In this work, we propose ORCA, a general cross-modal fine-tuning
framework that extends the applicability of a single large-scale pretrained
model to diverse modalities. ORCA adapts to a target task via an
align-then-refine workflow: given the target input, ORCA first learns an
embedding network that aligns the embedded feature distribution with the
pretraining modality. The pretrained model is then fine-tuned on the embedded
data to exploit the knowledge shared across modalities. Through extensive
experiments, we show that ORCA obtains state-of-the-art results on 3 benchmarks
containing over 60 datasets from 12 modalities, outperforming a wide range of
hand-designed, AutoML, general-purpose, and task-specific methods. We highlight
the importance of data alignment via a series of ablation studies and
demonstrate ORCA's utility in data-limited regimes.
| [
{
"version": "v1",
"created": "Sat, 11 Feb 2023 16:32:28 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 17:05:04 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Shen",
"Junhong",
""
],
[
"Li",
"Liam",
""
],
[
"Dery",
"Lucio M.",
""
],
[
"Staten",
"Corey",
""
],
[
"Khodak",
"Mikhail",
""
],
[
"Neubig",
"Graham",
""
],
[
"Talwalkar",
"Ameet",
""
]
] | TITLE: Cross-Modal Fine-Tuning: Align then Refine
ABSTRACT: Fine-tuning large-scale pretrained models has led to tremendous progress in
well-studied modalities such as vision and NLP. However, similar gains have not
been observed in many other modalities due to a lack of relevant pretrained
models. In this work, we propose ORCA, a general cross-modal fine-tuning
framework that extends the applicability of a single large-scale pretrained
model to diverse modalities. ORCA adapts to a target task via an
align-then-refine workflow: given the target input, ORCA first learns an
embedding network that aligns the embedded feature distribution with the
pretraining modality. The pretrained model is then fine-tuned on the embedded
data to exploit the knowledge shared across modalities. Through extensive
experiments, we show that ORCA obtains state-of-the-art results on 3 benchmarks
containing over 60 datasets from 12 modalities, outperforming a wide range of
hand-designed, AutoML, general-purpose, and task-specific methods. We highlight
the importance of data alignment via a series of ablation studies and
demonstrate ORCA's utility in data-limited regimes.
| no_new_dataset | 0.944995 |
2302.06337 | Zhijun Chen | Zhijun Chen, Hailong Sun, Haoqian He, Pengpeng Chen | Learning from Noisy Crowd Labels with Logics | 12 pages, 7 figures, accepted by ICDE-2023 | null | null | null | cs.LG cs.AI cs.CL cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper explores the integration of symbolic logic knowledge into deep
neural networks for learning from noisy crowd labels. We introduce Logic-guided
Learning from Noisy Crowd Labels (Logic-LNCL), an EM-alike iterative logic
knowledge distillation framework that learns from both noisy labeled data and
logic rules of interest. Unlike traditional EM methods, our framework contains
a ``pseudo-E-step'' that distills from the logic rules a new type of learning
target, which is then used in the ``pseudo-M-step'' for training the
classifier. Extensive evaluations on two real-world datasets for text sentiment
classification and named entity recognition demonstrate that the proposed
framework improves the state-of-the-art and provides a new solution to learning
from noisy crowd labels.
| [
{
"version": "v1",
"created": "Mon, 13 Feb 2023 13:14:23 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Feb 2023 14:49:16 GMT"
},
{
"version": "v3",
"created": "Sun, 19 Mar 2023 13:01:47 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Chen",
"Zhijun",
""
],
[
"Sun",
"Hailong",
""
],
[
"He",
"Haoqian",
""
],
[
"Chen",
"Pengpeng",
""
]
] | TITLE: Learning from Noisy Crowd Labels with Logics
ABSTRACT: This paper explores the integration of symbolic logic knowledge into deep
neural networks for learning from noisy crowd labels. We introduce Logic-guided
Learning from Noisy Crowd Labels (Logic-LNCL), an EM-alike iterative logic
knowledge distillation framework that learns from both noisy labeled data and
logic rules of interest. Unlike traditional EM methods, our framework contains
a ``pseudo-E-step'' that distills from the logic rules a new type of learning
target, which is then used in the ``pseudo-M-step'' for training the
classifier. Extensive evaluations on two real-world datasets for text sentiment
classification and named entity recognition demonstrate that the proposed
framework improves the state-of-the-art and provides a new solution to learning
from noisy crowd labels.
