id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2306.06263 | Yaroslav Kivva | Yaroslav Kivva, Saber Salehkaleybar, Negar Kiyavash | A Cross-Moment Approach for Causal Effect Estimation | null | null | null | null | cs.CE math.ST stat.TH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of estimating the causal effect of a treatment on an
outcome in linear structural causal models (SCM) with latent confounders when
we have access to a single proxy variable. Several methods (such as
difference-in-difference (DiD) estimator or negative outcome control) have been
proposed in this setting in the literature. However, these approaches require
either restrictive assumptions on the data generating model or having access to
at least two proxy variables. We propose a method to estimate the causal effect
using cross moments between the treatment, the outcome, and the proxy variable.
In particular, we show that the causal effect can be identified with simple
arithmetic operations on the cross moments if the latent confounder in linear
SCM is non-Gaussian. In this setting, DiD estimator provides an unbiased
estimate only in the special case where the latent confounder has exactly the
same direct causal effects on the outcomes in the pre-treatment and
post-treatment phases. This translates to the common trend assumption in DiD,
which we effectively relax. Additionally, we provide an impossibility result
that shows the causal effect cannot be identified if the observational
distribution over the treatment, the outcome, and the proxy is jointly
Gaussian.
Our experiments on both synthetic and real-world datasets showcase the
effectiveness of the proposed approach in estimating the causal effect.
| [
{
"version": "v1",
"created": "Fri, 9 Jun 2023 21:22:37 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Kivva",
"Yaroslav",
""
],
[
"Salehkaleybar",
"Saber",
""
],
[
"Kiyavash",
"Negar",
""
]
] | TITLE: A Cross-Moment Approach for Causal Effect Estimation
ABSTRACT: We consider the problem of estimating the causal effect of a treatment on an
outcome in linear structural causal models (SCM) with latent confounders when
we have access to a single proxy variable. Several methods (such as
difference-in-difference (DiD) estimator or negative outcome control) have been
proposed in this setting in the literature. However, these approaches require
either restrictive assumptions on the data generating model or having access to
at least two proxy variables. We propose a method to estimate the causal effect
using cross moments between the treatment, the outcome, and the proxy variable.
In particular, we show that the causal effect can be identified with simple
arithmetic operations on the cross moments if the latent confounder in linear
SCM is non-Gaussian. In this setting, DiD estimator provides an unbiased
estimate only in the special case where the latent confounder has exactly the
same direct causal effects on the outcomes in the pre-treatment and
post-treatment phases. This translates to the common trend assumption in DiD,
which we effectively relax. Additionally, we provide an impossibility result
that shows the causal effect cannot be identified if the observational
distribution over the treatment, the outcome, and the proxy is jointly
Gaussian.
Our experiments on both synthetic and real-world datasets showcase the
effectiveness of the proposed approach in estimating the causal effect.
|
2306.06292 | Sean Cottrell | Sean Cottrell, Rui Wang, Guowei Wei | PLPCA: Persistent Laplacian Enhanced-PCA for Microarray Data Analysis | 24 pages, 15 figures | null | null | null | math.AT cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Over the years, Principal Component Analysis (PCA) has served as the baseline
approach for dimensionality reduction in gene expression data analysis. It
primary objective is to identify a subset of disease-causing genes from a vast
pool of thousands of genes. However, PCA possesses inherent limitations that
hinder its interpretability, introduce classification ambiguity, and fail to
capture complex geometric structures in the data. Although these limitations
have been partially addressed in the literature by incorporating various
regularizers such as graph Laplacian regularization, existing improved PCA
methods still face challenges related to multiscale analysis and capturing
higher-order interactions in the data. To address these challenges, we propose
a novel approach called Persistent Laplacian-enhanced Principal Component
Analysis (PLPCA). PLPCA amalgamates the advantages of earlier regularized PCA
methods with persistent spectral graph theory, specifically persistent
Laplacians derived from algebraic topology. In contrast to graph Laplacians,
persistent Laplacians enable multiscale analysis through filtration and
incorporate higher-order simplicial complexes to capture higher-order
interactions in the data. We evaluate and validate the performance of PLPCA
using benchmark microarray datasets that involve normal tissue samples and four
different cancer tissues. Our extensive studies demonstrate that PLPCA
outperforms all other state-of-the-art models for classification tasks after
dimensionality reduction.
| [
{
"version": "v1",
"created": "Fri, 9 Jun 2023 22:48:14 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Cottrell",
"Sean",
""
],
[
"Wang",
"Rui",
""
],
[
"Wei",
"Guowei",
""
]
] | TITLE: PLPCA: Persistent Laplacian Enhanced-PCA for Microarray Data Analysis
ABSTRACT: Over the years, Principal Component Analysis (PCA) has served as the baseline
approach for dimensionality reduction in gene expression data analysis. It
primary objective is to identify a subset of disease-causing genes from a vast
pool of thousands of genes. However, PCA possesses inherent limitations that
hinder its interpretability, introduce classification ambiguity, and fail to
capture complex geometric structures in the data. Although these limitations
have been partially addressed in the literature by incorporating various
regularizers such as graph Laplacian regularization, existing improved PCA
methods still face challenges related to multiscale analysis and capturing
higher-order interactions in the data. To address these challenges, we propose
a novel approach called Persistent Laplacian-enhanced Principal Component
Analysis (PLPCA). PLPCA amalgamates the advantages of earlier regularized PCA
methods with persistent spectral graph theory, specifically persistent
Laplacians derived from algebraic topology. In contrast to graph Laplacians,
persistent Laplacians enable multiscale analysis through filtration and
incorporate higher-order simplicial complexes to capture higher-order
interactions in the data. We evaluate and validate the performance of PLPCA
using benchmark microarray datasets that involve normal tissue samples and four
different cancer tissues. Our extensive studies demonstrate that PLPCA
outperforms all other state-of-the-art models for classification tasks after
dimensionality reduction.
|
2306.06294 | Jiong Yang | Jiong Yang, Arijit Shaw, Teodora Baluta, Mate Soos, and Kuldeep S.
Meel | Explaining SAT Solving Using Causal Reasoning | 17 pages, 3 figures, to be published in SAT23 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The past three decades have witnessed notable success in designing efficient
SAT solvers, with modern solvers capable of solving industrial benchmarks
containing millions of variables in just a few seconds. The success of modern
SAT solvers owes to the widely-used CDCL algorithm, which lacks comprehensive
theoretical investigation. Furthermore, it has been observed that CDCL solvers
still struggle to deal with specific classes of benchmarks comprising only
hundreds of variables, which contrasts with their widespread use in real-world
applications. Consequently, there is an urgent need to uncover the inner
workings of these seemingly weak yet powerful black boxes.
In this paper, we present a first step towards this goal by introducing an
approach called CausalSAT, which employs causal reasoning to gain insights into
the functioning of modern SAT solvers. CausalSAT initially generates
observational data from the execution of SAT solvers and learns a structured
graph representing the causal relationships between the components of a SAT
solver. Subsequently, given a query such as whether a clause with low literals
blocks distance (LBD) has a higher clause utility, CausalSAT calculates the
causal effect of LBD on clause utility and provides an answer to the question.
We use CausalSAT to quantitatively verify hypotheses previously regarded as
"rules of thumb" or empirical findings such as the query above. Moreover,
CausalSAT can address previously unexplored questions, like which branching
heuristic leads to greater clause utility in order to study the relationship
between branching and clause management. Experimental evaluations using
practical benchmarks demonstrate that CausalSAT effectively fits the data,
verifies four "rules of thumb", and provides answers to three questions closely
related to implementing modern solvers.
| [
{
"version": "v1",
"created": "Fri, 9 Jun 2023 22:53:16 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Yang",
"Jiong",
""
],
[
"Shaw",
"Arijit",
""
],
[
"Baluta",
"Teodora",
""
],
[
"Soos",
"Mate",
""
],
[
"Meel",
"Kuldeep S.",
""
]
] | TITLE: Explaining SAT Solving Using Causal Reasoning
ABSTRACT: The past three decades have witnessed notable success in designing efficient
SAT solvers, with modern solvers capable of solving industrial benchmarks
containing millions of variables in just a few seconds. The success of modern
SAT solvers owes to the widely-used CDCL algorithm, which lacks comprehensive
theoretical investigation. Furthermore, it has been observed that CDCL solvers
still struggle to deal with specific classes of benchmarks comprising only
hundreds of variables, which contrasts with their widespread use in real-world
applications. Consequently, there is an urgent need to uncover the inner
workings of these seemingly weak yet powerful black boxes.
In this paper, we present a first step towards this goal by introducing an
approach called CausalSAT, which employs causal reasoning to gain insights into
the functioning of modern SAT solvers. CausalSAT initially generates
observational data from the execution of SAT solvers and learns a structured
graph representing the causal relationships between the components of a SAT
solver. Subsequently, given a query such as whether a clause with low literals
blocks distance (LBD) has a higher clause utility, CausalSAT calculates the
causal effect of LBD on clause utility and provides an answer to the question.
We use CausalSAT to quantitatively verify hypotheses previously regarded as
"rules of thumb" or empirical findings such as the query above. Moreover,
CausalSAT can address previously unexplored questions, like which branching
heuristic leads to greater clause utility in order to study the relationship
between branching and clause management. Experimental evaluations using
practical benchmarks demonstrate that CausalSAT effectively fits the data,
verifies four "rules of thumb", and provides answers to three questions closely
related to implementing modern solvers.
|
2306.06296 | Michael Shvartsman | Michael Shvartsman, Benjamin Letham, Stephen Keeley | Response Time Improves Choice Prediction and Function Estimation for
Gaussian Process Models of Perception and Preferences | 18 pages incl. references and supplement; 11 figures | null | null | null | q-bio.NC cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Models for human choice prediction in preference learning and psychophysics
often consider only binary response data, requiring many samples to accurately
learn preferences or perceptual detection thresholds. The response time (RT) to
make each choice captures additional information about the decision process,
however existing models incorporating RTs for choice prediction do so in fully
parametric settings or over discrete stimulus sets. This is in part because the
de-facto standard model for choice RTs, the diffusion decision model (DDM),
does not admit tractable, differentiable inference. The DDM thus cannot be
easily integrated with flexible models for continuous, multivariate function
approximation, particularly Gaussian process (GP) models. We propose a novel
differentiable approximation to the DDM likelihood using a family of known,
skewed three-parameter distributions. We then use this new likelihood to
incorporate RTs into GP models for binary choices. Our RT-choice GPs enable
both better latent value estimation and held-out choice prediction relative to
baselines, which we demonstrate on three real-world multivariate datasets
covering both human psychophysics and preference learning applications.
| [
{
"version": "v1",
"created": "Fri, 9 Jun 2023 23:22:49 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shvartsman",
"Michael",
""
],
[
"Letham",
"Benjamin",
""
],
[
"Keeley",
"Stephen",
""
]
] | TITLE: Response Time Improves Choice Prediction and Function Estimation for
Gaussian Process Models of Perception and Preferences
ABSTRACT: Models for human choice prediction in preference learning and psychophysics
often consider only binary response data, requiring many samples to accurately
learn preferences or perceptual detection thresholds. The response time (RT) to
make each choice captures additional information about the decision process,
however existing models incorporating RTs for choice prediction do so in fully
parametric settings or over discrete stimulus sets. This is in part because the
de-facto standard model for choice RTs, the diffusion decision model (DDM),
does not admit tractable, differentiable inference. The DDM thus cannot be
easily integrated with flexible models for continuous, multivariate function
approximation, particularly Gaussian process (GP) models. We propose a novel
differentiable approximation to the DDM likelihood using a family of known,
skewed three-parameter distributions. We then use this new likelihood to
incorporate RTs into GP models for binary choices. Our RT-choice GPs enable
both better latent value estimation and held-out choice prediction relative to
baselines, which we demonstrate on three real-world multivariate datasets
covering both human psychophysics and preference learning applications.
|
2306.06322 | Abdelhamid Haouhat | Abdelhamid Haouhat, Slimane Bellaouar, Attia Nehar, Hadda Cherroun | Towards Arabic Multimodal Dataset for Sentiment Analysis | 8 pages | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Multimodal Sentiment Analysis (MSA) has recently become a centric research
direction for many real-world applications. This proliferation is due to the
fact that opinions are central to almost all human activities and are key
influencers of our behaviors. In addition, the recent deployment of Deep
Learning-based (DL) models has proven their high efficiency for a wide range of
Western languages. In contrast, Arabic DL-based multimodal sentiment analysis
(MSA) is still in its infantile stage due, mainly, to the lack of standard
datasets. In this paper, our investigation is twofold. First, we design a
pipeline that helps building our Arabic Multimodal dataset leveraging both
state-of-the-art transformers and feature extraction tools within word
alignment techniques. Thereafter, we validate our dataset using
state-of-the-art transformer-based model dealing with multimodality. Despite
the small size of the outcome dataset, experiments show that Arabic
multimodality is very promising
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 00:13:09 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Haouhat",
"Abdelhamid",
""
],
[
"Bellaouar",
"Slimane",
""
],
[
"Nehar",
"Attia",
""
],
[
"Cherroun",
"Hadda",
""
]
] | TITLE: Towards Arabic Multimodal Dataset for Sentiment Analysis
ABSTRACT: Multimodal Sentiment Analysis (MSA) has recently become a centric research
direction for many real-world applications. This proliferation is due to the
fact that opinions are central to almost all human activities and are key
influencers of our behaviors. In addition, the recent deployment of Deep
Learning-based (DL) models has proven their high efficiency for a wide range of
Western languages. In contrast, Arabic DL-based multimodal sentiment analysis
(MSA) is still in its infantile stage due, mainly, to the lack of standard
datasets. In this paper, our investigation is twofold. First, we design a
pipeline that helps building our Arabic Multimodal dataset leveraging both
state-of-the-art transformers and feature extraction tools within word
alignment techniques. Thereafter, we validate our dataset using
state-of-the-art transformer-based model dealing with multimodality. Despite
the small size of the outcome dataset, experiments show that Arabic
multimodality is very promising
|
2306.06329 | Shixi Lian | Shixi Lian, Yi Ma, Jinyi Liu, Yan Zheng, Zhaopeng Meng | HIPODE: Enhancing Offline Reinforcement Learning with High-Quality
Synthetic Data from a Policy-Decoupled Approach | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Offline reinforcement learning (ORL) has gained attention as a means of
training reinforcement learning models using pre-collected static data. To
address the issue of limited data and improve downstream ORL performance,
recent work has attempted to expand the dataset's coverage through data
augmentation. However, most of these methods are tied to a specific policy
(policy-dependent), where the generated data can only guarantee to support the
current downstream ORL policy, limiting its usage scope on other downstream
policies. Moreover, the quality of synthetic data is often not well-controlled,
which limits the potential for further improving the downstream policy. To
tackle these issues, we propose \textbf{HI}gh-quality
\textbf{PO}licy-\textbf{DE}coupled~(HIPODE), a novel data augmentation method
for ORL. On the one hand, HIPODE generates high-quality synthetic data by
selecting states near the dataset distribution with potentially high value
among candidate states using the negative sampling technique. On the other
hand, HIPODE is policy-decoupled, thus can be used as a common plug-in method
for any downstream ORL process. We conduct experiments on the widely studied
TD3BC and CQL algorithms, and the results show that HIPODE outperforms the
state-of-the-art policy-decoupled data augmentation method and most prevalent
model-based ORL methods on D4RL benchmarks.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 01:49:01 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Lian",
"Shixi",
""
],
[
"Ma",
"Yi",
""
],
[
"Liu",
"Jinyi",
""
],
[
"Zheng",
"Yan",
""
],
[
"Meng",
"Zhaopeng",
""
]
] | TITLE: HIPODE: Enhancing Offline Reinforcement Learning with High-Quality
Synthetic Data from a Policy-Decoupled Approach
ABSTRACT: Offline reinforcement learning (ORL) has gained attention as a means of
training reinforcement learning models using pre-collected static data. To
address the issue of limited data and improve downstream ORL performance,
recent work has attempted to expand the dataset's coverage through data
augmentation. However, most of these methods are tied to a specific policy
(policy-dependent), where the generated data can only guarantee to support the
current downstream ORL policy, limiting its usage scope on other downstream
policies. Moreover, the quality of synthetic data is often not well-controlled,
which limits the potential for further improving the downstream policy. To
tackle these issues, we propose \textbf{HI}gh-quality
\textbf{PO}licy-\textbf{DE}coupled~(HIPODE), a novel data augmentation method
for ORL. On the one hand, HIPODE generates high-quality synthetic data by
selecting states near the dataset distribution with potentially high value
among candidate states using the negative sampling technique. On the other
hand, HIPODE is policy-decoupled, thus can be used as a common plug-in method
for any downstream ORL process. We conduct experiments on the widely studied
TD3BC and CQL algorithms, and the results show that HIPODE outperforms the
state-of-the-art policy-decoupled data augmentation method and most prevalent
model-based ORL methods on D4RL benchmarks.
|
2306.06367 | Shuo Huang | Shuo Huang, Jia Jia, Zongxin Yang, Wei Wang, Haozhe Wu, Yi Yang,
Junliang Xing | Shuffled Autoregression For Motion Interpolation | accepted by ICASSP 2023 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work aims to provide a deep-learning solution for the motion
interpolation task. Previous studies solve it with geometric weight functions.
Some other works propose neural networks for different problem settings with
consecutive pose sequences as input. However, motion interpolation is a more
complex problem that takes isolated poses (e.g., only one start pose and one
end pose) as input. When applied to motion interpolation, these deep learning
methods have limited performance since they do not leverage the flexible
dependencies between interpolation frames as the original geometric formulas
do. To realize this interpolation characteristic, we propose a novel framework,
referred to as \emph{Shuffled AutoRegression}, which expands the autoregression
to generate in arbitrary (shuffled) order and models any inter-frame
dependencies as a directed acyclic graph. We further propose an approach to
constructing a particular kind of dependency graph, with three stages assembled
into an end-to-end spatial-temporal motion Transformer. Experimental results on
one of the current largest datasets show that our model generates vivid and
coherent motions from only one start frame to one end frame and outperforms
competing methods by a large margin. The proposed model is also extensible to
multiple keyframes' motion interpolation tasks and other areas' interpolation.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 07:14:59 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Huang",
"Shuo",
""
],
[
"Jia",
"Jia",
""
],
[
"Yang",
"Zongxin",
""
],
[
"Wang",
"Wei",
""
],
[
"Wu",
"Haozhe",
""
],
[
"Yang",
"Yi",
""
],
[
"Xing",
"Junliang",
""
]
] | TITLE: Shuffled Autoregression For Motion Interpolation
ABSTRACT: This work aims to provide a deep-learning solution for the motion
interpolation task. Previous studies solve it with geometric weight functions.
Some other works propose neural networks for different problem settings with
consecutive pose sequences as input. However, motion interpolation is a more
complex problem that takes isolated poses (e.g., only one start pose and one
end pose) as input. When applied to motion interpolation, these deep learning
methods have limited performance since they do not leverage the flexible
dependencies between interpolation frames as the original geometric formulas
do. To realize this interpolation characteristic, we propose a novel framework,
referred to as \emph{Shuffled AutoRegression}, which expands the autoregression
to generate in arbitrary (shuffled) order and models any inter-frame
dependencies as a directed acyclic graph. We further propose an approach to
constructing a particular kind of dependency graph, with three stages assembled
into an end-to-end spatial-temporal motion Transformer. Experimental results on
one of the current largest datasets show that our model generates vivid and
coherent motions from only one start frame to one end frame and outperforms
competing methods by a large margin. The proposed model is also extensible to
multiple keyframes' motion interpolation tasks and other areas' interpolation.
|
2306.06370 | Tal Shaharabany | Tal Shaharabany, Aviad Dahan, Raja Giryes and Lior Wolf | AutoSAM: Adapting SAM to Medical Images by Overloading the Prompt
Encoder | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recently introduced Segment Anything Model (SAM) combines a clever
architecture and large quantities of training data to obtain remarkable image
segmentation capabilities. However, it fails to reproduce such results for
Out-Of-Distribution (OOD) domains such as medical images. Moreover, while SAM
is conditioned on either a mask or a set of points, it may be desirable to have
a fully automatic solution. In this work, we replace SAM's conditioning with an
encoder that operates on the same input image. By adding this encoder and
without further fine-tuning SAM, we obtain state-of-the-art results on multiple
medical images and video benchmarks. This new encoder is trained via gradients
provided by a frozen SAM. For inspecting the knowledge within it, and providing
a lightweight segmentation solution, we also learn to decode it into a mask by
a shallow deconvolution network.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 07:27:00 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shaharabany",
"Tal",
""
],
[
"Dahan",
"Aviad",
""
],
[
"Giryes",
"Raja",
""
],
[
"Wolf",
"Lior",
""
]
] | TITLE: AutoSAM: Adapting SAM to Medical Images by Overloading the Prompt
Encoder
ABSTRACT: The recently introduced Segment Anything Model (SAM) combines a clever
architecture and large quantities of training data to obtain remarkable image
segmentation capabilities. However, it fails to reproduce such results for
Out-Of-Distribution (OOD) domains such as medical images. Moreover, while SAM
is conditioned on either a mask or a set of points, it may be desirable to have
a fully automatic solution. In this work, we replace SAM's conditioning with an
encoder that operates on the same input image. By adding this encoder and
without further fine-tuning SAM, we obtain state-of-the-art results on multiple
medical images and video benchmarks. This new encoder is trained via gradients
provided by a frozen SAM. For inspecting the knowledge within it, and providing
a lightweight segmentation solution, we also learn to decode it into a mask by
a shallow deconvolution network.
|
2306.06371 | El Hassane Ettifouri | Jessica L\'opez Espejel, Mahaman Sanoussi Yahaya Alassan, El Mehdi
Chouham, Walid Dahhane, El Hassane Ettifouri | A Comprehensive Review of State-of-The-Art Methods for Java Code
Generation from Natural Language Text | Published at Elsevier's Natural Language Processing Journal | Natural Language Processing Journal, Volume 3, 2023, 100013, ISSN
2949-7191 | 10.1016/j.nlp.2023.100013 | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Java Code Generation consists in generating automatically Java code from a
Natural Language Text. This NLP task helps in increasing programmers'
productivity by providing them with immediate solutions to the simplest and
most repetitive tasks. Code generation is a challenging task because of the
hard syntactic rules and the necessity of a deep understanding of the semantic
aspect of the programming language. Many works tried to tackle this task using
either RNN-based, or Transformer-based models. The latter achieved remarkable
advancement in the domain and they can be divided into three groups: (1)
encoder-only models, (2) decoder-only models, and (3) encoder-decoder models.
In this paper, we provide a comprehensive review of the evolution and progress
of deep learning models in Java code generation task. We focus on the most
important methods and present their merits and limitations, as well as the
objective functions used by the community. In addition, we provide a detailed
description of datasets and evaluation metrics used in the literature. Finally,
we discuss results of different models on CONCODE dataset, then propose some
future directions.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 07:27:51 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Espejel",
"Jessica López",
""
],
[
"Alassan",
"Mahaman Sanoussi Yahaya",
""
],
[
"Chouham",
"El Mehdi",
""
],
[
"Dahhane",
"Walid",
""
],
[
"Ettifouri",
"El Hassane",
""
]
] | TITLE: A Comprehensive Review of State-of-The-Art Methods for Java Code
Generation from Natural Language Text
ABSTRACT: Java Code Generation consists in generating automatically Java code from a
Natural Language Text. This NLP task helps in increasing programmers'
productivity by providing them with immediate solutions to the simplest and
most repetitive tasks. Code generation is a challenging task because of the
hard syntactic rules and the necessity of a deep understanding of the semantic
aspect of the programming language. Many works tried to tackle this task using
either RNN-based, or Transformer-based models. The latter achieved remarkable
advancement in the domain and they can be divided into three groups: (1)
encoder-only models, (2) decoder-only models, and (3) encoder-decoder models.