| no_new_dataset | 0.953579 |
2302.08063 | Raghav Goyal | Raghav Goyal, Effrosyni Mavroudi, Xitong Yang, Sainbayar Sukhbaatar,
Leonid Sigal, Matt Feiszli, Lorenzo Torresani, Du Tran | MINOTAUR: Multi-task Video Grounding From Multimodal Queries | 22 pages, 8 figures and 13 tables | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Video understanding tasks take many forms, from action detection to visual
query localization and spatio-temporal grounding of sentences. These tasks
differ in the type of inputs (only video, or video-query pair where query is an
image region or sentence) and outputs (temporal segments or spatio-temporal
tubes). However, at their core they require the same fundamental understanding
of the video, i.e., the actors and objects in it, their actions and
interactions. So far these tasks have been tackled in isolation with
individual, highly specialized architectures, which do not exploit the
interplay between tasks. In contrast, in this paper, we present a single,
unified model for tackling query-based video understanding in long-form videos.
In particular, our model can address all three tasks of the Ego4D Episodic
Memory benchmark which entail queries of three different forms: given an
egocentric video and a visual, textual or activity query, the goal is to
determine when and where the answer can be seen within the video. Our model
design is inspired by recent query-based approaches to spatio-temporal
grounding, and contains modality-specific query encoders and task-specific
sliding window inference that allow multi-task training with diverse input
modalities and different structured outputs. We exhaustively analyze
relationships among the tasks and illustrate that cross-task learning leads to
improved performance on each individual task, as well as the ability to
generalize to unseen tasks, such as zero-shot spatial localization of language
queries.
| [
{
"version": "v1",
"created": "Thu, 16 Feb 2023 04:00:03 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Mar 2023 20:46:33 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Goyal",
"Raghav",
""
],
[
"Mavroudi",
"Effrosyni",
""
],
[
"Yang",
"Xitong",
""
],
[
"Sukhbaatar",
"Sainbayar",
""
],
[
"Sigal",
"Leonid",
""
],
[
"Feiszli",
"Matt",
""
],
[
"Torresani",
"Lorenzo",
""
],
[
"Tran",
"Du",
""
]
] | TITLE: MINOTAUR: Multi-task Video Grounding From Multimodal Queries
ABSTRACT: Video understanding tasks take many forms, from action detection to visual
query localization and spatio-temporal grounding of sentences. These tasks
differ in the type of inputs (only video, or video-query pair where query is an
image region or sentence) and outputs (temporal segments or spatio-temporal
tubes). However, at their core they require the same fundamental understanding
of the video, i.e., the actors and objects in it, their actions and
interactions. So far these tasks have been tackled in isolation with
individual, highly specialized architectures, which do not exploit the
interplay between tasks. In contrast, in this paper, we present a single,
unified model for tackling query-based video understanding in long-form videos.
In particular, our model can address all three tasks of the Ego4D Episodic
Memory benchmark which entail queries of three different forms: given an
egocentric video and a visual, textual or activity query, the goal is to
determine when and where the answer can be seen within the video. Our model
design is inspired by recent query-based approaches to spatio-temporal
grounding, and contains modality-specific query encoders and task-specific
sliding window inference that allow multi-task training with diverse input
modalities and different structured outputs. We exhaustively analyze
relationships among the tasks and illustrate that cross-task learning leads to
improved performance on each individual task, as well as the ability to
generalize to unseen tasks, such as zero-shot spatial localization of language
queries.
| no_new_dataset | 0.943034 |
2302.09315 | Rong Du | Rong Du, Qingqing Ye, Yue Fu, Haibo Hu, Jin Li, Chengfang Fang, Jie
Shi | Differential Aggregation against General Colluding Attackers | This paper has been accepted by ICDE2023 | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Local Differential Privacy (LDP) is now widely adopted in large-scale systems
to collect and analyze sensitive data while preserving users' privacy. However,
almost all LDP protocols rely on a semi-trust model where users are
curious-but-honest, which rarely holds in real-world scenarios. Recent works
show poor estimation accuracy of many LDP protocols under malicious threat
models. Although a few works have proposed some countermeasures to address
these attacks, they all require prior knowledge of either the attacking pattern
or the poison value distribution, which is impractical as they can be easily
evaded by the attackers.