In this paper, we provide a comprehensive review of the evolution and progress
of deep learning models in Java code generation task. We focus on the most
important methods and present their merits and limitations, as well as the
objective functions used by the community. In addition, we provide a detailed
description of datasets and evaluation metrics used in the literature. Finally,
we discuss results of different models on CONCODE dataset, then propose some
future directions.
|
2306.06376 | Marco Montali | Sander J. J. Leemans, Fabrizio M. Maggi, Marco Montali | Enjoy the Silence: Analysis of Stochastic Petri Nets with Silent
Transitions | null | null | null | null | cs.LO cs.AI | http://creativecommons.org/licenses/by/4.0/ | Capturing stochastic behaviors in business and work processes is essential to
quantitatively understand how nondeterminism is resolved when taking decisions
within the process. This is of special interest in process mining, where event
data tracking the actual execution of the process are related to process
models, and can then provide insights on frequencies and probabilities.
Variants of stochastic Petri nets provide a natural formal basis for this.
However, when capturing processes, such nets need to be labelled with (possibly
duplicated) activities, and equipped with silent transitions that model
internal, non-logged steps related to the orchestration of the process. At the
same time, they have to be analyzed in a finite-trace semantics, matching the
fact that each process execution consists of finitely many steps. These two
aspects impede the direct application of existing techniques for stochastic
Petri nets, calling for a novel characterization that incorporates labels and
silent transitions in a finite-trace semantics. In this article, we provide
such a characterization starting from generalized stochastic Petri nets and
obtaining the framework of labelled stochastic processes (LSPs). On top of this
framework, we introduce different key analysis tasks on the traces of LSPs and
their probabilities. We show that all such analysis tasks can be solved
analytically, in particular reducing them to a single method that combines
automata-based techniques to single out the behaviors of interest within a LSP,
with techniques based on absorbing Markov chains to reason on their
probabilities. Finally, we demonstrate the significance of how our approach in
the context of stochastic conformance checking, illustrating practical
feasibility through a proof-of-concept implementation and its application to
different datasets.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 07:57:24 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Leemans",
"Sander J. J.",
""
],
[
"Maggi",
"Fabrizio M.",
""
],
[
"Montali",
"Marco",
""
]
] | TITLE: Enjoy the Silence: Analysis of Stochastic Petri Nets with Silent
Transitions
ABSTRACT: Capturing stochastic behaviors in business and work processes is essential to
quantitatively understand how nondeterminism is resolved when taking decisions
within the process. This is of special interest in process mining, where event
data tracking the actual execution of the process are related to process
models, and can then provide insights on frequencies and probabilities.
Variants of stochastic Petri nets provide a natural formal basis for this.
However, when capturing processes, such nets need to be labelled with (possibly
duplicated) activities, and equipped with silent transitions that model
internal, non-logged steps related to the orchestration of the process. At the
same time, they have to be analyzed in a finite-trace semantics, matching the
fact that each process execution consists of finitely many steps. These two
aspects impede the direct application of existing techniques for stochastic
Petri nets, calling for a novel characterization that incorporates labels and
silent transitions in a finite-trace semantics. In this article, we provide
such a characterization starting from generalized stochastic Petri nets and
obtaining the framework of labelled stochastic processes (LSPs). On top of this
framework, we introduce different key analysis tasks on the traces of LSPs and
their probabilities. We show that all such analysis tasks can be solved
analytically, in particular reducing them to a single method that combines
automata-based techniques to single out the behaviors of interest within a LSP,
with techniques based on absorbing Markov chains to reason on their
probabilities. Finally, we demonstrate the significance of how our approach in
the context of stochastic conformance checking, illustrating practical
feasibility through a proof-of-concept implementation and its application to
different datasets.
|
2306.06381 | Wenhao Zhu | Wenhao Zhu, Jingjing Xu, Shujian Huang, Lingpeng Kong, Jiajun Chen | INK: Injecting kNN Knowledge in Nearest Neighbor Machine Translation | 10 pages, ACL-2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural machine translation has achieved promising results on many translation
tasks. However, previous studies have shown that neural models induce a
non-smooth representation space, which harms its generalization results.
Recently, kNN-MT has provided an effective paradigm to smooth the prediction
based on neighbor representations during inference. Despite promising results,
kNN-MT usually requires large inference overhead. We propose an effective
training framework INK to directly smooth the representation space via
adjusting representations of kNN neighbors with a small number of new
parameters. The new parameters are then used to refresh the whole
representation datastore to get new kNN knowledge asynchronously. This loop
keeps running until convergence. Experiments on four benchmark datasets show
that \method achieves average gains of 1.99 COMET and 1.0 BLEU, outperforming
the state-of-the-art kNN-MT system with 0.02x memory space and 1.9x inference
speedup.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 08:39:16 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhu",
"Wenhao",
""
],
[
"Xu",
"Jingjing",
""
],
[
"Huang",
"Shujian",
""
],
[
"Kong",
"Lingpeng",
""
],
[
"Chen",
"Jiajun",
""
]
] | TITLE: INK: Injecting kNN Knowledge in Nearest Neighbor Machine Translation
ABSTRACT: Neural machine translation has achieved promising results on many translation
tasks. However, previous studies have shown that neural models induce a
non-smooth representation space, which harms its generalization results.
Recently, kNN-MT has provided an effective paradigm to smooth the prediction
based on neighbor representations during inference. Despite promising results,
kNN-MT usually requires large inference overhead. We propose an effective
training framework INK to directly smooth the representation space via
adjusting representations of kNN neighbors with a small number of new
parameters. The new parameters are then used to refresh the whole
representation datastore to get new kNN knowledge asynchronously. This loop
keeps running until convergence. Experiments on four benchmark datasets show
that \method achieves average gains of 1.99 COMET and 1.0 BLEU, outperforming
the state-of-the-art kNN-MT system with 0.02x memory space and 1.9x inference
speedup.
|
2306.06384 | Vineet Bhat | Vineet Bhat, Preethi Jyothi and Pushpak Bhattacharyya | Adversarial Training For Low-Resource Disfluency Correction | Accepted for Findings of ACL 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Disfluencies commonly occur in conversational speech. Speech with
disfluencies can result in noisy Automatic Speech Recognition (ASR)
transcripts, which affects downstream tasks like machine translation. In this
paper, we propose an adversarially-trained sequence-tagging model for
Disfluency Correction (DC) that utilizes a small amount of labeled real
disfluent data in conjunction with a large amount of unlabeled data. We show
the benefit of our proposed technique, which crucially depends on synthetically
generated disfluent data, by evaluating it for DC in three Indian languages-
Bengali, Hindi, and Marathi (all from the Indo-Aryan family). Our technique
also performs well in removing stuttering disfluencies in ASR transcripts
introduced by speech impairments. We achieve an average 6.15 points improvement
in F1-score over competitive baselines across all three languages mentioned. To
the best of our knowledge, we are the first to utilize adversarial training for
DC and use it to correct stuttering disfluencies in English, establishing a new
benchmark for this task.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 08:58:53 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Bhat",
"Vineet",
""
],
[
"Jyothi",
"Preethi",
""
],
[
"Bhattacharyya",
"Pushpak",
""
]
] | TITLE: Adversarial Training For Low-Resource Disfluency Correction
ABSTRACT: Disfluencies commonly occur in conversational speech. Speech with
disfluencies can result in noisy Automatic Speech Recognition (ASR)
transcripts, which affects downstream tasks like machine translation. In this
paper, we propose an adversarially-trained sequence-tagging model for
Disfluency Correction (DC) that utilizes a small amount of labeled real
disfluent data in conjunction with a large amount of unlabeled data. We show
the benefit of our proposed technique, which crucially depends on synthetically
generated disfluent data, by evaluating it for DC in three Indian languages-
Bengali, Hindi, and Marathi (all from the Indo-Aryan family). Our technique
also performs well in removing stuttering disfluencies in ASR transcripts
introduced by speech impairments. We achieve an average 6.15 points improvement
in F1-score over competitive baselines across all three languages mentioned. To
the best of our knowledge, we are the first to utilize adversarial training for
DC and use it to correct stuttering disfluencies in English, establishing a new
benchmark for this task.
|
2306.06393 | Ramoni Adeogun | Gilberto Berardinelli, Ramoni Adeogun | On the required radio resources for ultra-reliable communication in
highly interfered scenarios | null | in IEEE Communications Letters, vol. 27, no. 6, pp. 1639-1643,
June 2023 | 10.1109/LCOMM.2023.3265265 | null | eess.SP cs.IT math.IT | http://creativecommons.org/licenses/by/4.0/ | Future wireless systems are expected to support mission-critical services
demanding higher and higher reliability. In this letter, we dimension the radio
resources needed to achieve a given failure probability target for
ultra-reliable wireless systems in high interference conditions, assuming a
protocol with frequency hopping combined with packet repetitions. We resort to
packet erasure channel models and derive the minimum amount of resource units
in the case of receiver with and without collision resolution capability, as
well as the number of packet repetitions needed for achieving the failure
probability target. Analytical results are numerically validated and can be
used as a benchmark for realistic system simulations
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 09:40:36 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Berardinelli",
"Gilberto",
""
],
[
"Adeogun",
"Ramoni",
""
]
] | TITLE: On the required radio resources for ultra-reliable communication in
highly interfered scenarios
ABSTRACT: Future wireless systems are expected to support mission-critical services
demanding higher and higher reliability. In this letter, we dimension the radio
resources needed to achieve a given failure probability target for
ultra-reliable wireless systems in high interference conditions, assuming a
protocol with frequency hopping combined with packet repetitions. We resort to
packet erasure channel models and derive the minimum amount of resource units
in the case of receiver with and without collision resolution capability, as
well as the number of packet repetitions needed for achieving the failure
probability target. Analytical results are numerically validated and can be
used as a benchmark for realistic system simulations
|
2306.06423 | Juan M. Gandarias | Francisco Pastor (1), Jorge Garc\'ia-Gonz\'alez (2), Juan M. Gandarias
(1), Daniel Medina (3), Pau Closas (4), Alfonso J. Garc\'ia-Cerezo (1),
Jes\'us M. G\'omez-de-Gabriel (1) ((1) Robotics and Mechatronics Group,
University of Malaga, Spain, (2) Department of Computer Languages and
Computer Science, University of Malaga, Spain, (3) Institute of
Communications and Navigation, German Aerospace Center (DLR), Germany, (4)
Department of Electrical and Computer Engineering, Northeastern University,
Boston, USA) | Bayesian and Neural Inference on LSTM-based Object Recognition from
Tactile and Kinesthetic Information | null | null | 10.1109/LRA.2020.3038377 | null | cs.RO cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in the field of intelligent robotic manipulation pursue
providing robotic hands with touch sensitivity. Haptic perception encompasses
the sensing modalities encountered in the sense of touch (e.g., tactile and
kinesthetic sensations). This letter focuses on multimodal object recognition
and proposes analytical and data-driven methodologies to fuse tactile- and
kinesthetic-based classification results. The procedure is as follows: a
three-finger actuated gripper with an integrated high-resolution tactile sensor
performs squeeze-and-release Exploratory Procedures (EPs). The tactile images
and kinesthetic information acquired using angular sensors on the finger joints
constitute the time-series datasets of interest. Each temporal dataset is fed
to a Long Short-term Memory (LSTM) Neural Network, which is trained to classify
in-hand objects. The LSTMs provide an estimation of the posterior probability
of each object given the corresponding measurements, which after fusion allows
to estimate the object through Bayesian and Neural inference approaches. An
experiment with 36-classes is carried out to evaluate and compare the
performance of the fused, tactile, and kinesthetic perception systems.The
results show that the Bayesian-based classifiers improves capabilities for
object recognition and outperforms the Neural-based approach.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 12:29:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Pastor",
"Francisco",
""
],
[
"García-González",
"Jorge",
""
],
[
"Gandarias",
"Juan M.",
""
],
[
"Medina",
"Daniel",
""
],
[
"Closas",
"Pau",
""
],
[
"García-Cerezo",
"Alfonso J.",
""
],
[
"Gómez-de-Gabriel",
"Jesús M.",
""
]
] | TITLE: Bayesian and Neural Inference on LSTM-based Object Recognition from
Tactile and Kinesthetic Information
ABSTRACT: Recent advances in the field of intelligent robotic manipulation pursue
providing robotic hands with touch sensitivity. Haptic perception encompasses
the sensing modalities encountered in the sense of touch (e.g., tactile and
kinesthetic sensations). This letter focuses on multimodal object recognition
and proposes analytical and data-driven methodologies to fuse tactile- and
kinesthetic-based classification results. The procedure is as follows: a
three-finger actuated gripper with an integrated high-resolution tactile sensor
performs squeeze-and-release Exploratory Procedures (EPs). The tactile images
and kinesthetic information acquired using angular sensors on the finger joints
constitute the time-series datasets of interest. Each temporal dataset is fed
to a Long Short-term Memory (LSTM) Neural Network, which is trained to classify
in-hand objects. The LSTMs provide an estimation of the posterior probability
of each object given the corresponding measurements, which after fusion allows
to estimate the object through Bayesian and Neural inference approaches. An
experiment with 36-classes is carried out to evaluate and compare the
performance of the fused, tactile, and kinesthetic perception systems.The
results show that the Bayesian-based classifiers improves capabilities for
object recognition and outperforms the Neural-based approach.
|
2306.06461 | Ji Won Kim | Ji Won Kim, Sang Won Son, Yoonah Song, Hong Kook Kim, Il Hoon Song,
Jeong Eun Lim | Semi-supervsied Learning-based Sound Event Detection using Freuqency
Dynamic Convolution with Large Kernel Attention for DCASE Challenge 2023 Task
4 | DCASE 2023 Challenge Task 4A, 5 pages | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This report proposes a frequency dynamic convolution (FDY) with a large
kernel attention (LKA)-convolutional recurrent neural network (CRNN) with a
pre-trained bidirectional encoder representation from audio transformers
(BEATs) embedding-based sound event detection (SED) model that employs a
mean-teacher and pseudo-label approach to address the challenge of limited
labeled data for DCASE 2023 Task 4. The proposed FDY with LKA integrates the
FDY and LKA module to effectively capture time-frequency patterns, long-term
dependencies, and high-level semantic information in audio signals. The
proposed FDY with LKA-CRNN with a BEATs embedding network is initially trained
on the entire DCASE 2023 Task 4 dataset using the mean-teacher approach,
generating pseudo-labels for weakly labeled, unlabeled, and the AudioSet.
Subsequently, the proposed SED model is retrained using the same pseudo-label
approach. A subset of these models is selected for submission, demonstrating
superior F1-scores and polyphonic SED score performance on the DCASE 2023
Challenge Task 4 validation dataset.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 15:05:11 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Kim",
"Ji Won",
""
],
[
"Son",
"Sang Won",
""
],
[
"Song",
"Yoonah",
""
],
[
"Kim",
"Hong Kook",
""
],
[
"Song",
"Il Hoon",
""
],
[
"Lim",
"Jeong Eun",
""
]
] | TITLE: Semi-supervsied Learning-based Sound Event Detection using Freuqency
Dynamic Convolution with Large Kernel Attention for DCASE Challenge 2023 Task
4
ABSTRACT: This report proposes a frequency dynamic convolution (FDY) with a large
kernel attention (LKA)-convolutional recurrent neural network (CRNN) with a
pre-trained bidirectional encoder representation from audio transformers
(BEATs) embedding-based sound event detection (SED) model that employs a
mean-teacher and pseudo-label approach to address the challenge of limited
labeled data for DCASE 2023 Task 4. The proposed FDY with LKA integrates the
FDY and LKA module to effectively capture time-frequency patterns, long-term
dependencies, and high-level semantic information in audio signals. The
proposed FDY with LKA-CRNN with a BEATs embedding network is initially trained
on the entire DCASE 2023 Task 4 dataset using the mean-teacher approach,
generating pseudo-labels for weakly labeled, unlabeled, and the AudioSet.
Subsequently, the proposed SED model is retrained using the same pseudo-label
approach. A subset of these models is selected for submission, demonstrating
superior F1-scores and polyphonic SED score performance on the DCASE 2023
Challenge Task 4 validation dataset.
|
2306.06470 | Zefeng Chen | Zefeng Chen, Wensheng Gan, Gengsen Huang, Zhenlian Qi, Yan Li, Philip
S. Yu | TALENT: Targeted Mining of Non-overlapping Sequential Patterns | Preprint. 9 figures, 5 tables | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the widespread application of efficient pattern mining algorithms,
sequential patterns that allow gap constraints have become a valuable tool to
discover knowledge from biological data such as DNA and protein sequences.
Among all kinds of gap-constrained mining, non-overlapping sequence mining can
mine interesting patterns and satisfy the anti-monotonic property (the Apriori
property). However, existing algorithms do not search for targeted sequential
patterns, resulting in unnecessary and redundant pattern generation. Targeted
pattern mining can not only mine patterns that are more interesting to users
but also reduce the unnecessary redundant sequence generated, which can greatly
avoid irrelevant computation. In this paper, we define and formalize the
problem of targeted non-overlapping sequential pattern mining and propose an
algorithm named TALENT (TArgeted mining of sequentiaL pattErN with
consTraints). Two search methods including breadth-first and depth-first
searching are designed to troubleshoot the generation of patterns. Furthermore,
several pruning strategies to reduce the reading of sequences and items in the
data and terminate redundant pattern extensions are presented. Finally, we
select a series of datasets with different characteristics and conduct
extensive experiments to compare the TALENT algorithm with the existing
algorithms for mining non-overlapping sequential patterns. The experimental
results demonstrate that the proposed targeted mining algorithm, TALENT, has
excellent mining efficiency and can deal efficiently with many different query
settings.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 15:47:05 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Chen",
"Zefeng",
""
],
[
"Gan",
"Wensheng",
""
],
[
"Huang",
"Gengsen",
""
],
[
"Qi",
"Zhenlian",
""
],
[
"Li",
"Yan",
""
],
[
"Yu",
"Philip S.",
""
]
] | TITLE: TALENT: Targeted Mining of Non-overlapping Sequential Patterns
ABSTRACT: With the widespread application of efficient pattern mining algorithms,
sequential patterns that allow gap constraints have become a valuable tool to
discover knowledge from biological data such as DNA and protein sequences.
Among all kinds of gap-constrained mining, non-overlapping sequence mining can
mine interesting patterns and satisfy the anti-monotonic property (the Apriori
property). However, existing algorithms do not search for targeted sequential
patterns, resulting in unnecessary and redundant pattern generation. Targeted
pattern mining can not only mine patterns that are more interesting to users
but also reduce the unnecessary redundant sequence generated, which can greatly
avoid irrelevant computation. In this paper, we define and formalize the
problem of targeted non-overlapping sequential pattern mining and propose an
algorithm named TALENT (TArgeted mining of sequentiaL pattErN with
consTraints). Two search methods including breadth-first and depth-first
searching are designed to troubleshoot the generation of patterns. Furthermore,
several pruning strategies to reduce the reading of sequences and items in the
data and terminate redundant pattern extensions are presented. Finally, we
select a series of datasets with different characteristics and conduct
extensive experiments to compare the TALENT algorithm with the existing
algorithms for mining non-overlapping sequential patterns. The experimental
results demonstrate that the proposed targeted mining algorithm, TALENT, has
excellent mining efficiency and can deal efficiently with many different query
settings.
|
2306.06476 | Abdelhamid Haouhat | Abdelhamid Haouhat, Slimane Bellaouar, Attia Nehar, Hadda Cherroun | Modality Influence in Multimodal Machine Learning | 10 pages | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Multimodal Machine Learning has emerged as a prominent research direction
across various applications such as Sentiment Analysis, Emotion Recognition,
Machine Translation, Hate Speech Recognition, and Movie Genre Classification.
This approach has shown promising results by utilizing modern deep learning
architectures. Despite the achievements made, challenges remain in data
representation, alignment techniques, reasoning, generation, and quantification
within multimodal learning. Additionally, assumptions about the dominant role
of textual modality in decision-making have been made. However, limited
investigations have been conducted on the influence of different modalities in
Multimodal Machine Learning systems. This paper aims to address this gap by
studying the impact of each modality on multimodal learning tasks. The research
focuses on verifying presumptions and gaining insights into the usage of
different modalities. The main contribution of this work is the proposal of a
methodology to determine the effect of each modality on several Multimodal
Machine Learning models and datasets from various tasks. Specifically, the
study examines Multimodal Sentiment Analysis, Multimodal Emotion Recognition,
Multimodal Hate Speech Recognition, and Multimodal Disease Detection. The study
objectives include training SOTA MultiModal Machine Learning models with masked
modalities to evaluate their impact on performance. Furthermore, the research
aims to identify the most influential modality or set of modalities for each
task and draw conclusions for diverse multimodal classification tasks. By
undertaking these investigations, this research contributes to a better
understanding of the role of individual modalities in multi-modal learning and
provides valuable insights for future advancements in this field.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 16:28:52 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Haouhat",
"Abdelhamid",
""
],
[
"Bellaouar",
"Slimane",
""
],
[
"Nehar",
"Attia",
""
],
[
"Cherroun",
"Hadda",
""
]
] | TITLE: Modality Influence in Multimodal Machine Learning
ABSTRACT: Multimodal Machine Learning has emerged as a prominent research direction
across various applications such as Sentiment Analysis, Emotion Recognition,
Machine Translation, Hate Speech Recognition, and Movie Genre Classification.
This approach has shown promising results by utilizing modern deep learning
architectures. Despite the achievements made, challenges remain in data
representation, alignment techniques, reasoning, generation, and quantification
within multimodal learning. Additionally, assumptions about the dominant role
of textual modality in decision-making have been made. However, limited
investigations have been conducted on the influence of different modalities in
Multimodal Machine Learning systems. This paper aims to address this gap by
studying the impact of each modality on multimodal learning tasks. The research
focuses on verifying presumptions and gaining insights into the usage of
different modalities. The main contribution of this work is the proposal of a
methodology to determine the effect of each modality on several Multimodal
Machine Learning models and datasets from various tasks. Specifically, the
study examines Multimodal Sentiment Analysis, Multimodal Emotion Recognition,
Multimodal Hate Speech Recognition, and Multimodal Disease Detection. The study
objectives include training SOTA MultiModal Machine Learning models with masked
modalities to evaluate their impact on performance. Furthermore, the research
aims to identify the most influential modality or set of modalities for each
task and draw conclusions for diverse multimodal classification tasks. By
undertaking these investigations, this research contributes to a better
understanding of the role of individual modalities in multi-modal learning and
provides valuable insights for future advancements in this field.
|
2306.06477 | Raviraj Joshi | Maithili Sabane, Aparna Ranade, Onkar Litake, Parth Patil, Raviraj
Joshi, Dipali Kadam | Enhancing Low Resource NER Using Assisting Language And Transfer
Learning | Accepted at International Conference on Applied Artificial
Intelligence and Computing (ICAAIC) 2023 | null | 10.1109/ICAAIC56838.2023.10141204 | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Named Entity Recognition (NER) is a fundamental task in NLP that is used to
locate the key information in text and is primarily applied in conversational
and search systems. In commercial applications, NER or comparable slot-filling
methods have been widely deployed for popular languages. NER is used in
applications such as human resources, customer service, search engines, content
classification, and academia. In this paper, we draw focus on identifying name
entities for low-resource Indian languages that are closely related, like Hindi
and Marathi. We use various adaptations of BERT such as baseBERT, AlBERT, and
RoBERTa to train a supervised NER model. We also compare multilingual models
with monolingual models and establish a baseline. In this work, we show the
assisting capabilities of the Hindi and Marathi languages for the NER task. We
show that models trained using multiple languages perform better than a single
language. However, we also observe that blind mixing of all datasets doesn't
necessarily provide improvements and data selection methods may be required.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 16:31:04 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Sabane",
"Maithili",
""
],
[
"Ranade",
"Aparna",
""
],
[
"Litake",
"Onkar",
""
],
[
"Patil",
"Parth",
""
],
[
"Joshi",
"Raviraj",
""
],
[
"Kadam",
"Dipali",
""
]
] | TITLE: Enhancing Low Resource NER Using Assisting Language And Transfer
Learning
ABSTRACT: Named Entity Recognition (NER) is a fundamental task in NLP that is used to
locate the key information in text and is primarily applied in conversational
and search systems. In commercial applications, NER or comparable slot-filling
methods have been widely deployed for popular languages. NER is used in
applications such as human resources, customer service, search engines, content
classification, and academia. In this paper, we draw focus on identifying name
entities for low-resource Indian languages that are closely related, like Hindi
and Marathi. We use various adaptations of BERT such as baseBERT, AlBERT, and
RoBERTa to train a supervised NER model. We also compare multilingual models
with monolingual models and establish a baseline. In this work, we show the
assisting capabilities of the Hindi and Marathi languages for the NER task. We
show that models trained using multiple languages perform better than a single
language. However, we also observe that blind mixing of all datasets doesn't
necessarily provide improvements and data selection methods may be required.
|
2306.06480 | Wei Liu | Wei Liu and Michael Strube | Annotation-Inspired Implicit Discourse Relation Classification with
Auxiliary Discourse Connective Generation | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Implicit discourse relation classification is a challenging task due to the
absence of discourse connectives. To overcome this issue, we design an
end-to-end neural model to explicitly generate discourse connectives for the
task, inspired by the annotation process of PDTB. Specifically, our model
jointly learns to generate discourse connectives between arguments and predict
discourse relations based on the arguments and the generated connectives. To
prevent our relation classifier from being misled by poor connectives generated
at the early stage of training while alleviating the discrepancy between
training and inference, we adopt Scheduled Sampling to the joint learning. We
evaluate our method on three benchmarks, PDTB 2.0, PDTB 3.0, and PCC. Results
show that our joint model significantly outperforms various baselines on three
datasets, demonstrating its superiority for the task.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 16:38:46 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Liu",
"Wei",
""
],
[
"Strube",
"Michael",
""
]
] | TITLE: Annotation-Inspired Implicit Discourse Relation Classification with
Auxiliary Discourse Connective Generation
ABSTRACT: Implicit discourse relation classification is a challenging task due to the
absence of discourse connectives. To overcome this issue, we design an
end-to-end neural model to explicitly generate discourse connectives for the
task, inspired by the annotation process of PDTB. Specifically, our model
jointly learns to generate discourse connectives between arguments and predict
discourse relations based on the arguments and the generated connectives. To
prevent our relation classifier from being misled by poor connectives generated
at the early stage of training while alleviating the discrepancy between
training and inference, we adopt Scheduled Sampling to the joint learning. We
evaluate our method on three benchmarks, PDTB 2.0, PDTB 3.0, and PCC. Results
show that our joint model significantly outperforms various baselines on three
datasets, demonstrating its superiority for the task.
|
2306.06508 | Wenxuan Bao | Wenxuan Bao, Haohan Wang, Jun Wu, Jingrui He | Optimizing the Collaboration Structure in Cross-Silo Federated Learning | Accepted by ICML 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In federated learning (FL), multiple clients collaborate to train machine
learning models together while keeping their data decentralized. Through
utilizing more training data, FL suffers from the potential negative transfer
problem: the global FL model may even perform worse than the models trained
with local data only. In this paper, we propose FedCollab, a novel FL framework
that alleviates negative transfer by clustering clients into non-overlapping
coalitions based on their distribution distances and data quantities. As a
result, each client only collaborates with the clients having similar data
distributions, and tends to collaborate with more clients when it has less
data. We evaluate our framework with a variety of datasets, models, and types
of non-IIDness. Our results demonstrate that FedCollab effectively mitigates
negative transfer across a wide range of FL algorithms and consistently
outperforms other clustered FL algorithms.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 18:59:50 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Bao",
"Wenxuan",
""
],
[
"Wang",
"Haohan",
""
],
[
"Wu",
"Jun",
""
],
[
"He",
"Jingrui",
""
]
] | TITLE: Optimizing the Collaboration Structure in Cross-Silo Federated Learning
ABSTRACT: In federated learning (FL), multiple clients collaborate to train machine
learning models together while keeping their data decentralized. Through
utilizing more training data, FL suffers from the potential negative transfer
problem: the global FL model may even perform worse than the models trained
with local data only. In this paper, we propose FedCollab, a novel FL framework
that alleviates negative transfer by clustering clients into non-overlapping
coalitions based on their distribution distances and data quantities. As a
result, each client only collaborates with the clients having similar data
distributions, and tends to collaborate with more clients when it has less
data. We evaluate our framework with a variety of datasets, models, and types
of non-IIDness. Our results demonstrate that FedCollab effectively mitigates
negative transfer across a wide range of FL algorithms and consistently
outperforms other clustered FL algorithms.
|
2306.06522 | Philipp Hallgarten | Philipp Hallgarten, David Bethge, Ozan \"Ozdenizci, Tobias
Grosse-Puppendahl, Enkelejda Kasneci | TS-MoCo: Time-Series Momentum Contrast for Self-Supervised Physiological
Representation Learning | 31st European Signal Processing Conference (EUSIPCO) | null | null | null | cs.LG cs.HC eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Limited availability of labeled physiological data often prohibits the use of
powerful supervised deep learning models in the biomedical machine intelligence
domain. We approach this problem and propose a novel encoding framework that
relies on self-supervised learning with momentum contrast to learn
representations from multivariate time-series of various physiological domains
without needing labels. Our model uses a transformer architecture that can be
easily adapted to classification problems by optimizing a linear output
classification layer. We experimentally evaluate our framework using two
publicly available physiological datasets from different domains, i.e., human
activity recognition from embedded inertial sensory and emotion recognition
from electroencephalography. We show that our self-supervised learning approach
can indeed learn discriminative features which can be exploited in downstream
classification tasks. Our work enables the development of domain-agnostic
intelligent systems that can effectively analyze multivariate time-series data
from physiological domains.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 21:17:42 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Hallgarten",
"Philipp",
""
],
[
"Bethge",
"David",
""
],
[
"Özdenizci",
"Ozan",
""
],
[
"Grosse-Puppendahl",
"Tobias",
""
],
[
"Kasneci",
"Enkelejda",
""
]
] | TITLE: TS-MoCo: Time-Series Momentum Contrast for Self-Supervised Physiological
Representation Learning
ABSTRACT: Limited availability of labeled physiological data often prohibits the use of
powerful supervised deep learning models in the biomedical machine intelligence
domain. We approach this problem and propose a novel encoding framework that
relies on self-supervised learning with momentum contrast to learn
representations from multivariate time-series of various physiological domains
without needing labels. Our model uses a transformer architecture that can be
easily adapted to classification problems by optimizing a linear output
classification layer. We experimentally evaluate our framework using two
publicly available physiological datasets from different domains, i.e., human
activity recognition from embedded inertial sensory and emotion recognition
from electroencephalography. We show that our self-supervised learning approach
can indeed learn discriminative features which can be exploited in downstream
classification tasks. Our work enables the development of domain-agnostic
intelligent systems that can effectively analyze multivariate time-series data
from physiological domains.
|
2306.06527 | Amine Bendahmane | Amine Bendahmane | Contribution \`a l'Optimisation d'un Comportement Collectif pour un
Groupe de Robots Autonomes | PhD Thesis (French version) | null | 10.13140/RG.2.2.34066.63684 | null | cs.RO cs.AI cs.MA | http://creativecommons.org/licenses/by/4.0/ | This thesis studies the domain of collective robotics, and more particularly
the optimization problems of multirobot systems in the context of exploration,
path planning and coordination. It includes two contributions. The first one is
the use of the Butterfly Optimization Algorithm (BOA) to solve the Unknown Area
Exploration problem with energy constraints in dynamic environments. This
algorithm was never used for solving robotics problems before, as far as we
know. We proposed a new version of this algorithm called xBOA based on the
crossover operator to improve the diversity of the candidate solutions and
speed up the convergence of the algorithm. The second contribution is the
development of a new simulation framework for benchmarking dynamic incremental
problems in robotics such as exploration tasks. The framework is made in such a
manner to be generic to quickly compare different metaheuristics with minimum
modifications, and to adapt easily to single and multi-robot scenarios. Also,
it provides researchers with tools to automate their experiments and generate
visuals, which will allow them to focus on more important tasks such as
modeling new algorithms. We conducted a series of experiments that showed
promising results and allowed us to validate our approach and model.
| [
{
"version": "v1",
"created": "Sat, 10 Jun 2023 21:49:08 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Bendahmane",
"Amine",
""
]
] | TITLE: Contribution \`a l'Optimisation d'un Comportement Collectif pour un
Groupe de Robots Autonomes
ABSTRACT: This thesis studies the domain of collective robotics, and more particularly
the optimization problems of multirobot systems in the context of exploration,
path planning and coordination. It includes two contributions. The first one is
the use of the Butterfly Optimization Algorithm (BOA) to solve the Unknown Area
Exploration problem with energy constraints in dynamic environments. This
algorithm was never used for solving robotics problems before, as far as we
know. We proposed a new version of this algorithm called xBOA based on the
crossover operator to improve the diversity of the candidate solutions and
speed up the convergence of the algorithm. The second contribution is the
development of a new simulation framework for benchmarking dynamic incremental
problems in robotics such as exploration tasks. The framework is made in such a
manner to be generic to quickly compare different metaheuristics with minimum
modifications, and to adapt easily to single and multi-robot scenarios. Also,
it provides researchers with tools to automate their experiments and generate
visuals, which will allow them to focus on more important tasks such as
modeling new algorithms. We conducted a series of experiments that showed
promising results and allowed us to validate our approach and model.
|
2306.06578 | Weizhe Chen | Weizhe Chen and Lantao Liu | Long-Term Autonomous Ocean Monitoring with Streaming Samples | Proceedings of OCEANS 2019, SEATTLE | null | null | null | cs.RO | http://creativecommons.org/licenses/by/4.0/ | In the autonomous ocean monitoring task, the sampling robot moves in the
environment and accumulates data continuously. The widely adopted spatial
modeling method - standard Gaussian process (GP) regression - becomes
inadequate in processing the growing sensing data of a large size. To overcome
the computational challenge, this paper presents an environmental modeling
framework using a sparse variant of GP called streaming sparse GP (SSGP). The
SSGP is able to handle streaming data in an online and incremental manner, and
is therefore suitable for long-term autonomous environmental monitoring. The
SSGP summarizes the collected data using a small set of pseudo data points that
best represent the whole dataset, and updates the hyperparameters and pseudo
point locations in a streaming fashion, leading to high-quality approximation
of the underlying environmental model with significantly reduced computational
cost and memory demand.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 03:59:26 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Chen",
"Weizhe",
""
],
[
"Liu",
"Lantao",
""
]
] | TITLE: Long-Term Autonomous Ocean Monitoring with Streaming Samples
ABSTRACT: In the autonomous ocean monitoring task, the sampling robot moves in the
environment and accumulates data continuously. The widely adopted spatial
modeling method - standard Gaussian process (GP) regression - becomes
inadequate in processing the growing sensing data of a large size. To overcome
the computational challenge, this paper presents an environmental modeling
framework using a sparse variant of GP called streaming sparse GP (SSGP). The
SSGP is able to handle streaming data in an online and incremental manner, and
is therefore suitable for long-term autonomous environmental monitoring. The
SSGP summarizes the collected data using a small set of pseudo data points that
best represent the whole dataset, and updates the hyperparameters and pseudo
point locations in a streaming fashion, leading to high-quality approximation
of the underlying environmental model with significantly reduced computational
cost and memory demand.
|
2306.06580 | Ahmed BaHammam | Ahmed S. BaHammam | Unlocking the Power of Health Datasets and Registries: The Need for
Urgent Institutional and National Ownership and Governance Regulations for
Research Advancement | 1 Figure, (in press) | Journal of Nature and Science of Medicine 2023, Volume 6, Issue 3 | null | null | cs.CY q-bio.OT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Health datasets have immense potential to drive research advancements and
improve healthcare outcomes. However, realizing this potential requires careful
consideration of governance and ownership frameworks. This article explores the
importance of nurturing governance and ownership models that facilitate
responsible and ethical use of health datasets for research purposes. We
highlight the importance of adopting governance and ownership models that
enable responsible and ethical utilization of health datasets and clinical data
registries for research purposes. The article addresses the important local and
international regulations related to the utilization of health data/medical
records in research, and emphasizes the urgent need for developing clear
institutional and national guidelines on data access, sharing, and utilization,
ensuring transparency, privacy, and data protection. By establishing robust
governance structures and fostering ownership among stakeholders,
collaboration, innovation, and equitable access to health data can be promoted,
ultimately unlocking its full power for transformative research and improving
global health outcomes.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 04:01:42 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"BaHammam",
"Ahmed S.",
""
]
] | TITLE: Unlocking the Power of Health Datasets and Registries: The Need for
Urgent Institutional and National Ownership and Governance Regulations for
Research Advancement
ABSTRACT: Health datasets have immense potential to drive research advancements and
improve healthcare outcomes. However, realizing this potential requires careful
consideration of governance and ownership frameworks. This article explores the
importance of nurturing governance and ownership models that facilitate
responsible and ethical use of health datasets for research purposes. We
highlight the importance of adopting governance and ownership models that
enable responsible and ethical utilization of health datasets and clinical data
registries for research purposes. The article addresses the important local and
international regulations related to the utilization of health data/medical
records in research, and emphasizes the urgent need for developing clear
institutional and national guidelines on data access, sharing, and utilization,
ensuring transparency, privacy, and data protection. By establishing robust
governance structures and fostering ownership among stakeholders,
collaboration, innovation, and equitable access to health data can be promoted,
ultimately unlocking its full power for transformative research and improving
global health outcomes.
|
2306.06582 | Yue Gao | Yue Gao, Garvesh Raskutti, Rebecca Willet | Fast, Distribution-free Predictive Inference for Neural Networks with
Coverage Guarantees | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper introduces a novel, computationally-efficient algorithm for
predictive inference (PI) that requires no distributional assumptions on the
data and can be computed faster than existing bootstrap-type methods for neural
networks. Specifically, if there are $n$ training samples, bootstrap methods
require training a model on each of the $n$ subsamples of size $n-1$; for large
models like neural networks, this process can be computationally prohibitive.
In contrast, our proposed method trains one neural network on the full dataset
with $(\epsilon, \delta)$-differential privacy (DP) and then approximates each
leave-one-out model efficiently using a linear approximation around the
differentially-private neural network estimate. With exchangeable data, we
prove that our approach has a rigorous coverage guarantee that depends on the
preset privacy parameters and the stability of the neural network, regardless
of the data distribution. Simulations and experiments on real data demonstrate
that our method satisfies the coverage guarantees with substantially reduced
computation compared to bootstrap methods.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 04:03:58 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Gao",
"Yue",
""
],
[
"Raskutti",
"Garvesh",
""
],
[
"Willet",
"Rebecca",
""
]
] | TITLE: Fast, Distribution-free Predictive Inference for Neural Networks with
Coverage Guarantees
ABSTRACT: This paper introduces a novel, computationally-efficient algorithm for
predictive inference (PI) that requires no distributional assumptions on the
data and can be computed faster than existing bootstrap-type methods for neural
networks. Specifically, if there are $n$ training samples, bootstrap methods
require training a model on each of the $n$ subsamples of size $n-1$; for large
models like neural networks, this process can be computationally prohibitive.
In contrast, our proposed method trains one neural network on the full dataset
with $(\epsilon, \delta)$-differential privacy (DP) and then approximates each
leave-one-out model efficiently using a linear approximation around the
differentially-private neural network estimate. With exchangeable data, we
prove that our approach has a rigorous coverage guarantee that depends on the
preset privacy parameters and the stability of the neural network, regardless
of the data distribution. Simulations and experiments on real data demonstrate
that our method satisfies the coverage guarantees with substantially reduced
computation compared to bootstrap methods.
|
2306.06583 | Siyang Song | Siyang Song, Micol Spitale, Cheng Luo, German Barquero, Cristina
Palmero, Sergio Escalera, Michel Valstar, Tobias Baur, Fabien Ringeval,
Elisabeth Andre and Hatice Gunes | REACT2023: the first Multi-modal Multiple Appropriate Facial Reaction
Generation Challenge | null | null | null | null | cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | The Multi-modal Multiple Appropriate Facial Reaction Generation Challenge
(REACT2023) is the first competition event focused on evaluating multimedia
processing and machine learning techniques for generating human-appropriate
facial reactions in various dyadic interaction scenarios, with all participants
competing strictly under the same conditions. The goal of the challenge is to
provide the first benchmark test set for multi-modal information processing and
to foster collaboration among the audio, visual, and audio-visual affective
computing communities, to compare the relative merits of the approaches to
automatic appropriate facial reaction generation under different spontaneous
dyadic interaction conditions. This paper presents: (i) novelties,
contributions and guidelines of the REACT2023 challenge; (ii) the dataset
utilized in the challenge; and (iii) the performance of baseline systems on the
two proposed sub-challenges: Offline Multiple Appropriate Facial Reaction
Generation and Online Multiple Appropriate Facial Reaction Generation,
respectively. The challenge baseline code is publicly available at
\url{https://github.com/reactmultimodalchallenge/baseline_react2023}.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 04:15:56 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Song",
"Siyang",
""
],
[
"Spitale",
"Micol",
""
],
[
"Luo",
"Cheng",
""
],
[
"Barquero",
"German",
""
],
[
"Palmero",
"Cristina",
""
],
[
"Escalera",
"Sergio",
""
],
[
"Valstar",
"Michel",
""
],
[
"Baur",
"Tobias",
""
],
[
"Ringeval",
"Fabien",
""
],
[
"Andre",
"Elisabeth",
""
],
[
"Gunes",
"Hatice",
""
]
] | TITLE: REACT2023: the first Multi-modal Multiple Appropriate Facial Reaction
Generation Challenge
ABSTRACT: The Multi-modal Multiple Appropriate Facial Reaction Generation Challenge
(REACT2023) is the first competition event focused on evaluating multimedia
processing and machine learning techniques for generating human-appropriate
facial reactions in various dyadic interaction scenarios, with all participants
competing strictly under the same conditions. The goal of the challenge is to
provide the first benchmark test set for multi-modal information processing and
to foster collaboration among the audio, visual, and audio-visual affective
computing communities, to compare the relative merits of the approaches to
automatic appropriate facial reaction generation under different spontaneous
dyadic interaction conditions. This paper presents: (i) novelties,
contributions and guidelines of the REACT2023 challenge; (ii) the dataset
utilized in the challenge; and (iii) the performance of baseline systems on the
two proposed sub-challenges: Offline Multiple Appropriate Facial Reaction
Generation and Online Multiple Appropriate Facial Reaction Generation,
respectively. The challenge baseline code is publicly available at
\url{https://github.com/reactmultimodalchallenge/baseline_react2023}.
|
2306.06584 | Qiang Lyu | Qiang Lyu, Weiqiang Wang | Compositional Prototypical Networks for Few-Shot Classification | Accepted by AAAI 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is assumed that pre-training provides the feature extractor with strong
class transferability and that high novel class generalization can be achieved
by simply reusing the transferable feature extractor. In this work, our
motivation is to explicitly learn some fine-grained and transferable
meta-knowledge so that feature reusability can be further improved. Concretely,
inspired by the fact that humans can use learned concepts or components to help
them recognize novel classes, we propose Compositional Prototypical Networks
(CPN) to learn a transferable prototype for each human-annotated attribute,
which we call a component prototype. We empirically demonstrate that the
learned component prototypes have good class transferability and can be reused
to construct compositional prototypes for novel classes. Then a learnable
weight generator is utilized to adaptively fuse the compositional and visual
prototypes. Extensive experiments demonstrate that our method can achieve
state-of-the-art results on different datasets and settings. The performance
gains are especially remarkable in the 5-way 1-shot setting. The code is
available at https://github.com/fikry102/CPN.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 04:16:12 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Lyu",
"Qiang",
""
],
[
"Wang",
"Weiqiang",
""
]
] | TITLE: Compositional Prototypical Networks for Few-Shot Classification
ABSTRACT: It is assumed that pre-training provides the feature extractor with strong
class transferability and that high novel class generalization can be achieved
by simply reusing the transferable feature extractor. In this work, our
motivation is to explicitly learn some fine-grained and transferable
meta-knowledge so that feature reusability can be further improved. Concretely,
inspired by the fact that humans can use learned concepts or components to help
them recognize novel classes, we propose Compositional Prototypical Networks
(CPN) to learn a transferable prototype for each human-annotated attribute,
which we call a component prototype. We empirically demonstrate that the
learned component prototypes have good class transferability and can be reused
to construct compositional prototypes for novel classes. Then a learnable
weight generator is utilized to adaptively fuse the compositional and visual
prototypes. Extensive experiments demonstrate that our method can achieve
state-of-the-art results on different datasets and settings. The performance
gains are especially remarkable in the 5-way 1-shot setting. The code is
available at https://github.com/fikry102/CPN.
|
2306.06601 | Ting Zhang | Ting Zhang, Zhuang Chen, Ming Zhong and Tieyun Qian | Mimicking the Thinking Process for Emotion Recognition in Conversation
with Prompts and Paraphrasing | Accepted to IJCAI 2023, AI and Social Good track | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Emotion recognition in conversation, which aims to predict the emotion for
all utterances, has attracted considerable research attention in recent years.
It is a challenging task since the recognition of the emotion in one utterance
involves many complex factors, such as the conversational context, the
speaker's background, and the subtle difference between emotion labels. In this
paper, we propose a novel framework which mimics the thinking process when
modeling these factors. Specifically, we first comprehend the conversational
context with a history-oriented prompt to selectively gather information from
predecessors of the target utterance. We then model the speaker's background
with an experience-oriented prompt to retrieve the similar utterances from all
conversations. We finally differentiate the subtle label semantics with a
paraphrasing mechanism to elicit the intrinsic label related knowledge. We
conducted extensive experiments on three benchmarks. The empirical results
demonstrate the superiority of our proposed framework over the state-of-the-art
baselines.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 06:36:19 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhang",
"Ting",
""
],
[
"Chen",
"Zhuang",
""
],
[
"Zhong",
"Ming",
""
],
[
"Qian",
"Tieyun",
""
]
] | TITLE: Mimicking the Thinking Process for Emotion Recognition in Conversation
with Prompts and Paraphrasing
ABSTRACT: Emotion recognition in conversation, which aims to predict the emotion for
all utterances, has attracted considerable research attention in recent years.