In this paper, we adopt a general opportunistic-and-colluding threat model
and propose a multi-group Differential Aggregation Protocol (DAP) to improve
the accuracy of mean estimation under LDP. Different from all existing works
that detect poison values on individual basis, DAP mitigates the overall impact
of poison values on the estimated mean. It relies on a new probing mechanism
EMF (i.e., Expectation-Maximization Filter) to estimate features of the
attackers. In addition to EMF, DAP also consists of two EMF post-processing
procedures (EMF* and CEMF*), and a group-wise mean aggregation scheme to
optimize the final estimated mean to achieve the smallest variance. Extensive
experimental results on both synthetic and real-world datasets demonstrate the
superior performance of DAP over state-of-the-art solutions.
| [
{
"version": "v1",
"created": "Sat, 18 Feb 2023 12:21:45 GMT"
},
{
"version": "v2",
"created": "Fri, 24 Feb 2023 10:27:18 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 09:24:35 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Du",
"Rong",
""
],
[
"Ye",
"Qingqing",
""
],
[
"Fu",
"Yue",
""
],
[
"Hu",
"Haibo",
""
],
[
"Li",
"Jin",
""
],
[
"Fang",
"Chengfang",
""
],
[
"Shi",
"Jie",
""
]
] | TITLE: Differential Aggregation against General Colluding Attackers
ABSTRACT: Local Differential Privacy (LDP) is now widely adopted in large-scale systems
to collect and analyze sensitive data while preserving users' privacy. However,
almost all LDP protocols rely on a semi-trust model where users are
curious-but-honest, which rarely holds in real-world scenarios. Recent works
show poor estimation accuracy of many LDP protocols under malicious threat
models. Although a few works have proposed some countermeasures to address
these attacks, they all require prior knowledge of either the attacking pattern
or the poison value distribution, which is impractical as they can be easily
evaded by the attackers.
In this paper, we adopt a general opportunistic-and-colluding threat model
and propose a multi-group Differential Aggregation Protocol (DAP) to improve
the accuracy of mean estimation under LDP. Different from all existing works
that detect poison values on individual basis, DAP mitigates the overall impact
of poison values on the estimated mean. It relies on a new probing mechanism
EMF (i.e., Expectation-Maximization Filter) to estimate features of the
attackers. In addition to EMF, DAP also consists of two EMF post-processing
procedures (EMF* and CEMF*), and a group-wise mean aggregation scheme to
optimize the final estimated mean to achieve the smallest variance. Extensive
experimental results on both synthetic and real-world datasets demonstrate the
superior performance of DAP over state-of-the-art solutions.
| no_new_dataset | 0.942029 |
2302.09462 | Omid Nejati Manzari | Omid Nejati Manzari, Hamid Ahmadabadi, Hossein Kashiani, Shahriar B.
Shokouhi, Ahmad Ayatollahi | MedViT: A Robust Vision Transformer for Generalized Medical Image
Classification | null | Computers in Biology and Medicine 2023 | 10.1016/j.compbiomed.2023.106791 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Convolutional Neural Networks (CNNs) have advanced existing medical systems
for automatic disease diagnosis. However, there are still concerns about the
reliability of deep medical diagnosis systems against the potential threats of
adversarial attacks since inaccurate diagnosis could lead to disastrous
consequences in the safety realm. In this study, we propose a highly robust yet
efficient CNN-Transformer hybrid model which is equipped with the locality of
CNNs as well as the global connectivity of vision Transformers. To mitigate the
high quadratic complexity of the self-attention mechanism while jointly
attending to information in various representation subspaces, we construct our
attention mechanism by means of an efficient convolution operation. Moreover,
to alleviate the fragility of our Transformer model against adversarial
attacks, we attempt to learn smoother decision boundaries. To this end, we
augment the shape information of an image in the high-level feature space by
permuting the feature mean and variance within mini-batches. With less
computational complexity, our proposed hybrid model demonstrates its high
robustness and generalization ability compared to the state-of-the-art studies
on a large-scale collection of standardized MedMNIST-2D datasets.