It is a challenging task since the recognition of the emotion in one utterance
involves many complex factors, such as the conversational context, the
speaker's background, and the subtle difference between emotion labels. In this
paper, we propose a novel framework which mimics the thinking process when
modeling these factors. Specifically, we first comprehend the conversational
context with a history-oriented prompt to selectively gather information from
predecessors of the target utterance. We then model the speaker's background
with an experience-oriented prompt to retrieve the similar utterances from all
conversations. We finally differentiate the subtle label semantics with a
paraphrasing mechanism to elicit the intrinsic label related knowledge. We
conducted extensive experiments on three benchmarks. The empirical results
demonstrate the superiority of our proposed framework over the state-of-the-art
baselines.
|
2306.06620 | Son Nguyen | Son Nguyen, Cuong Tran Manh, Kien T. Tran, Tan M. Nguyen, Thu-Trang
Nguyen, Kien-Tuan Ngo and Hieu Dinh Vo | ARIST: An Effective API Argument Recommendation Approach | null | null | null | null | cs.SE cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Learning and remembering to use APIs are difficult. Several techniques have
been proposed to assist developers in using APIs. Most existing techniques
focus on recommending the right API methods to call, but very few techniques
focus on recommending API arguments. In this paper, we propose ARIST, a novel
automated argument recommendation approach which suggests arguments by
predicting developers' expectations when they define and use API methods. To
implement this idea in the recommendation process, ARIST combines program
analysis (PA), language models (LMs), and several features specialized for the
recommendation task which consider the functionality of formal parameters and
the positional information of code elements (e.g., variables or method calls)
in the given context. In ARIST, the LMs and the recommending features are used
to suggest the promising candidates identified by PA. Meanwhile, PA navigates
the LMs and the features working on the set of the valid candidates which
satisfy syntax, accessibility, and type-compatibility constraints defined by
the programming language in use. Our evaluation on a large dataset of
real-world projects shows that ARIST improves the state-of-the-art approach by
19% and 18% in top-1 precision and recall for recommending arguments of
frequently-used libraries. For general argument recommendation task, i.e.,
recommending arguments for every method call, ARIST outperforms the baseline
approaches by up to 125% top-1 accuracy. Moreover, for newly-encountered
projects, ARIST achieves more than 60% top-3 accuracy when evaluating on a
larger dataset. For working/maintaining projects, with a personalized LM to
capture developers' coding practice, ARIST can productively rank the expected
arguments at the top-1 position in 7/10 requests.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 08:41:37 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Nguyen",
"Son",
""
],
[
"Manh",
"Cuong Tran",
""
],
[
"Tran",
"Kien T.",
""
],
[
"Nguyen",
"Tan M.",
""
],
[
"Nguyen",
"Thu-Trang",
""
],
[
"Ngo",
"Kien-Tuan",
""
],
[
"Vo",
"Hieu Dinh",
""
]
] | TITLE: ARIST: An Effective API Argument Recommendation Approach
ABSTRACT: Learning and remembering to use APIs are difficult. Several techniques have
been proposed to assist developers in using APIs. Most existing techniques
focus on recommending the right API methods to call, but very few techniques
focus on recommending API arguments. In this paper, we propose ARIST, a novel
automated argument recommendation approach which suggests arguments by
predicting developers' expectations when they define and use API methods. To
implement this idea in the recommendation process, ARIST combines program
analysis (PA), language models (LMs), and several features specialized for the
recommendation task which consider the functionality of formal parameters and
the positional information of code elements (e.g., variables or method calls)
in the given context. In ARIST, the LMs and the recommending features are used
to suggest the promising candidates identified by PA. Meanwhile, PA navigates
the LMs and the features working on the set of the valid candidates which
satisfy syntax, accessibility, and type-compatibility constraints defined by
the programming language in use. Our evaluation on a large dataset of
real-world projects shows that ARIST improves the state-of-the-art approach by
19% and 18% in top-1 precision and recall for recommending arguments of
frequently-used libraries. For general argument recommendation task, i.e.,
recommending arguments for every method call, ARIST outperforms the baseline
approaches by up to 125% top-1 accuracy. Moreover, for newly-encountered
projects, ARIST achieves more than 60% top-3 accuracy when evaluating on a
larger dataset. For working/maintaining projects, with a personalized LM to
capture developers' coding practice, ARIST can productively rank the expected
arguments at the top-1 position in 7/10 requests.
|
2306.06625 | Shicheng Tan | Shicheng Tan, Weng Lam Tam, Yuanchun Wang, Wenwen Gong, Shu Zhao, Peng
Zhang, Jie Tang | Are Intermediate Layers and Labels Really Necessary? A General Language
Model Distillation Method | Accepted to Findings of ACL2023 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The large scale of pre-trained language models poses a challenge for their
deployment on various devices, with a growing emphasis on methods to compress
these models, particularly knowledge distillation. However, current knowledge
distillation methods rely on the model's intermediate layer features and the
golden labels (also called hard labels), which usually require aligned model
architecture and enough labeled data respectively. Moreover, the parameters of
vocabulary are usually neglected in existing methods. To address these
problems, we propose a general language model distillation (GLMD) method that
performs two-stage word prediction distillation and vocabulary compression,
which is simple and surprisingly shows extremely strong performance.
Specifically, GLMD supports more general application scenarios by eliminating
the constraints of dimension and structure between models and the need for
labeled datasets through the absence of intermediate layers and golden labels.
Meanwhile, based on the long-tailed distribution of word frequencies in the
data, GLMD designs a strategy of vocabulary compression through decreasing
vocabulary size instead of dimensionality. Experimental results show that our
method outperforms 25 state-of-the-art methods on the SuperGLUE benchmark,
achieving an average score that surpasses the best method by 3%.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 08:53:27 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Tan",
"Shicheng",
""
],
[
"Tam",
"Weng Lam",
""
],
[
"Wang",
"Yuanchun",
""
],
[
"Gong",
"Wenwen",
""
],
[
"Zhao",
"Shu",
""
],
[
"Zhang",
"Peng",
""
],
[
"Tang",
"Jie",
""
]
] | TITLE: Are Intermediate Layers and Labels Really Necessary? A General Language
Model Distillation Method
ABSTRACT: The large scale of pre-trained language models poses a challenge for their
deployment on various devices, with a growing emphasis on methods to compress
these models, particularly knowledge distillation. However, current knowledge
distillation methods rely on the model's intermediate layer features and the
golden labels (also called hard labels), which usually require aligned model
architecture and enough labeled data respectively. Moreover, the parameters of
vocabulary are usually neglected in existing methods. To address these
problems, we propose a general language model distillation (GLMD) method that
performs two-stage word prediction distillation and vocabulary compression,
which is simple and surprisingly shows extremely strong performance.
Specifically, GLMD supports more general application scenarios by eliminating
the constraints of dimension and structure between models and the need for
labeled datasets through the absence of intermediate layers and golden labels.
Meanwhile, based on the long-tailed distribution of word frequencies in the
data, GLMD designs a strategy of vocabulary compression through decreasing
vocabulary size instead of dimensionality. Experimental results show that our
method outperforms 25 state-of-the-art methods on the SuperGLUE benchmark,
achieving an average score that surpasses the best method by 3%.
|
2306.06626 | Neta Shaul | Neta Shaul, Ricky T. Q. Chen, Maximilian Nickel, Matt Le, Yaron Lipman | On Kinetic Optimal Probability Paths for Generative Models | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent successful generative models are trained by fitting a neural network
to an a-priori defined tractable probability density path taking noise to
training examples. In this paper we investigate the space of Gaussian
probability paths, which includes diffusion paths as an instance, and look for
an optimal member in some useful sense. In particular, minimizing the Kinetic
Energy (KE) of a path is known to make particles' trajectories simple, hence
easier to sample, and empirically improve performance in terms of likelihood of
unseen data and sample generation quality. We investigate Kinetic Optimal (KO)
Gaussian paths and offer the following observations: (i) We show the KE takes a
simplified form on the space of Gaussian paths, where the data is incorporated
only through a single, one dimensional scalar function, called the \emph{data
separation function}. (ii) We characterize the KO solutions with a one
dimensional ODE. (iii) We approximate data-dependent KO paths by approximating
the data separation function and minimizing the KE. (iv) We prove that the data
separation function converges to $1$ in the general case of arbitrary
normalized dataset consisting of $n$ samples in $d$ dimension as
$n/\sqrt{d}\rightarrow 0$. A consequence of this result is that the Conditional
Optimal Transport (Cond-OT) path becomes \emph{kinetic optimal} as
$n/\sqrt{d}\rightarrow 0$. We further support this theory with empirical
experiments on ImageNet.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 08:54:12 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shaul",
"Neta",
""
],
[
"Chen",
"Ricky T. Q.",
""
],
[
"Nickel",
"Maximilian",
""
],
[
"Le",
"Matt",
""
],
[
"Lipman",
"Yaron",
""
]
] | TITLE: On Kinetic Optimal Probability Paths for Generative Models
ABSTRACT: Recent successful generative models are trained by fitting a neural network
to an a-priori defined tractable probability density path taking noise to
training examples. In this paper we investigate the space of Gaussian
probability paths, which includes diffusion paths as an instance, and look for
an optimal member in some useful sense. In particular, minimizing the Kinetic
Energy (KE) of a path is known to make particles' trajectories simple, hence
easier to sample, and empirically improve performance in terms of likelihood of
unseen data and sample generation quality. We investigate Kinetic Optimal (KO)
Gaussian paths and offer the following observations: (i) We show the KE takes a
simplified form on the space of Gaussian paths, where the data is incorporated
only through a single, one dimensional scalar function, called the \emph{data
separation function}. (ii) We characterize the KO solutions with a one
dimensional ODE. (iii) We approximate data-dependent KO paths by approximating
the data separation function and minimizing the KE. (iv) We prove that the data
separation function converges to $1$ in the general case of arbitrary
normalized dataset consisting of $n$ samples in $d$ dimension as
$n/\sqrt{d}\rightarrow 0$. A consequence of this result is that the Conditional
Optimal Transport (Cond-OT) path becomes \emph{kinetic optimal} as
$n/\sqrt{d}\rightarrow 0$. We further support this theory with empirical
experiments on ImageNet.
|
2306.06634 | Defang Chen | Hailin Zhang, Defang Chen, Can Wang | Adaptive Multi-Teacher Knowledge Distillation with Meta-Learning | ICME 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-Teacher knowledge distillation provides students with additional
supervision from multiple pre-trained teachers with diverse information
sources. Most existing methods explore different weighting strategies to obtain
a powerful ensemble teacher, while ignoring the student with poor learning
ability may not benefit from such specialized integrated knowledge. To address
this problem, we propose Adaptive Multi-teacher Knowledge Distillation with
Meta-Learning (MMKD) to supervise student with appropriate knowledge from a
tailored ensemble teacher. With the help of a meta-weight network, the diverse
yet compatible teacher knowledge in the output layer and intermediate layers is
jointly leveraged to enhance the student performance. Extensive experiments on
multiple benchmark datasets validate the effectiveness and flexibility of our
methods. Code is available: https://github.com/Rorozhl/MMKD.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 09:38:45 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhang",
"Hailin",
""
],
[
"Chen",
"Defang",
""
],
[
"Wang",
"Can",
""
]
] | TITLE: Adaptive Multi-Teacher Knowledge Distillation with Meta-Learning
ABSTRACT: Multi-Teacher knowledge distillation provides students with additional
supervision from multiple pre-trained teachers with diverse information
sources. Most existing methods explore different weighting strategies to obtain
a powerful ensemble teacher, while ignoring the student with poor learning
ability may not benefit from such specialized integrated knowledge. To address
this problem, we propose Adaptive Multi-teacher Knowledge Distillation with
Meta-Learning (MMKD) to supervise student with appropriate knowledge from a
tailored ensemble teacher. With the help of a meta-weight network, the diverse
yet compatible teacher knowledge in the output layer and intermediate layers is
jointly leveraged to enhance the student performance. Extensive experiments on
multiple benchmark datasets validate the effectiveness and flexibility of our
methods. Code is available: https://github.com/Rorozhl/MMKD.
|
2306.06635 | Itamar Zimerman | Ethan Baron, Itamar Zimerman, Lior Wolf | 2-D SSM: A General Spatial Layer for Visual Transformers | 16 pages, 5 figures | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A central objective in computer vision is to design models with appropriate
2-D inductive bias. Desiderata for 2D inductive bias include two-dimensional
position awareness, dynamic spatial locality, and translation and permutation
invariance. To address these goals, we leverage an expressive variation of the
multidimensional State Space Model (SSM). Our approach introduces efficient
parameterization, accelerated computation, and a suitable normalization scheme.
Empirically, we observe that incorporating our layer at the beginning of each
transformer block of Vision Transformers (ViT) significantly enhances
performance for multiple ViT backbones and across datasets. The new layer is
effective even with a negligible amount of additional parameters and inference
time. Ablation studies and visualizations demonstrate that the layer has a
strong 2-D inductive bias. For example, vision transformers equipped with our
layer exhibit effective performance even without positional encoding
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 09:41:37 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Baron",
"Ethan",
""
],
[
"Zimerman",
"Itamar",
""
],
[
"Wolf",
"Lior",
""
]
] | TITLE: 2-D SSM: A General Spatial Layer for Visual Transformers
ABSTRACT: A central objective in computer vision is to design models with appropriate
2-D inductive bias. Desiderata for 2D inductive bias include two-dimensional
position awareness, dynamic spatial locality, and translation and permutation
invariance. To address these goals, we leverage an expressive variation of the
multidimensional State Space Model (SSM). Our approach introduces efficient
parameterization, accelerated computation, and a suitable normalization scheme.
Empirically, we observe that incorporating our layer at the beginning of each
transformer block of Vision Transformers (ViT) significantly enhances
performance for multiple ViT backbones and across datasets. The new layer is
effective even with a negligible amount of additional parameters and inference
time. Ablation studies and visualizations demonstrate that the layer has a
strong 2-D inductive bias. For example, vision transformers equipped with our
layer exhibit effective performance even without positional encoding
|
2306.06638 | Dani Valevski | Dani Valevski, Danny Wasserman, Yossi Matias, Yaniv Leviathan | Face0: Instantaneously Conditioning a Text-to-Image Model on a Face | null | null | null | null | cs.CV cs.AI cs.GR cs.LG | http://creativecommons.org/licenses/by/4.0/ | We present Face0, a novel way to instantaneously condition a text-to-image
generation model on a face, in sample time, without any optimization procedures
such as fine-tuning or inversions. We augment a dataset of annotated images
with embeddings of the included faces and train an image generation model, on
the augmented dataset. Once trained, our system is practically identical at
inference time to the underlying base model, and is therefore able to generate
images, given a user-supplied face image and a prompt, in just a couple of
seconds. Our method achieves pleasing results, is remarkably simple, extremely
fast, and equips the underlying model with new capabilities, like controlling
the generated images both via text or via direct manipulation of the input face
embeddings. In addition, when using a fixed random vector instead of a face
embedding from a user supplied image, our method essentially solves the problem
of consistent character generation across images. Finally, while requiring
further research, we hope that our method, which decouples the model's textual
biases from its biases on faces, might be a step towards some mitigation of
biases in future text-to-image models.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 09:52:03 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Valevski",
"Dani",
""
],
[
"Wasserman",
"Danny",
""
],
[
"Matias",
"Yossi",
""
],
[
"Leviathan",
"Yaniv",
""
]
] | TITLE: Face0: Instantaneously Conditioning a Text-to-Image Model on a Face
ABSTRACT: We present Face0, a novel way to instantaneously condition a text-to-image
generation model on a face, in sample time, without any optimization procedures
such as fine-tuning or inversions. We augment a dataset of annotated images
with embeddings of the included faces and train an image generation model, on
the augmented dataset. Once trained, our system is practically identical at
inference time to the underlying base model, and is therefore able to generate
images, given a user-supplied face image and a prompt, in just a couple of
seconds. Our method achieves pleasing results, is remarkably simple, extremely
fast, and equips the underlying model with new capabilities, like controlling
the generated images both via text or via direct manipulation of the input face
embeddings. In addition, when using a fixed random vector instead of a face
embedding from a user supplied image, our method essentially solves the problem
of consistent character generation across images. Finally, while requiring
further research, we hope that our method, which decouples the model's textual
biases from its biases on faces, might be a step towards some mitigation of
biases in future text-to-image models.
|
2306.06649 | Kartheek Bondugula | Kartheek Bondugula and Santiago Mazuelas and Aritz P\'erez | Efficient Learning of Minimax Risk Classifiers in High Dimensions | Accepted for the 39th Conference on Uncertainty in Artificial
Intelligence (UAI 2023) | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | High-dimensional data is common in multiple areas, such as health care and
genomics, where the number of features can be tens of thousands. In such
scenarios, the large number of features often leads to inefficient learning.
Constraint generation methods have recently enabled efficient learning of
L1-regularized support vector machines (SVMs). In this paper, we leverage such
methods to obtain an efficient learning algorithm for the recently proposed
minimax risk classifiers (MRCs). The proposed iterative algorithm also provides
a sequence of worst-case error probabilities and performs feature selection.
Experiments on multiple high-dimensional datasets show that the proposed
algorithm is efficient in high-dimensional scenarios. In addition, the
worst-case error probability provides useful information about the classifier
performance, and the features selected by the algorithm are competitive with
the state-of-the-art.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 11:08:20 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Bondugula",
"Kartheek",
""
],
[
"Mazuelas",
"Santiago",
""
],
[
"Pérez",
"Aritz",
""
]
] | TITLE: Efficient Learning of Minimax Risk Classifiers in High Dimensions
ABSTRACT: High-dimensional data is common in multiple areas, such as health care and
genomics, where the number of features can be tens of thousands. In such
scenarios, the large number of features often leads to inefficient learning.
Constraint generation methods have recently enabled efficient learning of
L1-regularized support vector machines (SVMs). In this paper, we leverage such
methods to obtain an efficient learning algorithm for the recently proposed
minimax risk classifiers (MRCs). The proposed iterative algorithm also provides
a sequence of worst-case error probabilities and performs feature selection.
Experiments on multiple high-dimensional datasets show that the proposed
algorithm is efficient in high-dimensional scenarios. In addition, the
worst-case error probability provides useful information about the classifier
performance, and the features selected by the algorithm are competitive with
the state-of-the-art.
|
2306.06669 | Shan Huang | Shan Huang, Xiaohong Liu, Tao Tan, Menghan Hu, Xiaoer Wei, Tingli
Chen, Bin Sheng | TransMRSR: Transformer-based Self-Distilled Generative Prior for Brain
MRI Super-Resolution | 2023 CGI | null | null | null | eess.IV cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Magnetic resonance images (MRI) acquired with low through-plane resolution
compromise time and cost. The poor resolution in one orientation is
insufficient to meet the requirement of high resolution for early diagnosis of
brain disease and morphometric study. The common Single image super-resolution
(SISR) solutions face two main challenges: (1) local detailed and global
anatomical structural information combination; and (2) large-scale restoration
when applied for reconstructing thick-slice MRI into high-resolution (HR)
iso-tropic data. To address these problems, we propose a novel two-stage
network for brain MRI SR named TransMRSR based on the convolutional blocks to
extract local information and transformer blocks to capture long-range
dependencies. TransMRSR consists of three modules: the shallow local feature
extraction, the deep non-local feature capture, and the HR image
reconstruction. We perform a generative task to encapsulate diverse priors into
a generative network (GAN), which is the decoder sub-module of the deep
non-local feature capture part, in the first stage. The pre-trained GAN is used
for the second stage of SR task. We further eliminate the potential latent
space shift caused by the two-stage training strategy through the
self-distilled truncation trick. The extensive experiments show that our method
achieves superior performance to other SSIR methods on both public and private
datasets. Code is released at https://github.com/goddesshs/TransMRSR.git .
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 12:41:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Huang",
"Shan",
""
],
[
"Liu",
"Xiaohong",
""
],
[
"Tan",
"Tao",
""
],
[
"Hu",
"Menghan",
""
],
[
"Wei",
"Xiaoer",
""
],
[
"Chen",
"Tingli",
""
],
[
"Sheng",
"Bin",
""
]
] | TITLE: TransMRSR: Transformer-based Self-Distilled Generative Prior for Brain
MRI Super-Resolution
ABSTRACT: Magnetic resonance images (MRI) acquired with low through-plane resolution
compromise time and cost. The poor resolution in one orientation is
insufficient to meet the requirement of high resolution for early diagnosis of
brain disease and morphometric study. The common Single image super-resolution
(SISR) solutions face two main challenges: (1) local detailed and global
anatomical structural information combination; and (2) large-scale restoration
when applied for reconstructing thick-slice MRI into high-resolution (HR)
iso-tropic data. To address these problems, we propose a novel two-stage
network for brain MRI SR named TransMRSR based on the convolutional blocks to
extract local information and transformer blocks to capture long-range
dependencies. TransMRSR consists of three modules: the shallow local feature
extraction, the deep non-local feature capture, and the HR image
reconstruction. We perform a generative task to encapsulate diverse priors into
a generative network (GAN), which is the decoder sub-module of the deep
non-local feature capture part, in the first stage. The pre-trained GAN is used
for the second stage of SR task. We further eliminate the potential latent
space shift caused by the two-stage training strategy through the
self-distilled truncation trick. The extensive experiments show that our method
achieves superior performance to other SSIR methods on both public and private
datasets. Code is released at https://github.com/goddesshs/TransMRSR.git .
|
2306.06696 | Mojtaba Kolahdouzi | Mojtaba Kolahdouzi and Ali Etemad | Toward Fair Facial Expression Recognition with Improved Distribution
Alignment | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We present a novel approach to mitigate bias in facial expression recognition
(FER) models. Our method aims to reduce sensitive attribute information such as
gender, age, or race, in the embeddings produced by FER models. We employ a
kernel mean shrinkage estimator to estimate the kernel mean of the
distributions of the embeddings associated with different sensitive attribute
groups, such as young and old, in the Hilbert space. Using this estimation, we
calculate the maximum mean discrepancy (MMD) distance between the distributions
and incorporate it in the classifier loss along with an adversarial loss, which
is then minimized through the learning process to improve the distribution
alignment. Our method makes sensitive attributes less recognizable for the
model, which in turn promotes fairness. Additionally, for the first time, we
analyze the notion of attractiveness as an important sensitive attribute in FER
models and demonstrate that FER models can indeed exhibit biases towards more
attractive faces. To prove the efficacy of our model in reducing bias regarding
different sensitive attributes (including the newly proposed attractiveness
attribute), we perform several experiments on two widely used datasets, CelebA
and RAF-DB. The results in terms of both accuracy and fairness measures
outperform the state-of-the-art in most cases, demonstrating the effectiveness
of the proposed method.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 14:59:20 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Kolahdouzi",
"Mojtaba",
""
],
[
"Etemad",
"Ali",
""
]
] | TITLE: Toward Fair Facial Expression Recognition with Improved Distribution
Alignment
ABSTRACT: We present a novel approach to mitigate bias in facial expression recognition
(FER) models. Our method aims to reduce sensitive attribute information such as
gender, age, or race, in the embeddings produced by FER models. We employ a
kernel mean shrinkage estimator to estimate the kernel mean of the
distributions of the embeddings associated with different sensitive attribute
groups, such as young and old, in the Hilbert space. Using this estimation, we
calculate the maximum mean discrepancy (MMD) distance between the distributions
and incorporate it in the classifier loss along with an adversarial loss, which
is then minimized through the learning process to improve the distribution
alignment. Our method makes sensitive attributes less recognizable for the
model, which in turn promotes fairness. Additionally, for the first time, we
analyze the notion of attractiveness as an important sensitive attribute in FER
models and demonstrate that FER models can indeed exhibit biases towards more
attractive faces. To prove the efficacy of our model in reducing bias regarding
different sensitive attributes (including the newly proposed attractiveness
attribute), we perform several experiments on two widely used datasets, CelebA
and RAF-DB. The results in terms of both accuracy and fairness measures
outperform the state-of-the-art in most cases, demonstrating the effectiveness
of the proposed method.
|
2306.06712 | Steffen Jung | Steffen Jung, Jovita Lukasik, Margret Keuper | Neural Architecture Design and Robustness: A Dataset | ICLR 2023; project page: http://robustness.vision/ | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning models have proven to be successful in a wide range of machine
learning tasks. Yet, they are often highly sensitive to perturbations on the
input data which can lead to incorrect decisions with high confidence,
hampering their deployment for practical use-cases. Thus, finding architectures
that are (more) robust against perturbations has received much attention in
recent years. Just like the search for well-performing architectures in terms
of clean accuracy, this usually involves a tedious trial-and-error process with
one additional challenge: the evaluation of a network's robustness is
significantly more expensive than its evaluation for clean accuracy. Thus, the
aim of this paper is to facilitate better streamlined research on architectural
design choices with respect to their impact on robustness as well as, for
example, the evaluation of surrogate measures for robustness. We therefore
borrow one of the most commonly considered search spaces for neural
architecture search for image classification, NAS-Bench-201, which contains a
manageable size of 6466 non-isomorphic network designs. We evaluate all these
networks on a range of common adversarial attacks and corruption types and
introduce a database on neural architecture design and robustness evaluations.
We further present three exemplary use cases of this dataset, in which we (i)
benchmark robustness measurements based on Jacobian and Hessian matrices for
their robustness predictability, (ii) perform neural architecture search on
robust accuracies, and (iii) provide an initial analysis of how architectural
design choices affect robustness. We find that carefully crafting the topology
of a network can have substantial impact on its robustness, where networks with
the same parameter count range in mean adversarial robust accuracy from
20%-41%. Code and data is available at http://robustness.vision/.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 16:02:14 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Jung",
"Steffen",
""
],
[
"Lukasik",
"Jovita",
""
],
[
"Keuper",
"Margret",
""
]
] | TITLE: Neural Architecture Design and Robustness: A Dataset
ABSTRACT: Deep learning models have proven to be successful in a wide range of machine
learning tasks. Yet, they are often highly sensitive to perturbations on the
input data which can lead to incorrect decisions with high confidence,
hampering their deployment for practical use-cases. Thus, finding architectures
that are (more) robust against perturbations has received much attention in
recent years. Just like the search for well-performing architectures in terms
of clean accuracy, this usually involves a tedious trial-and-error process with
one additional challenge: the evaluation of a network's robustness is
significantly more expensive than its evaluation for clean accuracy. Thus, the
aim of this paper is to facilitate better streamlined research on architectural
design choices with respect to their impact on robustness as well as, for
example, the evaluation of surrogate measures for robustness. We therefore
borrow one of the most commonly considered search spaces for neural
architecture search for image classification, NAS-Bench-201, which contains a
manageable size of 6466 non-isomorphic network designs. We evaluate all these
networks on a range of common adversarial attacks and corruption types and
introduce a database on neural architecture design and robustness evaluations.
We further present three exemplary use cases of this dataset, in which we (i)
benchmark robustness measurements based on Jacobian and Hessian matrices for
their robustness predictability, (ii) perform neural architecture search on
robust accuracies, and (iii) provide an initial analysis of how architectural
design choices affect robustness. We find that carefully crafting the topology
of a network can have substantial impact on its robustness, where networks with
the same parameter count range in mean adversarial robust accuracy from
20%-41%. Code and data is available at http://robustness.vision/.
|
2306.06716 | Anna Meyer | Anna P. Meyer, Dan Ley, Suraj Srinivas, Himabindu Lakkaraju | On Minimizing the Impact of Dataset Shifts on Actionable Explanations | 30 pages, 19 figures. To be published at UAI 2023 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | The Right to Explanation is an important regulatory principle that allows
individuals to request actionable explanations for algorithmic decisions.
However, several technical challenges arise when providing such actionable
explanations in practice. For instance, models are periodically retrained to
handle dataset shifts. This process may invalidate some of the previously
prescribed explanations, thus rendering them unactionable. But, it is unclear
if and when such invalidations occur, and what factors determine explanation
stability i.e., if an explanation remains unchanged amidst model retraining due
to dataset shifts. In this paper, we address the aforementioned gaps and
provide one of the first theoretical and empirical characterizations of the
factors influencing explanation stability. To this end, we conduct rigorous
theoretical analysis to demonstrate that model curvature, weight decay
parameters while training, and the magnitude of the dataset shift are key
factors that determine the extent of explanation (in)stability. Extensive
experimentation with real-world datasets not only validates our theoretical
results, but also demonstrates that the aforementioned factors dramatically
impact the stability of explanations produced by various state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 16:34:19 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Meyer",
"Anna P.",
""
],
[
"Ley",
"Dan",
""
],
[
"Srinivas",
"Suraj",
""
],
[
"Lakkaraju",
"Himabindu",
""
]
] | TITLE: On Minimizing the Impact of Dataset Shifts on Actionable Explanations
ABSTRACT: The Right to Explanation is an important regulatory principle that allows
individuals to request actionable explanations for algorithmic decisions.
However, several technical challenges arise when providing such actionable
explanations in practice. For instance, models are periodically retrained to
handle dataset shifts. This process may invalidate some of the previously
prescribed explanations, thus rendering them unactionable. But, it is unclear
if and when such invalidations occur, and what factors determine explanation
stability i.e., if an explanation remains unchanged amidst model retraining due
to dataset shifts. In this paper, we address the aforementioned gaps and
provide one of the first theoretical and empirical characterizations of the
factors influencing explanation stability. To this end, we conduct rigorous
theoretical analysis to demonstrate that model curvature, weight decay
parameters while training, and the magnitude of the dataset shift are key
factors that determine the extent of explanation (in)stability. Extensive
experimentation with real-world datasets not only validates our theoretical
results, but also demonstrates that the aforementioned factors dramatically
impact the stability of explanations produced by various state-of-the-art
methods.
|
2306.06731 | Daniel Jakubovitz | Daniel Jakubovitz, David Uliel, Miguel Rodrigues, Raja Giryes | An information-Theoretic Approach to Semi-supervised Transfer Learning | arXiv admin note: substantial text overlap with arXiv:1904.01670 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Transfer learning is a valuable tool in deep learning as it allows
propagating information from one "source dataset" to another "target dataset",
especially in the case of a small number of training examples in the latter.
Yet, discrepancies between the underlying distributions of the source and
target data are commonplace and are known to have a substantial impact on
algorithm performance. In this work we suggest novel information-theoretic
approaches for the analysis of the performance of deep neural networks in the
context of transfer learning. We focus on the task of semi-supervised transfer
learning, in which unlabeled samples from the target dataset are available
during network training on the source dataset. Our theory suggests that one may
improve the transferability of a deep neural network by incorporating
regularization terms on the target data based on information-theoretic
quantities, namely the Mutual Information and the Lautum Information. We
demonstrate the effectiveness of the proposed approaches in various
semi-supervised transfer learning experiments.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 17:45:46 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Jakubovitz",
"Daniel",
""
],
[
"Uliel",
"David",
""
],
[
"Rodrigues",
"Miguel",
""
],
[
"Giryes",
"Raja",
""
]
] | TITLE: An information-Theoretic Approach to Semi-supervised Transfer Learning
ABSTRACT: Transfer learning is a valuable tool in deep learning as it allows
propagating information from one "source dataset" to another "target dataset",
especially in the case of a small number of training examples in the latter.
Yet, discrepancies between the underlying distributions of the source and
target data are commonplace and are known to have a substantial impact on
algorithm performance. In this work we suggest novel information-theoretic
approaches for the analysis of the performance of deep neural networks in the
context of transfer learning. We focus on the task of semi-supervised transfer
learning, in which unlabeled samples from the target dataset are available
during network training on the source dataset. Our theory suggests that one may
improve the transferability of a deep neural network by incorporating
regularization terms on the target data based on information-theoretic
quantities, namely the Mutual Information and the Lautum Information. We
demonstrate the effectiveness of the proposed approaches in various
semi-supervised transfer learning experiments.
|
2306.06753 | Jinming Su | Jinming Su, Wangwang Yang, Junfeng Luo and Xiaolin Wei | 3rd Place Solution for PVUW Challenge 2023: Video Panoptic Segmentation | 3rd Place Solution for PVUW Challenge 2023: Video Panoptic
Segmentation | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In order to deal with the task of video panoptic segmentation in the wild, we
propose a robust integrated video panoptic segmentation solution. In our
solution, we regard the video panoptic segmentation task as a segmentation
target querying task, represent both semantic and instance targets as a set of
queries, and then combine these queries with video features extracted by neural
networks to predict segmentation masks. In order to improve the learning
accuracy and convergence speed of the solution, we add additional tasks of
video semantic segmentation and video instance segmentation for joint training.
In addition, we also add an additional image semantic segmentation model to
further improve the performance of semantic classes. In addition, we also add
some additional operations to improve the robustness of the model. Extensive
experiments on the VIPSeg dataset show that the proposed solution achieves
state-of-the-art performance with 50.04\% VPQ on the VIPSeg test set, which is
3rd place on the video panoptic segmentation track of the PVUW Challenge 2023.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 19:44:40 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Su",
"Jinming",
""
],
[
"Yang",
"Wangwang",
""
],
[
"Luo",
"Junfeng",
""
],
[
"Wei",
"Xiaolin",
""
]
] | TITLE: 3rd Place Solution for PVUW Challenge 2023: Video Panoptic Segmentation
ABSTRACT: In order to deal with the task of video panoptic segmentation in the wild, we
propose a robust integrated video panoptic segmentation solution. In our
solution, we regard the video panoptic segmentation task as a segmentation
target querying task, represent both semantic and instance targets as a set of
queries, and then combine these queries with video features extracted by neural
networks to predict segmentation masks. In order to improve the learning
accuracy and convergence speed of the solution, we add additional tasks of
video semantic segmentation and video instance segmentation for joint training.
In addition, we also add an additional image semantic segmentation model to
further improve the performance of semantic classes. In addition, we also add
some additional operations to improve the robustness of the model. Extensive
experiments on the VIPSeg dataset show that the proposed solution achieves
state-of-the-art performance with 50.04\% VPQ on the VIPSeg test set, which is
3rd place on the video panoptic segmentation track of the PVUW Challenge 2023.
|
2306.06764 | Spyridon Mastorakis | Hasniuj Zahan and Md Washik Al Azad and Ihsan Ali and Spyridon
Mastorakis | IoT-AD: A Framework To Detect Anomalies Among Interconnected IoT Devices | Accepted for publication by the IEEE Internet of Things Journal. The
copyright is with IEEE | null | null | null | cs.NI cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In an Internet of Things (IoT) environment (e.g., smart home), several IoT
devices may be available that are interconnected with each other. In such
interconnected environments, a faulty or compromised IoT device could impact
the operation of other IoT devices. In other words, anomalous behavior
exhibited by an IoT device could propagate to other devices in an IoT
environment. In this paper, we argue that mitigating the propagation of the
anomalous behavior exhibited by a device to other devices is equally important
to detecting this behavior in the first place. In line with this observation,
we present a framework, called IoT Anomaly Detector (IoT-AD), that can not only
detect the anomalous behavior of IoT devices, but also limit and recover from
anomalous behavior that might have affected other devices. We implemented a
prototype of IoT-AD, which we evaluated based on open-source IoT device
datasets as well as through real-world deployment on a small-scale IoT testbed
we have built. We have further evaluated IoT-AD in comparison to prior relevant
approaches. Our evaluation results show that IoT-AD can identify anomalous
behavior of IoT devices in less than 2.12 milliseconds and with up to 98% of
accuracy.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 20:29:05 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zahan",
"Hasniuj",
""
],
[
"Azad",
"Md Washik Al",
""
],
[
"Ali",
"Ihsan",
""
],
[
"Mastorakis",
"Spyridon",
""
]
] | TITLE: IoT-AD: A Framework To Detect Anomalies Among Interconnected IoT Devices
ABSTRACT: In an Internet of Things (IoT) environment (e.g., smart home), several IoT
devices may be available that are interconnected with each other. In such
interconnected environments, a faulty or compromised IoT device could impact
the operation of other IoT devices. In other words, anomalous behavior
exhibited by an IoT device could propagate to other devices in an IoT
environment. In this paper, we argue that mitigating the propagation of the
anomalous behavior exhibited by a device to other devices is equally important
to detecting this behavior in the first place. In line with this observation,
we present a framework, called IoT Anomaly Detector (IoT-AD), that can not only
detect the anomalous behavior of IoT devices, but also limit and recover from
anomalous behavior that might have affected other devices. We implemented a
prototype of IoT-AD, which we evaluated based on open-source IoT device
datasets as well as through real-world deployment on a small-scale IoT testbed
we have built. We have further evaluated IoT-AD in comparison to prior relevant
approaches. Our evaluation results show that IoT-AD can identify anomalous
behavior of IoT devices in less than 2.12 milliseconds and with up to 98% of
accuracy.
|
2306.06773 | Tina Kapur | Nicole M Duggan, Mike Jin, Maria Alejandra Duran Mendicuti, Stephen
Hallisey, Denie Bernier, Lauren A Selame, Ameneh Asgari-Targhi, Chanel E
Fischetti, Ruben Lucassen, Anthony E Samir, Erik Duhaime+, Tina Kapur, Andrew
J Goldsmith | Gamified Crowdsourcing as a Novel Approach to Lung Ultrasound Dataset
Labeling | 27 pages total | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Study Objective: Machine learning models have advanced medical image
processing and can yield faster, more accurate diagnoses. Despite a wealth of
available medical imaging data, high-quality labeled data for model training is
lacking. We investigated whether a gamified crowdsourcing platform enhanced
with inbuilt quality control metrics can produce lung ultrasound clip labels
comparable to those from clinical experts.
Methods: 2,384 lung ultrasound clips were retrospectively collected from 203
patients. Six lung ultrasound experts classified 393 of these clips as having
no B-lines, one or more discrete B-lines, or confluent B-lines to create two
sets of reference standard labels (195 training set clips and 198 test set
clips). Sets were respectively used to A) train users on a gamified
crowdsourcing platform, and B) compare concordance of the resulting crowd
labels to the concordance of individual experts to reference standards.
Results: 99,238 crowdsourced opinions on 2,384 lung ultrasound clips were
collected from 426 unique users over 8 days. On the 198 test set clips, mean
labeling concordance of individual experts relative to the reference standard
was 85.0% +/- 2.0 (SEM), compared to 87.9% crowdsourced label concordance
(p=0.15). When individual experts' opinions were compared to reference standard
labels created by majority vote excluding their own opinion, crowd concordance
was higher than the mean concordance of individual experts to reference
standards (87.4% vs. 80.8% +/- 1.6; p<0.001).
Conclusion: Crowdsourced labels for B-line classification via a gamified
approach achieved expert-level quality. Scalable, high-quality labeling
approaches may facilitate training dataset creation for machine learning model
development.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 20:59:48 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Duggan",
"Nicole M",
""
],
[
"Jin",
"Mike",
""
],
[
"Mendicuti",
"Maria Alejandra Duran",
""
],
[
"Hallisey",
"Stephen",
""
],
[
"Bernier",
"Denie",
""
],
[
"Selame",
"Lauren A",
""
],
[
"Asgari-Targhi",
"Ameneh",
""
],
[
"Fischetti",
"Chanel E",
""
],
[
"Lucassen",
"Ruben",
""
],
[
"Samir",
"Anthony E",
""
],
[
"Duhaime+",
"Erik",
""
],
[
"Kapur",
"Tina",
""
],
[
"Goldsmith",
"Andrew J",
""
]
] | TITLE: Gamified Crowdsourcing as a Novel Approach to Lung Ultrasound Dataset
Labeling
ABSTRACT: Study Objective: Machine learning models have advanced medical image
processing and can yield faster, more accurate diagnoses. Despite a wealth of
available medical imaging data, high-quality labeled data for model training is
lacking. We investigated whether a gamified crowdsourcing platform enhanced
with inbuilt quality control metrics can produce lung ultrasound clip labels
comparable to those from clinical experts.
Methods: 2,384 lung ultrasound clips were retrospectively collected from 203
patients. Six lung ultrasound experts classified 393 of these clips as having
no B-lines, one or more discrete B-lines, or confluent B-lines to create two
sets of reference standard labels (195 training set clips and 198 test set
clips). Sets were respectively used to A) train users on a gamified
crowdsourcing platform, and B) compare concordance of the resulting crowd
labels to the concordance of individual experts to reference standards.
Results: 99,238 crowdsourced opinions on 2,384 lung ultrasound clips were
collected from 426 unique users over 8 days. On the 198 test set clips, mean
labeling concordance of individual experts relative to the reference standard
was 85.0% +/- 2.0 (SEM), compared to 87.9% crowdsourced label concordance
(p=0.15). When individual experts' opinions were compared to reference standard
labels created by majority vote excluding their own opinion, crowd concordance
was higher than the mean concordance of individual experts to reference
standards (87.4% vs. 80.8% +/- 1.6; p<0.001).
Conclusion: Crowdsourced labels for B-line classification via a gamified
approach achieved expert-level quality. Scalable, high-quality labeling
approaches may facilitate training dataset creation for machine learning model
development.
|
2306.06779 | Hai Ye | Hai Ye, Qizhe Xie, Hwee Tou Ng | Multi-Source Test-Time Adaptation as Dueling Bandits for Extractive
Question Answering | Main conference of ACL 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we study multi-source test-time model adaptation from user
feedback, where K distinct models are established for adaptation. To allow
efficient adaptation, we cast the problem as a stochastic decision-making
process, aiming to determine the best adapted model after adaptation. We
discuss two frameworks: multi-armed bandit learning and multi-armed dueling
bandits. Compared to multi-armed bandit learning, the dueling framework allows
pairwise collaboration among K models, which is solved by a novel method named
Co-UCB proposed in this work. Experiments on six datasets of extractive
question answering (QA) show that the dueling framework using Co-UCB is more
effective than other strong baselines for our studied problem.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 21:18:50 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Ye",
"Hai",
""
],
[
"Xie",
"Qizhe",
""
],
[
"Ng",
"Hwee Tou",
""
]
] | TITLE: Multi-Source Test-Time Adaptation as Dueling Bandits for Extractive
Question Answering
ABSTRACT: In this work, we study multi-source test-time model adaptation from user
feedback, where K distinct models are established for adaptation. To allow
efficient adaptation, we cast the problem as a stochastic decision-making
process, aiming to determine the best adapted model after adaptation. We
discuss two frameworks: multi-armed bandit learning and multi-armed dueling
bandits. Compared to multi-armed bandit learning, the dueling framework allows
pairwise collaboration among K models, which is solved by a novel method named
Co-UCB proposed in this work. Experiments on six datasets of extractive
question answering (QA) show that the dueling framework using Co-UCB is more
effective than other strong baselines for our studied problem.
|
2306.06782 | Jie Hu | Jie Hu (University of California Riverside), Qian Zhang (University of
California Riverside), Heng Yin (University of California Riverside) | Augmenting Greybox Fuzzing with Generative AI | 20 pages, 11 figures | null | null | null | cs.CR cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Real-world programs expecting structured inputs often has a format-parsing
stage gating the deeper program space. Neither a mutation-based approach nor a
generative approach can provide a solution that is effective and scalable.
Large language models (LLM) pre-trained with an enormous amount of natural
language corpus have proved to be effective for understanding the implicit
format syntax and generating format-conforming inputs. In this paper, propose
ChatFuzz, a greybox fuzzer augmented by generative AI. More specifically, we
pick a seed in the fuzzer's seed pool and prompt ChatGPT generative models to
variations, which are more likely to be format-conforming and thus of high
quality. We conduct extensive experiments to explore the best practice for
harvesting the power of generative LLM models. The experiment results show that
our approach improves the edge coverage by 12.77\% over the SOTA greybox fuzzer
(AFL++) on 12 target programs from three well-tested benchmarks. As for
vulnerability detection, \sys is able to perform similar to or better than
AFL++ for programs with explicit syntax rules but not for programs with
non-trivial syntax.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 21:44:47 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Hu",
"Jie",
"",
"University of California Riverside"
],
[
"Zhang",
"Qian",
"",
"University of\n California Riverside"
],
[
"Yin",
"Heng",
"",
"University of California Riverside"
]
] | TITLE: Augmenting Greybox Fuzzing with Generative AI
ABSTRACT: Real-world programs expecting structured inputs often has a format-parsing
stage gating the deeper program space. Neither a mutation-based approach nor a
generative approach can provide a solution that is effective and scalable.
Large language models (LLM) pre-trained with an enormous amount of natural
language corpus have proved to be effective for understanding the implicit
format syntax and generating format-conforming inputs. In this paper, propose
ChatFuzz, a greybox fuzzer augmented by generative AI. More specifically, we
pick a seed in the fuzzer's seed pool and prompt ChatGPT generative models to
variations, which are more likely to be format-conforming and thus of high
quality. We conduct extensive experiments to explore the best practice for
harvesting the power of generative LLM models. The experiment results show that
our approach improves the edge coverage by 12.77\% over the SOTA greybox fuzzer
(AFL++) on 12 target programs from three well-tested benchmarks. As for
vulnerability detection, \sys is able to perform similar to or better than
AFL++ for programs with explicit syntax rules but not for programs with
non-trivial syntax.
|
2306.06797 | Jaskaran Singh | Jaskaran Singh | VBSF-TLD: Validation-Based Approach for Soft Computing-Inspired Transfer
Learning in Drone Detection | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | With the increasing utilization of Internet of Things (IoT) enabled drones in
diverse applications like photography, delivery, and surveillance, concerns
regarding privacy and security have become more prominent. Drones have the
ability to capture sensitive information, compromise privacy, and pose security
risks. As a result, the demand for advanced technology to automate drone
detection has become crucial. This paper presents a project on a transfer-based
drone detection scheme, which forms an integral part of a computer vision-based
module and leverages transfer learning to enhance performance. By harnessing
the knowledge of pre-trained models from a related domain, transfer learning
enables improved results even with limited training data. To evaluate the
scheme's performance, we conducted tests on benchmark datasets, including the
Drone-vs-Bird Dataset and the UAVDT dataset. Notably, the scheme's
effectiveness is highlighted by its IOU-based validation results, demonstrating
the potential of deep learning-based technology in automating drone detection
in critical areas such as airports, military bases, and other high-security
zones.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 22:30:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Singh",
"Jaskaran",
""
]
] | TITLE: VBSF-TLD: Validation-Based Approach for Soft Computing-Inspired Transfer
Learning in Drone Detection
ABSTRACT: With the increasing utilization of Internet of Things (IoT) enabled drones in
diverse applications like photography, delivery, and surveillance, concerns
regarding privacy and security have become more prominent. Drones have the
ability to capture sensitive information, compromise privacy, and pose security
risks. As a result, the demand for advanced technology to automate drone
detection has become crucial. This paper presents a project on a transfer-based
drone detection scheme, which forms an integral part of a computer vision-based
module and leverages transfer learning to enhance performance. By harnessing
the knowledge of pre-trained models from a related domain, transfer learning
enables improved results even with limited training data. To evaluate the
scheme's performance, we conducted tests on benchmark datasets, including the
Drone-vs-Bird Dataset and the UAVDT dataset. Notably, the scheme's
effectiveness is highlighted by its IOU-based validation results, demonstrating
the potential of deep learning-based technology in automating drone detection
in critical areas such as airports, military bases, and other high-security
zones.
|
2306.06801 | Josephine Lamp | Josephine Lamp, Yuxin Wu, Steven Lamp, Prince Afriyie, Kenneth
Bilchick, Lu Feng, Sula Mazimba | CARNA: Characterizing Advanced heart failure Risk and hemodyNAmic
phenotypes using learned multi-valued decision diagrams | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Early identification of high risk heart failure (HF) patients is key to
timely allocation of life-saving therapies. Hemodynamic assessments can
facilitate risk stratification and enhance understanding of HF trajectories.
However, risk assessment for HF is a complex, multi-faceted decision-making
process that can be challenging. Previous risk models for HF do not integrate
invasive hemodynamics or support missing data, and use statistical methods
prone to bias or machine learning methods that are not interpretable. To
address these limitations, this paper presents CARNA, a hemodynamic risk
stratification and phenotyping framework for advanced HF that takes advantage
of the explainability and expressivity of machine learned Multi-Valued Decision
Diagrams (MVDDs). This interpretable framework learns risk scores that predict
the probability of patient outcomes, and outputs descriptive patient phenotypes
(sets of features and thresholds) that characterize each predicted risk score.
CARNA incorporates invasive hemodynamics and can make predictions on missing
data. The CARNA models were trained and validated using a total of five
advanced HF patient cohorts collected from previous trials, and compared with
six established HF risk scores and three traditional ML risk models. CARNA
provides robust risk stratification, outperforming all previous benchmarks.
Although focused on advanced HF, the CARNA framework is general purpose and can
be used to learn risk stratifications for other diseases and medical
applications.
| [
{
"version": "v1",
"created": "Sun, 11 Jun 2023 22:56:59 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Lamp",
"Josephine",
""
],
[
"Wu",
"Yuxin",
""
],
[
"Lamp",
"Steven",
""
],
[
"Afriyie",
"Prince",
""
],
[
"Bilchick",
"Kenneth",
""
],
[
"Feng",
"Lu",
""
],
[
"Mazimba",
"Sula",
""
]
] | TITLE: CARNA: Characterizing Advanced heart failure Risk and hemodyNAmic
phenotypes using learned multi-valued decision diagrams
ABSTRACT: Early identification of high risk heart failure (HF) patients is key to
timely allocation of life-saving therapies. Hemodynamic assessments can
facilitate risk stratification and enhance understanding of HF trajectories.
However, risk assessment for HF is a complex, multi-faceted decision-making
process that can be challenging. Previous risk models for HF do not integrate
invasive hemodynamics or support missing data, and use statistical methods
prone to bias or machine learning methods that are not interpretable. To
address these limitations, this paper presents CARNA, a hemodynamic risk
stratification and phenotyping framework for advanced HF that takes advantage
of the explainability and expressivity of machine learned Multi-Valued Decision
Diagrams (MVDDs). This interpretable framework learns risk scores that predict
the probability of patient outcomes, and outputs descriptive patient phenotypes
(sets of features and thresholds) that characterize each predicted risk score.
CARNA incorporates invasive hemodynamics and can make predictions on missing
data. The CARNA models were trained and validated using a total of five
advanced HF patient cohorts collected from previous trials, and compared with
six established HF risk scores and three traditional ML risk models. CARNA
provides robust risk stratification, outperforming all previous benchmarks.