| [
{
"version": "v1",
"created": "Sun, 19 Feb 2023 02:55:45 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Manzari",
"Omid Nejati",
""
],
[
"Ahmadabadi",
"Hamid",
""
],
[
"Kashiani",
"Hossein",
""
],
[
"Shokouhi",
"Shahriar B.",
""
],
[
"Ayatollahi",
"Ahmad",
""
]
] | TITLE: MedViT: A Robust Vision Transformer for Generalized Medical Image
Classification
ABSTRACT: Convolutional Neural Networks (CNNs) have advanced existing medical systems
for automatic disease diagnosis. However, there are still concerns about the
reliability of deep medical diagnosis systems against the potential threats of
adversarial attacks since inaccurate diagnosis could lead to disastrous
consequences in the safety realm. In this study, we propose a highly robust yet
efficient CNN-Transformer hybrid model which is equipped with the locality of
CNNs as well as the global connectivity of vision Transformers. To mitigate the
high quadratic complexity of the self-attention mechanism while jointly
attending to information in various representation subspaces, we construct our
attention mechanism by means of an efficient convolution operation. Moreover,
to alleviate the fragility of our Transformer model against adversarial
attacks, we attempt to learn smoother decision boundaries. To this end, we
augment the shape information of an image in the high-level feature space by
permuting the feature mean and variance within mini-batches. With less
computational complexity, our proposed hybrid model demonstrates its high
robustness and generalization ability compared to the state-of-the-art studies
on a large-scale collection of standardized MedMNIST-2D datasets.
| no_new_dataset | 0.9434 |
2302.10121 | Prajwal Singh | Prajwal Singh, Pankaj Pandey, Krishna Miyapuram, Shanmuganathan Raman | EEG2IMAGE: Image Reconstruction from EEG Brain Signals | Accepted in ICASSP 2023 | null | null | null | cs.HC q-bio.NC | http://creativecommons.org/licenses/by/4.0/ | Reconstructing images using brain signals of imagined visuals may provide an
augmented vision to the disabled, leading to the advancement of Brain-Computer
Interface (BCI) technology. The recent progress in deep learning has boosted
the study area of synthesizing images from brain signals using Generative
Adversarial Networks (GAN). In this work, we have proposed a framework for
synthesizing the images from the brain activity recorded by an
electroencephalogram (EEG) using small-size EEG datasets. This brain activity
is recorded from the subject's head scalp using EEG when they ask to visualize
certain classes of Objects and English characters. We use a contrastive
learning method in the proposed framework to extract features from EEG signals
and synthesize the images from extracted features using conditional GAN. We
modify the loss function to train the GAN, which enables it to synthesize
128x128 images using a small number of images. Further, we conduct ablation
studies and experiments to show the effectiveness of our proposed framework
over other state-of-the-art methods using the small EEG dataset.
| [
{
"version": "v1",
"created": "Mon, 20 Feb 2023 17:37:22 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 04:31:02 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Singh",
"Prajwal",
""
],
[
"Pandey",
"Pankaj",
""
],
[
"Miyapuram",
"Krishna",
""
],
[
"Raman",
"Shanmuganathan",
""
]
] | TITLE: EEG2IMAGE: Image Reconstruction from EEG Brain Signals
ABSTRACT: Reconstructing images using brain signals of imagined visuals may provide an
augmented vision to the disabled, leading to the advancement of Brain-Computer
Interface (BCI) technology. The recent progress in deep learning has boosted
the study area of synthesizing images from brain signals using Generative
Adversarial Networks (GAN). In this work, we have proposed a framework for
synthesizing the images from the brain activity recorded by an
electroencephalogram (EEG) using small-size EEG datasets. This brain activity
is recorded from the subject's head scalp using EEG when they ask to visualize
certain classes of Objects and English characters. We use a contrastive
learning method in the proposed framework to extract features from EEG signals
and synthesize the images from extracted features using conditional GAN. We
modify the loss function to train the GAN, which enables it to synthesize
128x128 images using a small number of images. Further, we conduct ablation
studies and experiments to show the effectiveness of our proposed framework
over other state-of-the-art methods using the small EEG dataset.
| no_new_dataset | 0.944842 |
2302.10571 | Carlos Hern\'andez-P\'erez | Cristian Pach\'on-Garc\'ia, Carlos Hern\'andez-P\'erez, Pedro
Delicado, Ver\'onica Vilaplana | SurvLIMEpy: A Python package implementing SurvLIME | null | null | null | null | stat.ML cs.AI cs.LG cs.MS | http://creativecommons.org/licenses/by/4.0/ | In this paper we present SurvLIMEpy, an open-source Python package that
implements the SurvLIME algorithm. This method allows to compute local feature
importance for machine learning algorithms designed for modelling Survival
Analysis data. Our implementation takes advantage of the parallelisation
paradigm as all computations are performed in a matrix-wise fashion which
speeds up execution time. Additionally, SurvLIMEpy assists the user with
visualization tools to better understand the result of the algorithm. The
package supports a wide variety of survival models, from the Cox Proportional
Hazards Model to deep learning models such as DeepHit or DeepSurv. Two types of
experiments are presented in this paper. First, by means of simulated data, we
study the ability of the algorithm to capture the importance of the features.
Second, we use three open source survival datasets together with a set of
survival algorithms in order to demonstrate how SurvLIMEpy behaves when applied
to different models.
| [
{
"version": "v1",
"created": "Tue, 21 Feb 2023 09:54:32 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Mar 2023 21:41:38 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Pachón-García",
"Cristian",
""
],
[
"Hernández-Pérez",
"Carlos",
""
],
[
"Delicado",
"Pedro",
""
],
[
"Vilaplana",
"Verónica",
""
]
] | TITLE: SurvLIMEpy: A Python package implementing SurvLIME
ABSTRACT: In this paper we present SurvLIMEpy, an open-source Python package that
implements the SurvLIME algorithm. This method allows to compute local feature
importance for machine learning algorithms designed for modelling Survival
Analysis data. Our implementation takes advantage of the parallelisation
paradigm as all computations are performed in a matrix-wise fashion which
speeds up execution time. Additionally, SurvLIMEpy assists the user with
visualization tools to better understand the result of the algorithm. The
package supports a wide variety of survival models, from the Cox Proportional
Hazards Model to deep learning models such as DeepHit or DeepSurv. Two types of
experiments are presented in this paper. First, by means of simulated data, we
study the ability of the algorithm to capture the importance of the features.
Second, we use three open source survival datasets together with a set of
survival algorithms in order to demonstrate how SurvLIMEpy behaves when applied
to different models.
| no_new_dataset | 0.939415 |
2302.13562 | Yuhao Zhou | Yuhao Zhou, Mingjia Shi, Yuanxi Li, Qing Ye, Yanan Sun, Jiancheng Lv | Communication-efficient Federated Learning with Single-Step Synthetic
Features Compressor for Faster Convergence | null | null | null | null | cs.LG cs.AI cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reducing communication overhead in federated learning (FL) is challenging but
crucial for large-scale distributed privacy-preserving machine learning. While
methods utilizing sparsification or others can largely lower the communication
overhead, the convergence rate is also greatly compromised. In this paper, we
propose a novel method, named single-step synthetic features compressor (3SFC),
to achieve communication-efficient FL by directly constructing a tiny synthetic
dataset based on raw gradients. Thus, 3SFC can achieve an extremely low
compression rate when the constructed dataset contains only one data sample.
Moreover, 3SFC's compressing phase utilizes a similarity-based objective
function so that it can be optimized with just one step, thereby considerably
improving its performance and robustness. In addition, to minimize the
compressing error, error feedback (EF) is also incorporated into 3SFC.
Experiments on multiple datasets and models suggest that 3SFC owns
significantly better convergence rates compared to competing methods with lower
compression rates (up to 0.02%). Furthermore, ablation studies and
visualizations show that 3SFC can carry more information than competing methods
for every communication round, further validating its effectiveness.
| [
{
"version": "v1",
"created": "Mon, 27 Feb 2023 07:47:23 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 03:45:40 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Zhou",
"Yuhao",
""
],
[
"Shi",
"Mingjia",
""
],
[
"Li",
"Yuanxi",
""
],
[
"Ye",
"Qing",
""
],
[
"Sun",
"Yanan",
""
],
[
"Lv",
"Jiancheng",
""
]
] | TITLE: Communication-efficient Federated Learning with Single-Step Synthetic
Features Compressor for Faster Convergence
ABSTRACT: Reducing communication overhead in federated learning (FL) is challenging but
crucial for large-scale distributed privacy-preserving machine learning. While
methods utilizing sparsification or others can largely lower the communication
overhead, the convergence rate is also greatly compromised. In this paper, we
propose a novel method, named single-step synthetic features compressor (3SFC),
to achieve communication-efficient FL by directly constructing a tiny synthetic
dataset based on raw gradients. Thus, 3SFC can achieve an extremely low
compression rate when the constructed dataset contains only one data sample.