Although focused on advanced HF, the CARNA framework is general purpose and can
be used to learn risk stratifications for other diseases and medical
applications.
|
2306.06814 | Ji-Sang Hwang | Ji-Sang Hwang, Sang-Hoon Lee, and Seong-Whan Lee | HiddenSinger: High-Quality Singing Voice Synthesis via Neural Audio
Codec and Latent Diffusion Models | 11 pages, 5 figures, 5 tables, under review | null | null | null | eess.AS cs.AI cs.SD eess.SP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recently, denoising diffusion models have demonstrated remarkable performance
among generative models in various domains. However, in the speech domain, the
application of diffusion models for synthesizing time-varying audio faces
limitations in terms of complexity and controllability, as speech synthesis
requires very high-dimensional samples with long-term acoustic features. To
alleviate the challenges posed by model complexity in singing voice synthesis,
we propose HiddenSinger, a high-quality singing voice synthesis system using a
neural audio codec and latent diffusion models. To ensure high-fidelity audio,
we introduce an audio autoencoder that can encode audio into an audio codec as
a compressed representation and reconstruct the high-fidelity audio from the
low-dimensional compressed latent vector. Subsequently, we use the latent
diffusion models to sample a latent representation from a musical score. In
addition, our proposed model is extended to an unsupervised singing voice
learning framework, HiddenSinger-U, to train the model using an unlabeled
singing voice dataset. Experimental results demonstrate that our model
outperforms previous models in terms of audio quality. Furthermore, the
HiddenSinger-U can synthesize high-quality singing voices of speakers trained
solely on unlabeled data.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 01:21:41 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Hwang",
"Ji-Sang",
""
],
[
"Lee",
"Sang-Hoon",
""
],
[
"Lee",
"Seong-Whan",
""
]
] | TITLE: HiddenSinger: High-Quality Singing Voice Synthesis via Neural Audio
Codec and Latent Diffusion Models
ABSTRACT: Recently, denoising diffusion models have demonstrated remarkable performance
among generative models in various domains. However, in the speech domain, the
application of diffusion models for synthesizing time-varying audio faces
limitations in terms of complexity and controllability, as speech synthesis
requires very high-dimensional samples with long-term acoustic features. To
alleviate the challenges posed by model complexity in singing voice synthesis,
we propose HiddenSinger, a high-quality singing voice synthesis system using a
neural audio codec and latent diffusion models. To ensure high-fidelity audio,
we introduce an audio autoencoder that can encode audio into an audio codec as
a compressed representation and reconstruct the high-fidelity audio from the
low-dimensional compressed latent vector. Subsequently, we use the latent
diffusion models to sample a latent representation from a musical score. In
addition, our proposed model is extended to an unsupervised singing voice
learning framework, HiddenSinger-U, to train the model using an unlabeled
singing voice dataset. Experimental results demonstrate that our model
outperforms previous models in terms of audio quality. Furthermore, the
HiddenSinger-U can synthesize high-quality singing voices of speakers trained
solely on unlabeled data.
|
2306.06825 | Zahra Ghodsi | Ghada Almashaqbeh, Zahra Ghodsi | AnoFel: Supporting Anonymity for Privacy-Preserving Federated Learning | null | null | null | null | cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated learning enables users to collaboratively train a machine learning
model over their private datasets. Secure aggregation protocols are employed to
mitigate information leakage about the local datasets. This setup, however,
still leaks the participation of a user in a training iteration, which can also
be sensitive. Protecting user anonymity is even more challenging in dynamic
environments where users may (re)join or leave the training process at any
point of time. In this paper, we introduce AnoFel, the first framework to
support private and anonymous dynamic participation in federated learning.
AnoFel leverages several cryptographic primitives, the concept of anonymity
sets, differential privacy, and a public bulletin board to support anonymous
user registration, as well as unlinkable and confidential model updates
submission. Additionally, our system allows dynamic participation, where users
can join or leave at any time, without needing any recovery protocol or
interaction. To assess security, we formalize a notion for privacy and
anonymity in federated learning, and formally prove that AnoFel satisfies this
notion. To the best of our knowledge, our system is the first solution with
provable anonymity guarantees. To assess efficiency, we provide a concrete
implementation of AnoFel, and conduct experiments showing its ability to
support learning applications scaling to a large number of clients. For an
MNIST classification task with 512 clients, the client setup takes less than 3
sec, and a training iteration can be finished in 3.2 sec. We also compare our
system with prior work and demonstrate its practicality for contemporary
learning tasks.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 02:25:44 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Almashaqbeh",
"Ghada",
""
],
[
"Ghodsi",
"Zahra",
""
]
] | TITLE: AnoFel: Supporting Anonymity for Privacy-Preserving Federated Learning
ABSTRACT: Federated learning enables users to collaboratively train a machine learning
model over their private datasets. Secure aggregation protocols are employed to
mitigate information leakage about the local datasets. This setup, however,
still leaks the participation of a user in a training iteration, which can also
be sensitive. Protecting user anonymity is even more challenging in dynamic
environments where users may (re)join or leave the training process at any
point of time. In this paper, we introduce AnoFel, the first framework to
support private and anonymous dynamic participation in federated learning.
AnoFel leverages several cryptographic primitives, the concept of anonymity
sets, differential privacy, and a public bulletin board to support anonymous
user registration, as well as unlinkable and confidential model updates
submission. Additionally, our system allows dynamic participation, where users
can join or leave at any time, without needing any recovery protocol or
interaction. To assess security, we formalize a notion for privacy and
anonymity in federated learning, and formally prove that AnoFel satisfies this
notion. To the best of our knowledge, our system is the first solution with
provable anonymity guarantees. To assess efficiency, we provide a concrete
implementation of AnoFel, and conduct experiments showing its ability to
support learning applications scaling to a large number of clients. For an
MNIST classification task with 512 clients, the client setup takes less than 3
sec, and a training iteration can be finished in 3.2 sec. We also compare our
system with prior work and demonstrate its practicality for contemporary
learning tasks.
|
2306.06852 | Jiuling Zhang | Jiuling Zhang, Zhiming Ding | Rethink DARTS Search Space and Renovate a New Benchmark | Accepted by ICML 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | DARTS search space (DSS) has become a canonical benchmark for NAS whereas
some emerging works pointed out the issue of narrow accuracy range and claimed
it would hurt the method ranking. We observe some recent studies already suffer
from this issue that overshadows the meaning of scores. In this work, we first
propose and orchestrate a suite of improvements to frame a larger and harder
DSS, termed LHD, while retaining high efficiency in search. We step forward to
renovate a LHD-based new benchmark, taking care of both discernibility and
accessibility. Specifically, we re-implement twelve baselines and evaluate them
across twelve conditions by combining two underexpolored influential factors:
transductive robustness and discretization policy, to reasonably construct a
benchmark upon multi-condition evaluation. Considering that the tabular
benchmarks are always insufficient to adequately evaluate the methods of neural
architecture search (NAS), our work can serve as a crucial basis for the future
progress of NAS. https://github.com/chaoji90/LHD
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 03:56:09 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhang",
"Jiuling",
""
],
[
"Ding",
"Zhiming",
""
]
] | TITLE: Rethink DARTS Search Space and Renovate a New Benchmark
ABSTRACT: DARTS search space (DSS) has become a canonical benchmark for NAS whereas
some emerging works pointed out the issue of narrow accuracy range and claimed
it would hurt the method ranking. We observe some recent studies already suffer
from this issue that overshadows the meaning of scores. In this work, we first
propose and orchestrate a suite of improvements to frame a larger and harder
DSS, termed LHD, while retaining high efficiency in search. We step forward to
renovate a LHD-based new benchmark, taking care of both discernibility and
accessibility. Specifically, we re-implement twelve baselines and evaluate them
across twelve conditions by combining two underexpolored influential factors:
transductive robustness and discretization policy, to reasonably construct a
benchmark upon multi-condition evaluation. Considering that the tabular
benchmarks are always insufficient to adequately evaluate the methods of neural
architecture search (NAS), our work can serve as a crucial basis for the future
progress of NAS. https://github.com/chaoji90/LHD
|
2306.06866 | Jiaojiao Fan | Jiaojiao Fan and David Alvarez-Melis | Generating Synthetic Datasets by Interpolating along Generalized
Geodesics | null | Conference on Uncertainty in Artificial Intelligence (UAI) 2023 | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data for pretraining machine learning models often consists of collections of
heterogeneous datasets. Although training on their union is reasonable in
agnostic settings, it might be suboptimal when the target domain -- where the
model will ultimately be used -- is known in advance. In that case, one would
ideally pretrain only on the dataset(s) most similar to the target one. Instead
of limiting this choice to those datasets already present in the pretraining
collection, here we explore extending this search to all datasets that can be
synthesized as `combinations' of them. We define such combinations as
multi-dataset interpolations, formalized through the notion of generalized
geodesics from optimal transport (OT) theory. We compute these geodesics using
a recent notion of distance between labeled datasets, and derive alternative
interpolation schemes based on it: using either barycentric projections or
optimal transport maps, the latter computed using recent neural OT methods.
These methods are scalable, efficient, and -- notably -- can be used to
interpolate even between datasets with distinct and unrelated label sets.
Through various experiments in transfer learning in computer vision, we
demonstrate this is a promising new approach for targeted on-demand dataset
synthesis.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 04:46:44 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Fan",
"Jiaojiao",
""
],
[
"Alvarez-Melis",
"David",
""
]
] | TITLE: Generating Synthetic Datasets by Interpolating along Generalized
Geodesics
ABSTRACT: Data for pretraining machine learning models often consists of collections of
heterogeneous datasets. Although training on their union is reasonable in
agnostic settings, it might be suboptimal when the target domain -- where the
model will ultimately be used -- is known in advance. In that case, one would
ideally pretrain only on the dataset(s) most similar to the target one. Instead
of limiting this choice to those datasets already present in the pretraining
collection, here we explore extending this search to all datasets that can be
synthesized as `combinations' of them. We define such combinations as
multi-dataset interpolations, formalized through the notion of generalized
geodesics from optimal transport (OT) theory. We compute these geodesics using
a recent notion of distance between labeled datasets, and derive alternative
interpolation schemes based on it: using either barycentric projections or
optimal transport maps, the latter computed using recent neural OT methods.
These methods are scalable, efficient, and -- notably -- can be used to
interpolate even between datasets with distinct and unrelated label sets.
Through various experiments in transfer learning in computer vision, we
demonstrate this is a promising new approach for targeted on-demand dataset
synthesis.
|
2306.06870 | Sijie Zhao | Sijie Zhao, Yixiao Ge, Zhongang Qi, Lin Song, Xiaohan Ding, Zehua Xie,
Ying Shan | Sticker820K: Empowering Interactive Retrieval with Stickers | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Stickers have become a ubiquitous part of modern-day communication, conveying
complex emotions through visual imagery. To facilitate the development of more
powerful algorithms for analyzing stickers, we propose a large-scale Chinese
sticker dataset, namely Sticker820K, which consists of 820k image-text pairs.
Each sticker has rich and high-quality textual annotations, including
descriptions, optical characters, emotional labels, and style classifications.
Although vision-language tasks in the domain of natural images have been well
studied, directly applying the those models, such as CLIP, to sticker data is
not an optimal solution due to the discrepant nature between natural and
emotive image data. Therefore, we propose StickerCLIP as a benchmark model on
the Sticker820K dataset. For the text-to-image retrieval task, our StickerCLIP
demonstrates strong superiority over the CLIP, which achieves an absolute gain
of 66.0\% in mean recall on the Sticker820K test set. Additionally, we endeavor
to extend the recently popularized LLM by means of prompt tuning, integrating
its ability for sticker retrieval and allowing users to retrieve stickers
through instructions. We validate the feasibility of this method, demonstrating
the immense potential of prompt tuning in expanding LLM abilities while not
affecting the quality of upstream tasks.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 05:06:53 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhao",
"Sijie",
""
],
[
"Ge",
"Yixiao",
""
],
[
"Qi",
"Zhongang",
""
],
[
"Song",
"Lin",
""
],
[
"Ding",
"Xiaohan",
""
],
[
"Xie",
"Zehua",
""
],
[
"Shan",
"Ying",
""
]
] | TITLE: Sticker820K: Empowering Interactive Retrieval with Stickers
ABSTRACT: Stickers have become a ubiquitous part of modern-day communication, conveying
complex emotions through visual imagery. To facilitate the development of more
powerful algorithms for analyzing stickers, we propose a large-scale Chinese
sticker dataset, namely Sticker820K, which consists of 820k image-text pairs.
Each sticker has rich and high-quality textual annotations, including
descriptions, optical characters, emotional labels, and style classifications.
Although vision-language tasks in the domain of natural images have been well
studied, directly applying the those models, such as CLIP, to sticker data is
not an optimal solution due to the discrepant nature between natural and
emotive image data. Therefore, we propose StickerCLIP as a benchmark model on
the Sticker820K dataset. For the text-to-image retrieval task, our StickerCLIP
demonstrates strong superiority over the CLIP, which achieves an absolute gain
of 66.0\% in mean recall on the Sticker820K test set. Additionally, we endeavor
to extend the recently popularized LLM by means of prompt tuning, integrating
its ability for sticker retrieval and allowing users to retrieve stickers
through instructions. We validate the feasibility of this method, demonstrating
the immense potential of prompt tuning in expanding LLM abilities while not
affecting the quality of upstream tasks.
|
2306.06872 | Hao Sun | Hao Sun, Yang Li, Liwei Deng, Bowen Li, Binyuan Hui, Binhua Li, Yunshi
Lan, Yan Zhang, Yongbin Li | History Semantic Graph Enhanced Conversational KBQA with Temporal
Information Modeling | Accepted to ACL 2023 Main Conference | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Context information modeling is an important task in conversational KBQA.
However, existing methods usually assume the independence of utterances and
model them in isolation. In this paper, we propose a History Semantic Graph
Enhanced KBQA model (HSGE) that is able to effectively model long-range
semantic dependencies in conversation history while maintaining low
computational cost. The framework incorporates a context-aware encoder, which
employs a dynamic memory decay mechanism and models context at different levels
of granularity. We evaluate HSGE on a widely used benchmark dataset for complex
sequential question answering. Experimental results demonstrate that it
outperforms existing baselines averaged on all question types.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 05:10:58 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Sun",
"Hao",
""
],
[
"Li",
"Yang",
""
],
[
"Deng",
"Liwei",
""
],
[
"Li",
"Bowen",
""
],
[
"Hui",
"Binyuan",
""
],
[
"Li",
"Binhua",
""
],
[
"Lan",
"Yunshi",
""
],
[
"Zhang",
"Yan",
""
],
[
"Li",
"Yongbin",
""
]
] | TITLE: History Semantic Graph Enhanced Conversational KBQA with Temporal
Information Modeling
ABSTRACT: Context information modeling is an important task in conversational KBQA.
However, existing methods usually assume the independence of utterances and
model them in isolation. In this paper, we propose a History Semantic Graph
Enhanced KBQA model (HSGE) that is able to effectively model long-range
semantic dependencies in conversation history while maintaining low
computational cost. The framework incorporates a context-aware encoder, which
employs a dynamic memory decay mechanism and models context at different levels
of granularity. We evaluate HSGE on a widely used benchmark dataset for complex
sequential question answering. Experimental results demonstrate that it
outperforms existing baselines averaged on all question types.
|
2306.06877 | Anlan Sun | AnLan Sun, Zhao Zhang, Meng Lei, Yuting Dai, Dong Wang, Liwei Wang | Boosting Breast Ultrasound Video Classification by the Guidance of
Keyframe Feature Centers | Medical Image Computing and Computer-Assisted Intervention 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Breast ultrasound videos contain richer information than ultrasound images,
therefore it is more meaningful to develop video models for this diagnosis
task. However, the collection of ultrasound video datasets is much harder. In
this paper, we explore the feasibility of enhancing the performance of
ultrasound video classification using the static image dataset. To this end, we
propose KGA-Net and coherence loss. The KGA-Net adopts both video clips and
static images to train the network. The coherence loss uses the feature centers
generated by the static images to guide the frame attention in the video model.
Our KGA-Net boosts the performance on the public BUSV dataset by a large
margin. The visualization results of frame attention prove the explainability
of our method. The codes and model weights of our method will be made publicly
available.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 05:30:09 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Sun",
"AnLan",
""
],
[
"Zhang",
"Zhao",
""
],
[
"Lei",
"Meng",
""
],
[
"Dai",
"Yuting",
""
],
[
"Wang",
"Dong",
""
],
[
"Wang",
"Liwei",
""
]
] | TITLE: Boosting Breast Ultrasound Video Classification by the Guidance of
Keyframe Feature Centers
ABSTRACT: Breast ultrasound videos contain richer information than ultrasound images,
therefore it is more meaningful to develop video models for this diagnosis
task. However, the collection of ultrasound video datasets is much harder. In
this paper, we explore the feasibility of enhancing the performance of
ultrasound video classification using the static image dataset. To this end, we
propose KGA-Net and coherence loss. The KGA-Net adopts both video clips and
static images to train the network. The coherence loss uses the feature centers
generated by the static images to guide the frame attention in the video model.
Our KGA-Net boosts the performance on the public BUSV dataset by a large
margin. The visualization results of frame attention prove the explainability
of our method. The codes and model weights of our method will be made publicly
available.
|
2306.06885 | Yu Chen | Yu Chen, Yang Yu, Rongrong Ni, Yao Zhao, Haoliang Li | NPVForensics: Jointing Non-critical Phonemes and Visemes for Deepfake
Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Deepfake technologies empowered by deep learning are rapidly evolving,
creating new security concerns for society. Existing multimodal detection
methods usually capture audio-visual inconsistencies to expose Deepfake videos.
More seriously, the advanced Deepfake technology realizes the audio-visual
calibration of the critical phoneme-viseme regions, achieving a more realistic
tampering effect, which brings new challenges. To address this problem, we
propose a novel Deepfake detection method to mine the correlation between
Non-critical Phonemes and Visemes, termed NPVForensics. Firstly, we propose the
Local Feature Aggregation block with Swin Transformer (LFA-ST) to construct
non-critical phoneme-viseme and corresponding facial feature streams
effectively. Secondly, we design a loss function for the fine-grained motion of
the talking face to measure the evolutionary consistency of non-critical
phoneme-viseme. Next, we design a phoneme-viseme awareness module for
cross-modal feature fusion and representation alignment, so that the modality
gap can be reduced and the intrinsic complementarity of the two modalities can
be better explored. Finally, a self-supervised pre-training strategy is
leveraged to thoroughly learn the audio-visual correspondences in natural
videos. In this manner, our model can be easily adapted to the downstream
Deepfake datasets with fine-tuning. Extensive experiments on existing
benchmarks demonstrate that the proposed approach outperforms state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 06:06:05 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Chen",
"Yu",
""
],
[
"Yu",
"Yang",
""
],
[
"Ni",
"Rongrong",
""
],
[
"Zhao",
"Yao",
""
],
[
"Li",
"Haoliang",
""
]
] | TITLE: NPVForensics: Jointing Non-critical Phonemes and Visemes for Deepfake
Detection
ABSTRACT: Deepfake technologies empowered by deep learning are rapidly evolving,
creating new security concerns for society. Existing multimodal detection
methods usually capture audio-visual inconsistencies to expose Deepfake videos.
More seriously, the advanced Deepfake technology realizes the audio-visual
calibration of the critical phoneme-viseme regions, achieving a more realistic
tampering effect, which brings new challenges. To address this problem, we
propose a novel Deepfake detection method to mine the correlation between
Non-critical Phonemes and Visemes, termed NPVForensics. Firstly, we propose the
Local Feature Aggregation block with Swin Transformer (LFA-ST) to construct
non-critical phoneme-viseme and corresponding facial feature streams
effectively. Secondly, we design a loss function for the fine-grained motion of
the talking face to measure the evolutionary consistency of non-critical
phoneme-viseme. Next, we design a phoneme-viseme awareness module for
cross-modal feature fusion and representation alignment, so that the modality
gap can be reduced and the intrinsic complementarity of the two modalities can
be better explored. Finally, a self-supervised pre-training strategy is
leveraged to thoroughly learn the audio-visual correspondences in natural
videos. In this manner, our model can be easily adapted to the downstream
Deepfake datasets with fine-tuning. Extensive experiments on existing
benchmarks demonstrate that the proposed approach outperforms state-of-the-art
methods.
|
2306.06893 | Ba-Thinh Lam Mr | Huy T. Nguyen, Thinh B. Lam, Quan D.D. Tran, Minh T. Nguyen, Dat T.
Chung, and Vinh Q. Dinh | In-context Cross-Density Adaptation on Noisy Mammogram Abnormalities
Detection | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper investigates the impact of breast density distribution on the
generalization performance of deep-learning models on mammography images using
the VinDr-Mammo dataset. We explore the use of domain adaptation techniques,
specifically Domain Adaptive Object Detection (DAOD) with the Noise Latent
Transferability Exploration (NLTE) framework, to improve model performance
across breast densities under noisy labeling circumstances. We propose a robust
augmentation framework to bridge the domain gap between the source and target
inside a dataset. Our results show that DAOD-based methods, along with the
proposed augmentation framework, can improve the generalization performance of
deep-learning models (+5% overall mAP improvement approximately in our
experimental results compared to commonly used detection models). This paper
highlights the importance of domain adaptation techniques in medical imaging,
particularly in the context of breast density distribution, which is critical
in mammography.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 06:46:42 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Nguyen",
"Huy T.",
""
],
[
"Lam",
"Thinh B.",
""
],
[
"Tran",
"Quan D. D.",
""
],
[
"Nguyen",
"Minh T.",
""
],
[
"Chung",
"Dat T.",
""
],
[
"Dinh",
"Vinh Q.",
""
]
] | TITLE: In-context Cross-Density Adaptation on Noisy Mammogram Abnormalities
Detection
ABSTRACT: This paper investigates the impact of breast density distribution on the
generalization performance of deep-learning models on mammography images using
the VinDr-Mammo dataset. We explore the use of domain adaptation techniques,
specifically Domain Adaptive Object Detection (DAOD) with the Noise Latent
Transferability Exploration (NLTE) framework, to improve model performance
across breast densities under noisy labeling circumstances. We propose a robust
augmentation framework to bridge the domain gap between the source and target
inside a dataset. Our results show that DAOD-based methods, along with the
proposed augmentation framework, can improve the generalization performance of
deep-learning models (+5% overall mAP improvement approximately in our
experimental results compared to commonly used detection models). This paper
highlights the importance of domain adaptation techniques in medical imaging,
particularly in the context of breast density distribution, which is critical
in mammography.
|
2306.06894 | Senlin Shu | Senlin Shu, Shuo He, Haobo Wang, Hongxin Wei, Tao Xiang, Lei Feng | A Generalized Unbiased Risk Estimator for Learning with Augmented
Classes | Accepted by AAAI 2023, 17 pages, 4 tables, 4 figures | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | In contrast to the standard learning paradigm where all classes can be
observed in training data, learning with augmented classes (LAC) tackles the
problem where augmented classes unobserved in the training data may emerge in
the test phase. Previous research showed that given unlabeled data, an unbiased
risk estimator (URE) can be derived, which can be minimized for LAC with
theoretical guarantees. However, this URE is only restricted to the specific
type of one-versus-rest loss functions for multi-class classification, making
it not flexible enough when the loss needs to be changed with the dataset in
practice. In this paper, we propose a generalized URE that can be equipped with
arbitrary loss functions while maintaining the theoretical guarantees, given
unlabeled data for LAC. To alleviate the issue of negative empirical risk
commonly encountered by previous studies, we further propose a novel
risk-penalty regularization term. Experiments demonstrate the effectiveness of
our proposed method.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 06:52:04 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shu",
"Senlin",
""
],
[
"He",
"Shuo",
""
],
[
"Wang",
"Haobo",
""
],
[
"Wei",
"Hongxin",
""
],
[
"Xiang",
"Tao",
""
],
[
"Feng",
"Lei",
""
]
] | TITLE: A Generalized Unbiased Risk Estimator for Learning with Augmented
Classes
ABSTRACT: In contrast to the standard learning paradigm where all classes can be
observed in training data, learning with augmented classes (LAC) tackles the
problem where augmented classes unobserved in the training data may emerge in
the test phase. Previous research showed that given unlabeled data, an unbiased
risk estimator (URE) can be derived, which can be minimized for LAC with
theoretical guarantees. However, this URE is only restricted to the specific
type of one-versus-rest loss functions for multi-class classification, making
it not flexible enough when the loss needs to be changed with the dataset in
practice. In this paper, we propose a generalized URE that can be equipped with
arbitrary loss functions while maintaining the theoretical guarantees, given
unlabeled data for LAC. To alleviate the issue of negative empirical risk
commonly encountered by previous studies, we further propose a novel
risk-penalty regularization term. Experiments demonstrate the effectiveness of
our proposed method.
|
2306.06895 | Xing Wang | Xing Wang, Zhendong Wang, Kexin Yang, Junlan Feng, Zhiyan Song, Chao
Deng, Lin zhu | MPPN: Multi-Resolution Periodic Pattern Network For Long-Term Time
Series Forecasting | 21 pages | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Long-term time series forecasting plays an important role in various
real-world scenarios. Recent deep learning methods for long-term series
forecasting tend to capture the intricate patterns of time series by
decomposition-based or sampling-based methods. However, most of the extracted
patterns may include unpredictable noise and lack good interpretability.