Moreover, 3SFC's compressing phase utilizes a similarity-based objective
function so that it can be optimized with just one step, thereby considerably
improving its performance and robustness. In addition, to minimize the
compressing error, error feedback (EF) is also incorporated into 3SFC.
Experiments on multiple datasets and models suggest that 3SFC owns
significantly better convergence rates compared to competing methods with lower
compression rates (up to 0.02%). Furthermore, ablation studies and
visualizations show that 3SFC can carry more information than competing methods
for every communication round, further validating its effectiveness.
| no_new_dataset | 0.93646 |
2302.13770 | Xiao Pengxiang | Pengxiang Xiao, Shuai He, Limin Liu, Anlong Ming | Mask Reference Image Quality Assessment | 10 pages, 6 figures | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding semantic information is an essential step in knowing what is
being learned in both full-reference (FR) and no-reference (NR) image quality
assessment (IQA) methods. However, especially for many severely distorted
images, even if there is an undistorted image as a reference (FR-IQA), it is
difficult to perceive the lost semantic and texture information of distorted
images directly. In this paper, we propose a Mask Reference IQA (MR-IQA) method
that masks specific patches of a distorted image and supplements missing
patches with the reference image patches. In this way, our model only needs to
input the reconstructed image for quality assessment. First, we design a mask
generator to select the best candidate patches from reference images and
supplement the lost semantic information in distorted images, thus providing
more reference for quality assessment; in addition, the different masked
patches imply different data augmentations, which favors model training and
reduces overfitting. Second, we provide a Mask Reference Network (MRNet): the
dedicated modules can prevent disturbances due to masked patches and help
eliminate the patch discontinuity in the reconstructed image. Our method
achieves state-of-the-art performances on the benchmark KADID-10k, LIVE and
CSIQ datasets and has better generalization performance across datasets. The
code and results are available in the supplementary material.
| [
{
"version": "v1",
"created": "Mon, 27 Feb 2023 13:52:38 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 14:14:29 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Xiao",
"Pengxiang",
""
],
[
"He",
"Shuai",
""
],
[
"Liu",
"Limin",
""
],
[
"Ming",
"Anlong",
""
]
] | TITLE: Mask Reference Image Quality Assessment
ABSTRACT: Understanding semantic information is an essential step in knowing what is
being learned in both full-reference (FR) and no-reference (NR) image quality
assessment (IQA) methods. However, especially for many severely distorted
images, even if there is an undistorted image as a reference (FR-IQA), it is
difficult to perceive the lost semantic and texture information of distorted
images directly. In this paper, we propose a Mask Reference IQA (MR-IQA) method
that masks specific patches of a distorted image and supplements missing
patches with the reference image patches. In this way, our model only needs to
input the reconstructed image for quality assessment. First, we design a mask
generator to select the best candidate patches from reference images and
supplement the lost semantic information in distorted images, thus providing
more reference for quality assessment; in addition, the different masked
patches imply different data augmentations, which favors model training and
reduces overfitting. Second, we provide a Mask Reference Network (MRNet): the
dedicated modules can prevent disturbances due to masked patches and help
eliminate the patch discontinuity in the reconstructed image. Our method
achieves state-of-the-art performances on the benchmark KADID-10k, LIVE and
CSIQ datasets and has better generalization performance across datasets. The
code and results are available in the supplementary material.
| no_new_dataset | 0.946597 |
2303.00465 | Khabibulla Madatov | Khabibulla Madatov and Sanatbek Matlatipov and Mersaid Aripov | Uzbek text's correspondence with the educational potential of pupils: a
case study of the School corpus | Preprint of the paper accepted to The 10th Language & Technology
Conference: Human Language Technologies as a Challenge for Computer Science
and Linguistics. April 21-23, 2023, Poznan, Poland | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | One of the major challenges of an educational system is choosing appropriate
content considering pupils' age and intellectual potential. In this article the
experiment of primary school grades (from 1st to 4th grades) is considered for
automatically determining the correspondence of an educational materials
recommended for pupils by using the School corpus where it includes the dataset
of 25 school textbooks confirmed by the Ministry of preschool and school
education of the Republic of Uzbekistan. In this case, TF-IDF scores of the
texts are determined, they are converted into a vector representation, and the
given educational materials are compared with the corresponding class of the
School corpus using the cosine similarity algorithm. Based on the results of
the calculation, it is determined whether the given educational material is
appropriate or not appropriate for the pupils' educational potential.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 12:46:00 GMT"
},
{
"version": "v2",
"created": "Sat, 18 Mar 2023 06:54:57 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Madatov",
"Khabibulla",
""
],
[
"Matlatipov",
"Sanatbek",
""
],
[
"Aripov",
"Mersaid",
""
]
] | TITLE: Uzbek text's correspondence with the educational potential of pupils: a
case study of the School corpus
ABSTRACT: One of the major challenges of an educational system is choosing appropriate
content considering pupils' age and intellectual potential. In this article the
experiment of primary school grades (from 1st to 4th grades) is considered for
automatically determining the correspondence of an educational materials
recommended for pupils by using the School corpus where it includes the dataset
of 25 school textbooks confirmed by the Ministry of preschool and school
education of the Republic of Uzbekistan. In this case, TF-IDF scores of the
texts are determined, they are converted into a vector representation, and the
given educational materials are compared with the corresponding class of the
School corpus using the cosine similarity algorithm. Based on the results of
the calculation, it is determined whether the given educational material is
appropriate or not appropriate for the pupils' educational potential.
| no_new_dataset | 0.939692 |
2303.00554 | Sarthak Chakraborty | Sarthak Chakraborty, Shaddy Garg, Shubham Agarwal, Ayush Chauhan and
Shiv Kumar Saini | CausIL: Causal Graph for Instance Level Microservice Data | Accepted to the Proceedings of the ACM Web Conference 2023 (WWW '23) | null | 10.1145/3543507.3583274 | null | cs.DC | http://creativecommons.org/licenses/by/4.0/ | AI-based monitoring has become crucial for cloud-based services due to its
scale. A common approach to AI-based monitoring is to detect causal
relationships among service components and build a causal graph. Availability
of domain information makes cloud systems even better suited for such causal
detection approaches. In modern cloud systems, however, auto-scalers
dynamically change the number of microservice instances, and a load-balancer
manages the load on each instance. This poses a challenge for off-the-shelf
causal structure detection techniques as they neither incorporate the system
architectural domain information nor provide a way to model distributed compute
across varying numbers of service instances. To address this, we develop
CausIL, which detects a causal structure among service metrics by considering
compute distributed across dynamic instances and incorporating domain knowledge
derived from system architecture. Towards the application in cloud systems,
CausIL estimates a causal graph using instance-specific variations in
performance metrics, modeling multiple instances of a service as independent,
conditional on system assumptions. Simulation study shows the efficacy of
CausIL over baselines by improving graph estimation accuracy by ~25% as
measured by Structural Hamming Distance whereas the real-world dataset
demonstrates CausIL's applicability in deployment settings.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 14:56:20 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Mar 2023 14:14:17 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Chakraborty",
"Sarthak",
""
],
[
"Garg",
"Shaddy",
""
],
[
"Agarwal",
"Shubham",
""
],
[
"Chauhan",
"Ayush",
""
],
[
"Saini",
"Shiv Kumar",
""
]
] | TITLE: CausIL: Causal Graph for Instance Level Microservice Data
ABSTRACT: AI-based monitoring has become crucial for cloud-based services due to its
scale. A common approach to AI-based monitoring is to detect causal
relationships among service components and build a causal graph. Availability
of domain information makes cloud systems even better suited for such causal
detection approaches. In modern cloud systems, however, auto-scalers
dynamically change the number of microservice instances, and a load-balancer
manages the load on each instance. This poses a challenge for off-the-shelf
causal structure detection techniques as they neither incorporate the system
architectural domain information nor provide a way to model distributed compute
across varying numbers of service instances. To address this, we develop
CausIL, which detects a causal structure among service metrics by considering
compute distributed across dynamic instances and incorporating domain knowledge
derived from system architecture. Towards the application in cloud systems,
CausIL estimates a causal graph using instance-specific variations in
performance metrics, modeling multiple instances of a service as independent,
conditional on system assumptions. Simulation study shows the efficacy of
CausIL over baselines by improving graph estimation accuracy by ~25% as
measured by Structural Hamming Distance whereas the real-world dataset
demonstrates CausIL's applicability in deployment settings.