Moreover, the multivariate series forecasting methods usually ignore the
individual characteristics of each variate, which may affecting the prediction
accuracy. To capture the intrinsic patterns of time series, we propose a novel
deep learning network architecture, named Multi-resolution Periodic Pattern
Network (MPPN), for long-term series forecasting. We first construct
context-aware multi-resolution semantic units of time series and employ
multi-periodic pattern mining to capture the key patterns of time series. Then,
we propose a channel adaptive module to capture the perceptions of multivariate
towards different patterns. In addition, we present an entropy-based method for
evaluating the predictability of time series and providing an upper bound on
the prediction accuracy before forecasting. Our experimental evaluation on nine
real-world benchmarks demonstrated that MPPN significantly outperforms the
state-of-the-art Transformer-based, decomposition-based and sampling-based
methods for long-term series forecasting.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 07:00:37 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Wang",
"Xing",
""
],
[
"Wang",
"Zhendong",
""
],
[
"Yang",
"Kexin",
""
],
[
"Feng",
"Junlan",
""
],
[
"Song",
"Zhiyan",
""
],
[
"Deng",
"Chao",
""
],
[
"zhu",
"Lin",
""
]
] | TITLE: MPPN: Multi-Resolution Periodic Pattern Network For Long-Term Time
Series Forecasting
ABSTRACT: Long-term time series forecasting plays an important role in various
real-world scenarios. Recent deep learning methods for long-term series
forecasting tend to capture the intricate patterns of time series by
decomposition-based or sampling-based methods. However, most of the extracted
patterns may include unpredictable noise and lack good interpretability.
Moreover, the multivariate series forecasting methods usually ignore the
individual characteristics of each variate, which may affecting the prediction
accuracy. To capture the intrinsic patterns of time series, we propose a novel
deep learning network architecture, named Multi-resolution Periodic Pattern
Network (MPPN), for long-term series forecasting. We first construct
context-aware multi-resolution semantic units of time series and employ
multi-periodic pattern mining to capture the key patterns of time series. Then,
we propose a channel adaptive module to capture the perceptions of multivariate
towards different patterns. In addition, we present an entropy-based method for
evaluating the predictability of time series and providing an upper bound on
the prediction accuracy before forecasting. Our experimental evaluation on nine
real-world benchmarks demonstrated that MPPN significantly outperforms the
state-of-the-art Transformer-based, decomposition-based and sampling-based
methods for long-term series forecasting.
|
2306.06904 | Yuwen Deng | Yuwen Deng, Wang Kang, Wei W. Xing | Differentiable Multi-Fidelity Fusion: Efficient Learning of Physics
Simulations with Neural Architecture Search and Transfer Learning | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With rapid progress in deep learning, neural networks have been widely used
in scientific research and engineering applications as surrogate models.
Despite the great success of neural networks in fitting complex systems, two
major challenges still remain: i) the lack of generalization on different
problems/datasets, and ii) the demand for large amounts of simulation data that
are computationally expensive. To resolve these challenges, we propose the
differentiable \mf (DMF) model, which leverages neural architecture search
(NAS) to automatically search the suitable model architecture for different
problems, and transfer learning to transfer the learned knowledge from
low-fidelity (fast but inaccurate) data to high-fidelity (slow but accurate)
model. Novel and latest machine learning techniques such as hyperparameters
search and alternate learning are used to improve the efficiency and robustness
of DMF. As a result, DMF can efficiently learn the physics simulations with
only a few high-fidelity training samples, and outperform the state-of-the-art
methods with a significant margin (with up to 58$\%$ improvement in RMSE) based
on a variety of synthetic and practical benchmark problems.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 07:18:13 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Deng",
"Yuwen",
""
],
[
"Kang",
"Wang",
""
],
[
"Xing",
"Wei W.",
""
]
] | TITLE: Differentiable Multi-Fidelity Fusion: Efficient Learning of Physics
Simulations with Neural Architecture Search and Transfer Learning
ABSTRACT: With rapid progress in deep learning, neural networks have been widely used
in scientific research and engineering applications as surrogate models.
Despite the great success of neural networks in fitting complex systems, two
major challenges still remain: i) the lack of generalization on different
problems/datasets, and ii) the demand for large amounts of simulation data that
are computationally expensive. To resolve these challenges, we propose the
differentiable \mf (DMF) model, which leverages neural architecture search
(NAS) to automatically search the suitable model architecture for different
problems, and transfer learning to transfer the learned knowledge from
low-fidelity (fast but inaccurate) data to high-fidelity (slow but accurate)
model. Novel and latest machine learning techniques such as hyperparameters
search and alternate learning are used to improve the efficiency and robustness
of DMF. As a result, DMF can efficiently learn the physics simulations with
only a few high-fidelity training samples, and outperform the state-of-the-art
methods with a significant margin (with up to 58$\%$ improvement in RMSE) based
on a variety of synthetic and practical benchmark problems.
|
2306.06928 | Hong Liu | Hong Liu | Sparse-Inductive Generative Adversarial Hashing for Nearest Neighbor
Search | 18 pages, 8 figures, 7 tables | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Unsupervised hashing has received extensive research focus on the past
decade, which typically aims at preserving a predefined metric (i.e. Euclidean
metric) in the Hamming space. To this end, the encoding functions of the
existing hashing are typically quasi-isometric, which devote to reducing the
quantization loss from the target metric space to the discrete Hamming space.
However, it is indeed problematic to directly minimize such error, since such
mentioned two metric spaces are heterogeneous, and the quasi-isometric mapping
is non-linear. The former leads to inconsistent feature distributions, while
the latter leads to problematic optimization issues. In this paper, we propose
a novel unsupervised hashing method, termed Sparsity-Induced Generative
Adversarial Hashing (SiGAH), to encode large-scale high-dimensional features
into binary codes, which well solves the two problems through a generative
adversarial training framework. Instead of minimizing the quantization loss,
our key innovation lies in enforcing the learned Hamming space to have similar
data distribution to the target metric space via a generative model. In
particular, we formulate a ReLU-based neural network as a generator to output
binary codes and an MSE-loss based auto-encoder network as a discriminator,
upon which a generative adversarial learning is carried out to train hash
functions. Furthermore, to generate the synthetic features from the hash codes,
a compressed sensing procedure is introduced into the generative model, which
enforces the reconstruction boundary of binary codes to be consistent with that
of original features. Finally, such generative adversarial framework can be
trained via the Adam optimizer. Experimental results on four benchmarks, i.e.,
Tiny100K, GIST1M, Deep1M, and MNIST, have shown that the proposed SiGAH has
superior performance over the state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 08:07:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Liu",
"Hong",
""
]
] | TITLE: Sparse-Inductive Generative Adversarial Hashing for Nearest Neighbor
Search
ABSTRACT: Unsupervised hashing has received extensive research focus on the past
decade, which typically aims at preserving a predefined metric (i.e. Euclidean
metric) in the Hamming space. To this end, the encoding functions of the
existing hashing are typically quasi-isometric, which devote to reducing the
quantization loss from the target metric space to the discrete Hamming space.
However, it is indeed problematic to directly minimize such error, since such
mentioned two metric spaces are heterogeneous, and the quasi-isometric mapping
is non-linear. The former leads to inconsistent feature distributions, while
the latter leads to problematic optimization issues. In this paper, we propose
a novel unsupervised hashing method, termed Sparsity-Induced Generative
Adversarial Hashing (SiGAH), to encode large-scale high-dimensional features
into binary codes, which well solves the two problems through a generative
adversarial training framework. Instead of minimizing the quantization loss,
our key innovation lies in enforcing the learned Hamming space to have similar
data distribution to the target metric space via a generative model. In
particular, we formulate a ReLU-based neural network as a generator to output
binary codes and an MSE-loss based auto-encoder network as a discriminator,
upon which a generative adversarial learning is carried out to train hash
functions. Furthermore, to generate the synthetic features from the hash codes,
a compressed sensing procedure is introduced into the generative model, which
enforces the reconstruction boundary of binary codes to be consistent with that
of original features. Finally, such generative adversarial framework can be
trained via the Adam optimizer. Experimental results on four benchmarks, i.e.,
Tiny100K, GIST1M, Deep1M, and MNIST, have shown that the proposed SiGAH has
superior performance over the state-of-the-art approaches.
|
2306.06931 | Shiming Chen | Shiming Chen, Wenjin Hou, Ziming Hong, Xiaohan Ding, Yibing Song,
Xinge You, Tongliang Liu, Kun Zhang | Evolving Semantic Prototype Improves Generative Zero-Shot Learning | Accepted to ICML'23 | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In zero-shot learning (ZSL), generative methods synthesize class-related
sample features based on predefined semantic prototypes. They advance the ZSL
performance by synthesizing unseen class sample features for better training
the classifier. We observe that each class's predefined semantic prototype
(also referred to as semantic embedding or condition) does not accurately match
its real semantic prototype. So the synthesized visual sample features do not
faithfully represent the real sample features, limiting the classifier training
and existing ZSL performance. In this paper, we formulate this mismatch
phenomenon as the visual-semantic domain shift problem. We propose a dynamic
semantic prototype evolving (DSP) method to align the empirically predefined
semantic prototypes and the real prototypes for class-related feature
synthesis. The alignment is learned by refining sample features and semantic
prototypes in a unified framework and making the synthesized visual sample
features approach real sample features. After alignment, synthesized sample
features from unseen classes are closer to the real sample features and benefit
DSP to improve existing generative ZSL methods by 8.5\%, 8.0\%, and 9.7\% on
the standard CUB, SUN AWA2 datasets, the significant performance improvement
indicates that evolving semantic prototype explores a virgin field in ZSL.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 08:11:06 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Chen",
"Shiming",
""
],
[
"Hou",
"Wenjin",
""
],
[
"Hong",
"Ziming",
""
],
[
"Ding",
"Xiaohan",
""
],
[
"Song",
"Yibing",
""
],
[
"You",
"Xinge",
""
],
[
"Liu",
"Tongliang",
""
],
[
"Zhang",
"Kun",
""
]
] | TITLE: Evolving Semantic Prototype Improves Generative Zero-Shot Learning
ABSTRACT: In zero-shot learning (ZSL), generative methods synthesize class-related
sample features based on predefined semantic prototypes. They advance the ZSL
performance by synthesizing unseen class sample features for better training
the classifier. We observe that each class's predefined semantic prototype
(also referred to as semantic embedding or condition) does not accurately match
its real semantic prototype. So the synthesized visual sample features do not
faithfully represent the real sample features, limiting the classifier training
and existing ZSL performance. In this paper, we formulate this mismatch
phenomenon as the visual-semantic domain shift problem. We propose a dynamic
semantic prototype evolving (DSP) method to align the empirically predefined
semantic prototypes and the real prototypes for class-related feature
synthesis. The alignment is learned by refining sample features and semantic
prototypes in a unified framework and making the synthesized visual sample
features approach real sample features. After alignment, synthesized sample
features from unseen classes are closer to the real sample features and benefit
DSP to improve existing generative ZSL methods by 8.5\%, 8.0\%, and 9.7\% on
the standard CUB, SUN AWA2 datasets, the significant performance improvement
indicates that evolving semantic prototype explores a virgin field in ZSL.
|
2306.06934 | Weidong Qiao | Wei-Dong Qiao, Yang Xu, and Hui Li | Scale-Rotation-Equivariant Lie Group Convolution Neural Networks (Lie
Group-CNNs) | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The weight-sharing mechanism of convolutional kernels ensures
translation-equivariance of convolution neural networks (CNNs). Recently,
rotation-equivariance has been investigated. However, research on
scale-equivariance or simultaneous scale-rotation-equivariance is insufficient.
This study proposes a Lie group-CNN, which can keep scale-rotation-equivariance
for image classification tasks. The Lie group-CNN includes a lifting module, a
series of group convolution modules, a global pooling layer, and a
classification layer. The lifting module transfers the input image from
Euclidean space to Lie group space, and the group convolution is parameterized
through a fully connected network using Lie-algebra of Lie-group elements as
inputs to achieve scale-rotation-equivariance. The Lie group SIM(2) is utilized
to establish the Lie group-CNN with scale-rotation-equivariance.
Scale-rotation-equivariance of Lie group-CNN is verified and achieves the best
recognition accuracy on the blood cell dataset (97.50%) and the HAM10000
dataset (77.90%) superior to Lie algebra convolution network, dilation
convolution, spatial transformer network, and scale-equivariant steerable
network. In addition, the generalization ability of the Lie group-CNN on SIM(2)
on rotation-equivariance is verified on rotated-MNIST and rotated-CIFAR10, and
the robustness of the network is verified on SO(2) and SE(2). Therefore, the
Lie group-CNN can successfully extract geometric features and performs
equivariant recognition on images with rotation and scale transformations.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 08:14:12 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Qiao",
"Wei-Dong",
""
],
[
"Xu",
"Yang",
""
],
[
"Li",
"Hui",
""
]
] | TITLE: Scale-Rotation-Equivariant Lie Group Convolution Neural Networks (Lie
Group-CNNs)
ABSTRACT: The weight-sharing mechanism of convolutional kernels ensures
translation-equivariance of convolution neural networks (CNNs). Recently,
rotation-equivariance has been investigated. However, research on
scale-equivariance or simultaneous scale-rotation-equivariance is insufficient.
This study proposes a Lie group-CNN, which can keep scale-rotation-equivariance
for image classification tasks. The Lie group-CNN includes a lifting module, a
series of group convolution modules, a global pooling layer, and a
classification layer. The lifting module transfers the input image from
Euclidean space to Lie group space, and the group convolution is parameterized
through a fully connected network using Lie-algebra of Lie-group elements as
inputs to achieve scale-rotation-equivariance. The Lie group SIM(2) is utilized
to establish the Lie group-CNN with scale-rotation-equivariance.
Scale-rotation-equivariance of Lie group-CNN is verified and achieves the best
recognition accuracy on the blood cell dataset (97.50%) and the HAM10000
dataset (77.90%) superior to Lie algebra convolution network, dilation
convolution, spatial transformer network, and scale-equivariant steerable
network. In addition, the generalization ability of the Lie group-CNN on SIM(2)
on rotation-equivariance is verified on rotated-MNIST and rotated-CIFAR10, and
the robustness of the network is verified on SO(2) and SE(2). Therefore, the
Lie group-CNN can successfully extract geometric features and performs
equivariant recognition on images with rotation and scale transformations.
|
2306.06941 | Ana\"is Tack | Ana\"is Tack, Ekaterina Kochmar, Zheng Yuan, Serge Bibauw, Chris Piech | The BEA 2023 Shared Task on Generating AI Teacher Responses in
Educational Dialogues | to appear in the Proceedings of the 18th Workshop on Innovative Use
of NLP for Building Educational Applications, ACL 2023, Toronto, Canada | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper describes the results of the first shared task on the generation
of teacher responses in educational dialogues. The goal of the task was to
benchmark the ability of generative language models to act as AI teachers,
replying to a student in a teacher-student dialogue. Eight teams participated
in the competition hosted on CodaLab. They experimented with a wide variety of
state-of-the-art models, including Alpaca, Bloom, DialoGPT, DistilGPT-2,
Flan-T5, GPT-2, GPT-3, GPT- 4, LLaMA, OPT-2.7B, and T5-base. Their submissions
were automatically scored using BERTScore and DialogRPT metrics, and the top
three among them were further manually evaluated in terms of pedagogical
ability based on Tack and Piech (2022). The NAISTeacher system, which ranked
first in both automated and human evaluation, generated responses with GPT-3.5
using an ensemble of prompts and a DialogRPT-based ranking of responses for
given dialogue contexts. Despite the promising achievements of the
participating teams, the results also highlight the need for evaluation metrics
better suited to educational contexts.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 08:21:34 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Tack",
"Anaïs",
""
],
[
"Kochmar",
"Ekaterina",
""
],
[
"Yuan",
"Zheng",
""
],
[
"Bibauw",
"Serge",
""
],
[
"Piech",
"Chris",
""
]
] | TITLE: The BEA 2023 Shared Task on Generating AI Teacher Responses in
Educational Dialogues
ABSTRACT: This paper describes the results of the first shared task on the generation
of teacher responses in educational dialogues. The goal of the task was to
benchmark the ability of generative language models to act as AI teachers,
replying to a student in a teacher-student dialogue. Eight teams participated
in the competition hosted on CodaLab. They experimented with a wide variety of
state-of-the-art models, including Alpaca, Bloom, DialoGPT, DistilGPT-2,
Flan-T5, GPT-2, GPT-3, GPT- 4, LLaMA, OPT-2.7B, and T5-base. Their submissions
were automatically scored using BERTScore and DialogRPT metrics, and the top
three among them were further manually evaluated in terms of pedagogical
ability based on Tack and Piech (2022). The NAISTeacher system, which ranked
first in both automated and human evaluation, generated responses with GPT-3.5
using an ensemble of prompts and a DialogRPT-based ranking of responses for
given dialogue contexts. Despite the promising achievements of the
participating teams, the results also highlight the need for evaluation metrics
better suited to educational contexts.
|
2306.06949 | Nivedita Shrivastava | Nivedita Shrivastava and Smruti R. Sarangi | SecOComp: A Fast and Secure Simultaneous Compression and Encryption
Scheme | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We live in a data-driven era that involves the generation, collection and
processing of a massive amount of data. This data often contains valuable
intellectual property and sensitive user information that must be safeguarded.
There is a need to both encrypt and compress the data at line speed and
sometimes with added power constraints. The majority of the currently available
simultaneous compression and encryption (SCE) schemes are tailored for a
specific type of data such as images for instance. This reduces their generic
applicability. In this paper, we tackle this issue and propose a generic,
efficient, and secure simultaneous compression and encryption scheme where the
data is simultaneously encrypted using chaotic maps and compressed using a fast
lossless compression algorithm. We claim that employing multiple chaotic maps
and a lossless compression method can help us create not only an efficient
encryption scheme but also compress the data efficiently in a hardware-friendly
manner. We avoid all the known pitfalls of chaos theory based encryption that
have prevented its widespread usage. Our algorithm passes all the NIST tests
for nine different types of popular datasets. The proposed implementation uses
1.51x less storage as compared to the nearest computing work.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 08:33:58 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shrivastava",
"Nivedita",
""
],
[
"Sarangi",
"Smruti R.",
""
]
] | TITLE: SecOComp: A Fast and Secure Simultaneous Compression and Encryption
Scheme
ABSTRACT: We live in a data-driven era that involves the generation, collection and
processing of a massive amount of data. This data often contains valuable
intellectual property and sensitive user information that must be safeguarded.
There is a need to both encrypt and compress the data at line speed and
sometimes with added power constraints. The majority of the currently available
simultaneous compression and encryption (SCE) schemes are tailored for a
specific type of data such as images for instance. This reduces their generic
applicability. In this paper, we tackle this issue and propose a generic,
efficient, and secure simultaneous compression and encryption scheme where the
data is simultaneously encrypted using chaotic maps and compressed using a fast
lossless compression algorithm. We claim that employing multiple chaotic maps
and a lossless compression method can help us create not only an efficient
encryption scheme but also compress the data efficiently in a hardware-friendly
manner. We avoid all the known pitfalls of chaos theory based encryption that
have prevented its widespread usage. Our algorithm passes all the NIST tests
for nine different types of popular datasets. The proposed implementation uses
1.51x less storage as compared to the nearest computing work.
|
2306.06982 | Jian Wang | Jian Wang, Liang Qiao, Shichong Zhou, Jin Zhou, Jun Wang, Juncheng Li,
Shihui Ying, Cai Chang, and Jun Shi | Weakly Supervised Lesion Detection and Diagnosis for Breast Cancers with
Partially Annotated Ultrasound Images | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Deep learning (DL) has proven highly effective for ultrasound-based
computer-aided diagnosis (CAD) of breast cancers. In an automaticCAD system,
lesion detection is critical for the following diagnosis. However, existing
DL-based methods generally require voluminous manually-annotated region of
interest (ROI) labels and class labels to train both the lesion detection and
diagnosis models. In clinical practice, the ROI labels, i.e. ground truths, may
not always be optimal for the classification task due to individual experience
of sonologists, resulting in the issue of coarse annotation that limits the
diagnosis performance of a CAD model. To address this issue, a novel Two-Stage
Detection and Diagnosis Network (TSDDNet) is proposed based on weakly
supervised learning to enhance diagnostic accuracy of the ultrasound-based CAD
for breast cancers. In particular, all the ROI-level labels are considered as
coarse labels in the first training stage, and then a candidate selection
mechanism is designed to identify optimallesion areas for both the fully and
partially annotated samples. It refines the current ROI-level labels in the
fully annotated images and the detected ROIs in the partially annotated samples
with a weakly supervised manner under the guidance of class labels. In the
second training stage, a self-distillation strategy further is further proposed
to integrate the detection network and classification network into a unified
framework as the final CAD model for joint optimization, which then further
improves the diagnosis performance. The proposed TSDDNet is evaluated on a
B-mode ultrasound dataset, and the experimental results show that it achieves
the best performance on both lesion detection and diagnosis tasks, suggesting
promising application potential.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 09:26:54 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Wang",
"Jian",
""
],
[
"Qiao",
"Liang",
""
],
[
"Zhou",
"Shichong",
""
],
[
"Zhou",
"Jin",
""
],
[
"Wang",
"Jun",
""
],
[
"Li",
"Juncheng",
""
],
[
"Ying",
"Shihui",
""
],
[
"Chang",
"Cai",
""
],
[
"Shi",
"Jun",
""
]
] | TITLE: Weakly Supervised Lesion Detection and Diagnosis for Breast Cancers with
Partially Annotated Ultrasound Images
ABSTRACT: Deep learning (DL) has proven highly effective for ultrasound-based
computer-aided diagnosis (CAD) of breast cancers. In an automaticCAD system,
lesion detection is critical for the following diagnosis. However, existing
DL-based methods generally require voluminous manually-annotated region of
interest (ROI) labels and class labels to train both the lesion detection and
diagnosis models. In clinical practice, the ROI labels, i.e. ground truths, may
not always be optimal for the classification task due to individual experience
of sonologists, resulting in the issue of coarse annotation that limits the
diagnosis performance of a CAD model. To address this issue, a novel Two-Stage
Detection and Diagnosis Network (TSDDNet) is proposed based on weakly
supervised learning to enhance diagnostic accuracy of the ultrasound-based CAD
for breast cancers. In particular, all the ROI-level labels are considered as
coarse labels in the first training stage, and then a candidate selection
mechanism is designed to identify optimallesion areas for both the fully and
partially annotated samples. It refines the current ROI-level labels in the
fully annotated images and the detected ROIs in the partially annotated samples
with a weakly supervised manner under the guidance of class labels. In the
second training stage, a self-distillation strategy further is further proposed
to integrate the detection network and classification network into a unified
framework as the final CAD model for joint optimization, which then further
improves the diagnosis performance. The proposed TSDDNet is evaluated on a
B-mode ultrasound dataset, and the experimental results show that it achieves
the best performance on both lesion detection and diagnosis tasks, suggesting
promising application potential.
|
2306.06995 | Piersilvio De Bartolomeis | Piersilvio De Bartolomeis, Jacob Clarysse, Amartya Sanyal, Fanny Yang | How robust accuracy suffers from certified training with convex
relaxations | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Adversarial attacks pose significant threats to deploying state-of-the-art
classifiers in safety-critical applications. Two classes of methods have
emerged to address this issue: empirical defences and certified defences.