| no_new_dataset | 0.941654 |
2303.01498 | Dimitrios Kollias | Dimitrios Kollias and Panagiotis Tzirakis and Alice Baird and Alan
Cowen and Stefanos Zafeiriou | ABAW: Valence-Arousal Estimation, Expression Recognition, Action Unit
Detection & Emotional Reaction Intensity Estimation Challenges | arXiv admin note: text overlap with arXiv:2202.10659 | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The fifth Affective Behavior Analysis in-the-wild (ABAW) Competition is part
of the respective ABAW Workshop which will be held in conjunction with IEEE
Computer Vision and Pattern Recognition Conference (CVPR), 2023. The 5th ABAW
Competition is a continuation of the Competitions held at ECCV 2022, IEEE CVPR
2022, ICCV 2021, IEEE FG 2020 and CVPR 2017 Conferences, and is dedicated at
automatically analyzing affect. For this year's Competition, we feature two
corpora: i) an extended version of the Aff-Wild2 database and ii) the
Hume-Reaction dataset. The former database is an audiovisual one of around 600
videos of around 3M frames and is annotated with respect to:a) two continuous
affect dimensions -valence (how positive/negative a person is) and arousal (how
active/passive a person is)-; b) basic expressions (e.g. happiness, sadness,
neutral state); and c) atomic facial muscle actions (i.e., action units). The
latter dataset is an audiovisual one in which reactions of individuals to
emotional stimuli have been annotated with respect to seven emotional
expression intensities. Thus the 5th ABAW Competition encompasses four
Challenges: i) uni-task Valence-Arousal Estimation, ii) uni-task Expression
Classification, iii) uni-task Action Unit Detection, and iv) Emotional Reaction
Intensity Estimation. In this paper, we present these Challenges, along with
their corpora, we outline the evaluation metrics, we present the baseline
systems and illustrate their obtained performance.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 18:58:15 GMT"
},
{
"version": "v2",
"created": "Fri, 10 Mar 2023 04:49:02 GMT"
},
{
"version": "v3",
"created": "Mon, 20 Mar 2023 15:25:09 GMT"
}
] | 2023-03-21T00:00:00 | [
[
"Kollias",
"Dimitrios",
""
],
[
"Tzirakis",
"Panagiotis",
""
],
[
"Baird",
"Alice",
""
],
[
"Cowen",
"Alan",
""
],
[
"Zafeiriou",
"Stefanos",
""
]
] | TITLE: ABAW: Valence-Arousal Estimation, Expression Recognition, Action Unit
Detection & Emotional Reaction Intensity Estimation Challenges
ABSTRACT: The fifth Affective Behavior Analysis in-the-wild (ABAW) Competition is part
of the respective ABAW Workshop which will be held in conjunction with IEEE
Computer Vision and Pattern Recognition Conference (CVPR), 2023. The 5th ABAW
Competition is a continuation of the Competitions held at ECCV 2022, IEEE CVPR
2022, ICCV 2021, IEEE FG 2020 and CVPR 2017 Conferences, and is dedicated at
automatically analyzing affect. For this year's Competition, we feature two
corpora: i) an extended version of the Aff-Wild2 database and ii) the
Hume-Reaction dataset. The former database is an audiovisual one of around 600
videos of around 3M frames and is annotated with respect to:a) two continuous
affect dimensions -valence (how positive/negative a person is) and arousal (how
active/passive a person is)-; b) basic expressions (e.g. happiness, sadness,
neutral state); and c) atomic facial muscle actions (i.e., action units). The
latter dataset is an audiovisual one in which reactions of individuals to
emotional stimuli have been annotated with respect to seven emotional
expression intensities. Thus the 5th ABAW Competition encompasses four
Challenges: i) uni-task Valence-Arousal Estimation, ii) uni-task Expression
Classification, iii) uni-task Action Unit Detection, and iv) Emotional Reaction
Intensity Estimation. In this paper, we present these Challenges, along with
their corpora, we outline the evaluation metrics, we present the baseline
systems and illustrate their obtained performance.
| new_dataset | 0.964389 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.