Although certified defences come with robustness guarantees, empirical defences
such as adversarial training enjoy much higher popularity among practitioners.
In this paper, we systematically compare the standard and robust error of these
two robust training paradigms across multiple computer vision tasks. We show
that in most tasks and for both $\mathscr{l}_\infty$-ball and
$\mathscr{l}_2$-ball threat models, certified training with convex relaxations
suffers from worse standard and robust error than adversarial training. We
further explore how the error gap between certified and adversarial training
depends on the threat model and the data distribution. In particular, besides
the perturbation budget, we identify as important factors the shape of the
perturbation set and the implicit margin of the data distribution. We support
our arguments with extensive ablations on both synthetic and image datasets.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 09:45:21 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"De Bartolomeis",
"Piersilvio",
""
],
[
"Clarysse",
"Jacob",
""
],
[
"Sanyal",
"Amartya",
""
],
[
"Yang",
"Fanny",
""
]
] | TITLE: How robust accuracy suffers from certified training with convex
relaxations
ABSTRACT: Adversarial attacks pose significant threats to deploying state-of-the-art
classifiers in safety-critical applications. Two classes of methods have
emerged to address this issue: empirical defences and certified defences.
Although certified defences come with robustness guarantees, empirical defences
such as adversarial training enjoy much higher popularity among practitioners.
In this paper, we systematically compare the standard and robust error of these
two robust training paradigms across multiple computer vision tasks. We show
that in most tasks and for both $\mathscr{l}_\infty$-ball and
$\mathscr{l}_2$-ball threat models, certified training with convex relaxations
suffers from worse standard and robust error than adversarial training. We
further explore how the error gap between certified and adversarial training
depends on the threat model and the data distribution. In particular, besides
the perturbation budget, we identify as important factors the shape of the
perturbation set and the implicit margin of the data distribution. We support
our arguments with extensive ablations on both synthetic and image datasets.
|
2306.07027 | Unai Mu\~noz-Aseguinolaza | Unai Mu\~noz-Aseguinolaza, Basilio Sierra and Naiara Aginako | Rotational augmentation techniques: a new perspective on ensemble
learning for image classification | 15 pages, 5 figures and 5 tables | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The popularity of data augmentation techniques in machine learning has
increased in recent years, as they enable the creation of new samples from
existing datasets. Rotational augmentation, in particular, has shown great
promise by revolving images and utilising them as additional data points for
training. This research study introduces a new approach to enhance the
performance of classification methods where the testing sets were generated
employing transformations on every image from the original dataset.
Subsequently, ensemble-based systems were implemented to determine the most
reliable outcome in each subset acquired from the augmentation phase to get a
final prediction for every original image. The findings of this study suggest
that rotational augmentation techniques can significantly improve the accuracy
of standard classification models; and the selection of a voting scheme can
considerably impact the model's performance. Overall, the study found that
using an ensemble-based voting system produced more accurate results than
simple voting.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 11:04:11 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Muñoz-Aseguinolaza",
"Unai",
""
],
[
"Sierra",
"Basilio",
""
],
[
"Aginako",
"Naiara",
""
]
] | TITLE: Rotational augmentation techniques: a new perspective on ensemble
learning for image classification
ABSTRACT: The popularity of data augmentation techniques in machine learning has
increased in recent years, as they enable the creation of new samples from
existing datasets. Rotational augmentation, in particular, has shown great
promise by revolving images and utilising them as additional data points for
training. This research study introduces a new approach to enhance the
performance of classification methods where the testing sets were generated
employing transformations on every image from the original dataset.
Subsequently, ensemble-based systems were implemented to determine the most
reliable outcome in each subset acquired from the augmentation phase to get a
final prediction for every original image. The findings of this study suggest
that rotational augmentation techniques can significantly improve the accuracy
of standard classification models; and the selection of a voting scheme can
considerably impact the model's performance. Overall, the study found that
using an ensemble-based voting system produced more accurate results than
simple voting.
|
2306.07032 | Lyuzhou Chen | Lyuzhou Chen, Taiyu Ban, Xiangyu Wang, Derui Lyu, Huanhuan Chen | Mitigating Prior Errors in Causal Structure Learning: Towards LLM driven
Prior Knowledge | 14 pages, 4 figures | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Causal structure learning, a prominent technique for encoding cause and
effect relationships among variables, through Bayesian Networks (BNs). Merely
recovering causal structures from real-world observed data lacks precision,
while the development of Large Language Models (LLM) is opening a new frontier
of causality. LLM presents strong capability in discovering causal
relationships between variables with the "text" inputs defining the
investigated variables, leading to a potential new hierarchy and new ladder of
causality. We aim an critical issue in the emerging topic of LLM based causal
structure learning, to tackle erroneous prior causal statements from LLM, which
is seldom considered in the current context of expert dominating prior
resources. As a pioneer attempt, we propose a BN learning strategy resilient to
prior errors without need of human intervention. Focusing on the edge-level
prior, we classify the possible prior errors into three types:
order-consistent, order-reversed, and irrelevant, and provide their theoretical
impact on the Structural Hamming Distance (SHD) under the presumption of
sufficient data. Intriguingly, we discover and prove that only the
order-reversed error contributes to an increase in a unique acyclic closed
structure, defined as a "quasi-circle". Leveraging this insight, a post-hoc
strategy is employed to identify the order-reversed prior error by its impact
on the increment of "quasi-circles". Through empirical evaluation on both real
and synthetic datasets, we demonstrate our strategy's robustness against prior
errors. Specifically, we highlight its substantial ability to resist
order-reversed errors while maintaining the majority of correct prior
knowledge.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 11:24:48 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Chen",
"Lyuzhou",
""
],
[
"Ban",
"Taiyu",
""
],
[
"Wang",
"Xiangyu",
""
],
[
"Lyu",
"Derui",
""
],
[
"Chen",
"Huanhuan",
""
]
] | TITLE: Mitigating Prior Errors in Causal Structure Learning: Towards LLM driven
Prior Knowledge
ABSTRACT: Causal structure learning, a prominent technique for encoding cause and
effect relationships among variables, through Bayesian Networks (BNs). Merely
recovering causal structures from real-world observed data lacks precision,
while the development of Large Language Models (LLM) is opening a new frontier
of causality. LLM presents strong capability in discovering causal
relationships between variables with the "text" inputs defining the
investigated variables, leading to a potential new hierarchy and new ladder of
causality. We aim an critical issue in the emerging topic of LLM based causal
structure learning, to tackle erroneous prior causal statements from LLM, which
is seldom considered in the current context of expert dominating prior
resources. As a pioneer attempt, we propose a BN learning strategy resilient to
prior errors without need of human intervention. Focusing on the edge-level
prior, we classify the possible prior errors into three types:
order-consistent, order-reversed, and irrelevant, and provide their theoretical
impact on the Structural Hamming Distance (SHD) under the presumption of
sufficient data. Intriguingly, we discover and prove that only the
order-reversed error contributes to an increase in a unique acyclic closed
structure, defined as a "quasi-circle". Leveraging this insight, a post-hoc
strategy is employed to identify the order-reversed prior error by its impact
on the increment of "quasi-circles". Through empirical evaluation on both real
and synthetic datasets, we demonstrate our strategy's robustness against prior
errors. Specifically, we highlight its substantial ability to resist
order-reversed errors while maintaining the majority of correct prior
knowledge.
|
2306.07036 | Xiaobo Xia | Yuhao Wu, Xiaobo Xia, Jun Yu, Bo Han, Gang Niu, Masashi Sugiyama,
Tongliang Liu | Making Binary Classification from Multiple Unlabeled Datasets Almost
Free of Supervision | 38 pages, 5 figures, 10 tables | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training a classifier exploiting a huge amount of supervised data is
expensive or even prohibited in a situation, where the labeling cost is high.
The remarkable progress in working with weaker forms of supervision is binary
classification from multiple unlabeled datasets which requires the knowledge of
exact class priors for all unlabeled datasets. However, the availability of
class priors is restrictive in many real-world scenarios. To address this
issue, we propose to solve a new problem setting, i.e., binary classification
from multiple unlabeled datasets with only one pairwise numerical relationship
of class priors (MU-OPPO), which knows the relative order (which unlabeled
dataset has a higher proportion of positive examples) of two class-prior
probabilities for two datasets among multiple unlabeled datasets. In MU-OPPO,
we do not need the class priors for all unlabeled datasets, but we only require
that there exists a pair of unlabeled datasets for which we know which
unlabeled dataset has a larger class prior. Clearly, this form of supervision
is easier to be obtained, which can make labeling costs almost free. We propose
a novel framework to handle the MU-OPPO problem, which consists of four
sequential modules: (i) pseudo label assignment; (ii) confident example
collection; (iii) class prior estimation; (iv) classifier training with
estimated class priors. Theoretically, we analyze the gap between estimated
class priors and true class priors under the proposed framework. Empirically,
we confirm the superiority of our framework with comprehensive experiments.
Experimental results demonstrate that our framework brings smaller estimation
errors of class priors and better performance of binary classification.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 11:33:46 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Wu",
"Yuhao",
""
],
[
"Xia",
"Xiaobo",
""
],
[
"Yu",
"Jun",
""
],
[
"Han",
"Bo",
""
],
[
"Niu",
"Gang",
""
],
[
"Sugiyama",
"Masashi",
""
],
[
"Liu",
"Tongliang",
""
]
] | TITLE: Making Binary Classification from Multiple Unlabeled Datasets Almost
Free of Supervision
ABSTRACT: Training a classifier exploiting a huge amount of supervised data is
expensive or even prohibited in a situation, where the labeling cost is high.
The remarkable progress in working with weaker forms of supervision is binary
classification from multiple unlabeled datasets which requires the knowledge of
exact class priors for all unlabeled datasets. However, the availability of
class priors is restrictive in many real-world scenarios. To address this
issue, we propose to solve a new problem setting, i.e., binary classification
from multiple unlabeled datasets with only one pairwise numerical relationship
of class priors (MU-OPPO), which knows the relative order (which unlabeled
dataset has a higher proportion of positive examples) of two class-prior
probabilities for two datasets among multiple unlabeled datasets. In MU-OPPO,
we do not need the class priors for all unlabeled datasets, but we only require
that there exists a pair of unlabeled datasets for which we know which
unlabeled dataset has a larger class prior. Clearly, this form of supervision
is easier to be obtained, which can make labeling costs almost free. We propose
a novel framework to handle the MU-OPPO problem, which consists of four
sequential modules: (i) pseudo label assignment; (ii) confident example
collection; (iii) class prior estimation; (iv) classifier training with
estimated class priors. Theoretically, we analyze the gap between estimated
class priors and true class priors under the proposed framework. Empirically,
we confirm the superiority of our framework with comprehensive experiments.
Experimental results demonstrate that our framework brings smaller estimation
errors of class priors and better performance of binary classification.
|
2306.07046 | Muhammad Arslan | Muhammad Arslan and Christophe Cruz | Imbalanced Multi-label Classification for Business-related Text with
Moderately Large Label Spaces | null | https://easychair.org/smart-program/FRCCS2023/2023-06-01.html | null | null | cs.IR cs.LG | http://creativecommons.org/licenses/by/4.0/ | In this study, we compared the performance of four different methods for
multi label text classification using a specific imbalanced business dataset.
The four methods we evaluated were fine tuned BERT, Binary Relevance,
Classifier Chains, and Label Powerset. The results show that fine tuned BERT
outperforms the other three methods by a significant margin, achieving high
values of accuracy, F1 Score, Precision, and Recall. Binary Relevance also
performs well on this dataset, while Classifier Chains and Label Powerset
demonstrate relatively poor performance. These findings highlight the
effectiveness of fine tuned BERT for multi label text classification tasks, and
suggest that it may be a useful tool for businesses seeking to analyze complex
and multifaceted texts.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 11:51:50 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Arslan",
"Muhammad",
""
],
[
"Cruz",
"Christophe",
""
]
] | TITLE: Imbalanced Multi-label Classification for Business-related Text with
Moderately Large Label Spaces
ABSTRACT: In this study, we compared the performance of four different methods for
multi label text classification using a specific imbalanced business dataset.
The four methods we evaluated were fine tuned BERT, Binary Relevance,
Classifier Chains, and Label Powerset. The results show that fine tuned BERT
outperforms the other three methods by a significant margin, achieving high
values of accuracy, F1 Score, Precision, and Recall. Binary Relevance also
performs well on this dataset, while Classifier Chains and Label Powerset
demonstrate relatively poor performance. These findings highlight the
effectiveness of fine tuned BERT for multi label text classification tasks, and
suggest that it may be a useful tool for businesses seeking to analyze complex
and multifaceted texts.
|
2306.07082 | Charalambos Konstantinou | Andres Intriago, Francesco Liberati, Nikos D. Hatziargyriou,
Charalambos Konstantinou | Residual-Based Detection of Attacks in Cyber-Physical Inverter-Based
Microgrids | Accepted at IEEE Transactions on Power Systems | null | null | null | eess.SY cs.CR cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper discusses the challenges faced by cyber-physical microgrids (MGs)
due to the inclusion of information and communication technologies in their
already complex, multi-layered systems. The work identifies a research gap in
modeling and analyzing stealthy intermittent integrity attacks in MGs, which
are designed to maximize damage and cancel secondary control objectives. To
address this, the paper proposes a nonlinear residual-based observer approach
to detect and mitigate such attacks. In order to ensure a stable operation of
the MG, the formulation then incorporates stability constraints along with the
detection observer. The proposed design is validated through case studies on a
MG benchmark with four distributed generators, demonstrating its effectiveness
in detecting attacks while satisfying network and stability constraints.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 12:48:44 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Intriago",
"Andres",
""
],
[
"Liberati",
"Francesco",
""
],
[
"Hatziargyriou",
"Nikos D.",
""
],
[
"Konstantinou",
"Charalambos",
""
]
] | TITLE: Residual-Based Detection of Attacks in Cyber-Physical Inverter-Based
Microgrids
ABSTRACT: This paper discusses the challenges faced by cyber-physical microgrids (MGs)
due to the inclusion of information and communication technologies in their
already complex, multi-layered systems. The work identifies a research gap in
modeling and analyzing stealthy intermittent integrity attacks in MGs, which
are designed to maximize damage and cancel secondary control objectives. To
address this, the paper proposes a nonlinear residual-based observer approach
to detect and mitigate such attacks. In order to ensure a stable operation of
the MG, the formulation then incorporates stability constraints along with the
detection observer. The proposed design is validated through case studies on a
MG benchmark with four distributed generators, demonstrating its effectiveness
in detecting attacks while satisfying network and stability constraints.
|
2306.07085 | Stefan Klessinger | Stefan Klessinger, Meike Klettke, Uta St\"orl, Stefanie Scherzinger | Extracting JSON Schemas with Tagged Unions | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With data lakes and schema-free NoSQL document stores, extracting a
descriptive schema from JSON data collections is an acute challenge. In this
paper, we target the discovery of tagged unions, a JSON Schema design pattern
where the value of one property of an object (the tag) conditionally implies
subschemas for sibling properties. We formalize these implications as
conditional functional dependencies and capture them using the JSON Schema
operators if-then-else. We further motivate our heuristics to avoid
overfitting. Experiments with our prototype implementation are promising, and
show that this form of tagged unions can successfully be detected in real-world
GeoJSON and TopoJSON datasets. In discussing future work, we outline how our
approach can be extended further.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 12:57:34 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Klessinger",
"Stefan",
""
],
[
"Klettke",
"Meike",
""
],
[
"Störl",
"Uta",
""
],
[
"Scherzinger",
"Stefanie",
""
]
] | TITLE: Extracting JSON Schemas with Tagged Unions
ABSTRACT: With data lakes and schema-free NoSQL document stores, extracting a
descriptive schema from JSON data collections is an acute challenge. In this
paper, we target the discovery of tagged unions, a JSON Schema design pattern
where the value of one property of an object (the tag) conditionally implies
subschemas for sibling properties. We formalize these implications as
conditional functional dependencies and capture them using the JSON Schema
operators if-then-else. We further motivate our heuristics to avoid
overfitting. Experiments with our prototype implementation are promising, and
show that this form of tagged unions can successfully be detected in real-world
GeoJSON and TopoJSON datasets. In discussing future work, we outline how our
approach can be extended further.
|
2306.07098 | Maxwell Jones | Dravyansh Sharma, Maxwell Jones | Efficiently Learning the Graph for Semi-supervised Learning | 29 pages, 9 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Computational efficiency is a major bottleneck in using classic graph-based
approaches for semi-supervised learning on datasets with a large number of
unlabeled examples. Known techniques to improve efficiency typically involve an
approximation of the graph regularization objective, but suffer two major
drawbacks - first the graph is assumed to be known or constructed with
heuristic hyperparameter values, second they do not provide a principled
approximation guarantee for learning over the full unlabeled dataset. Building
on recent work on learning graphs for semi-supervised learning from multiple
datasets for problems from the same domain, and leveraging techniques for fast
approximations for solving linear systems in the graph Laplacian matrix, we
propose algorithms that overcome both the above limitations.
We show a formal separation in the learning-theoretic complexity of sparse
and dense graph families. We further show how to approximately learn the best
graphs from the sparse families efficiently using the conjugate gradient
method.
Our approach can also be used to learn the graph efficiently online with
sub-linear regret, under mild smoothness assumptions. Our online learning
results are stated generally, and may be useful for approximate and efficient
parameter tuning in other problems. We implement our approach and demonstrate
significant ($\sim$10-100x) speedups over prior work on semi-supervised
learning with learned graphs on benchmark datasets.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 13:22:06 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Sharma",
"Dravyansh",
""
],
[
"Jones",
"Maxwell",
""
]
] | TITLE: Efficiently Learning the Graph for Semi-supervised Learning
ABSTRACT: Computational efficiency is a major bottleneck in using classic graph-based
approaches for semi-supervised learning on datasets with a large number of
unlabeled examples. Known techniques to improve efficiency typically involve an
approximation of the graph regularization objective, but suffer two major
drawbacks - first the graph is assumed to be known or constructed with
heuristic hyperparameter values, second they do not provide a principled
approximation guarantee for learning over the full unlabeled dataset. Building
on recent work on learning graphs for semi-supervised learning from multiple
datasets for problems from the same domain, and leveraging techniques for fast
approximations for solving linear systems in the graph Laplacian matrix, we
propose algorithms that overcome both the above limitations.
We show a formal separation in the learning-theoretic complexity of sparse
and dense graph families. We further show how to approximately learn the best
graphs from the sparse families efficiently using the conjugate gradient
method.
Our approach can also be used to learn the graph efficiently online with
sub-linear regret, under mild smoothness assumptions. Our online learning
results are stated generally, and may be useful for approximate and efficient
parameter tuning in other problems. We implement our approach and demonstrate
significant ($\sim$10-100x) speedups over prior work on semi-supervised
learning with learned graphs on benchmark datasets.
|
2306.07114 | Shuo Yu | Feng Xia, Xin Chen, Shuo Yu, Mingliang Hou, Mujie Liu, Linlin You | Coupled Attention Networks for Multivariate Time Series Anomaly
Detection | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multivariate time series anomaly detection (MTAD) plays a vital role in a
wide variety of real-world application domains. Over the past few years, MTAD
has attracted rapidly increasing attention from both academia and industry.
Many deep learning and graph learning models have been developed for effective
anomaly detection in multivariate time series data, which enable advanced
applications such as smart surveillance and risk management with unprecedented
capabilities. Nevertheless, MTAD is facing critical challenges deriving from
the dependencies among sensors and variables, which often change over time. To
address this issue, we propose a coupled attention-based neural network
framework (CAN) for anomaly detection in multivariate time series data
featuring dynamic variable relationships. We combine adaptive graph learning
methods with graph attention to generate a global-local graph that can
represent both global correlations and dynamic local correlations among
sensors. To capture inter-sensor relationships and temporal dependencies, a
convolutional neural network based on the global-local graph is integrated with
a temporal self-attention module to construct a coupled attention module. In
addition, we develop a multilevel encoder-decoder architecture that
accommodates reconstruction and prediction tasks to better characterize
multivariate time series data. Extensive experiments on real-world datasets
have been conducted to evaluate the performance of the proposed CAN approach,
and the results show that CAN significantly outperforms state-of-the-art
baselines.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 13:42:56 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Xia",
"Feng",
""
],
[
"Chen",
"Xin",
""
],
[
"Yu",
"Shuo",
""
],
[
"Hou",
"Mingliang",
""
],
[
"Liu",
"Mujie",
""
],
[
"You",
"Linlin",
""
]
] | TITLE: Coupled Attention Networks for Multivariate Time Series Anomaly
Detection
ABSTRACT: Multivariate time series anomaly detection (MTAD) plays a vital role in a
wide variety of real-world application domains. Over the past few years, MTAD
has attracted rapidly increasing attention from both academia and industry.
Many deep learning and graph learning models have been developed for effective
anomaly detection in multivariate time series data, which enable advanced
applications such as smart surveillance and risk management with unprecedented
capabilities. Nevertheless, MTAD is facing critical challenges deriving from
the dependencies among sensors and variables, which often change over time. To
address this issue, we propose a coupled attention-based neural network
framework (CAN) for anomaly detection in multivariate time series data
featuring dynamic variable relationships. We combine adaptive graph learning
methods with graph attention to generate a global-local graph that can
represent both global correlations and dynamic local correlations among
sensors. To capture inter-sensor relationships and temporal dependencies, a
convolutional neural network based on the global-local graph is integrated with
a temporal self-attention module to construct a coupled attention module. In
addition, we develop a multilevel encoder-decoder architecture that
accommodates reconstruction and prediction tasks to better characterize
multivariate time series data. Extensive experiments on real-world datasets
have been conducted to evaluate the performance of the proposed CAN approach,
and the results show that CAN significantly outperforms state-of-the-art
baselines.
|
2306.07115 | Th\'eo Deschamps-Berger | Th\'eo Deschamps-Berger, Lori Lamel, Laurence Devillers | Exploring Attention Mechanisms for Multimodal Emotion Recognition in an
Emergency Call Center Corpus | 5 pages, 2 figures, 4 tables | Published in ICASSP 2023 - 2023 IEEE International Conference on
Acoustics, Speech and Signal Processing (ICASSP) | 10.1109/ICASSP49357.2023.10096112 | null | cs.CL cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The emotion detection technology to enhance human decision-making is an
important research issue for real-world applications, but real-life emotion
datasets are relatively rare and small. The experiments conducted in this paper
use the CEMO, which was collected in a French emergency call center. Two
pre-trained models based on speech and text were fine-tuned for speech emotion
recognition. Using pre-trained Transformer encoders mitigates our data's
limited and sparse nature. This paper explores the different fusion strategies
of these modality-specific models. In particular, fusions with and without
cross-attention mechanisms were tested to gather the most relevant information
from both the speech and text encoders. We show that multimodal fusion brings
an absolute gain of 4-9% with respect to either single modality and that the
Symmetric multi-headed cross-attention mechanism performed better than late
classical fusion approaches. Our experiments also suggest that for the
real-life CEMO corpus, the audio component encodes more emotive information
than the textual one.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 13:43:20 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Deschamps-Berger",
"Théo",
""
],
[
"Lamel",
"Lori",
""
],
[
"Devillers",
"Laurence",
""
]
] | TITLE: Exploring Attention Mechanisms for Multimodal Emotion Recognition in an
Emergency Call Center Corpus
ABSTRACT: The emotion detection technology to enhance human decision-making is an
important research issue for real-world applications, but real-life emotion
datasets are relatively rare and small. The experiments conducted in this paper
use the CEMO, which was collected in a French emergency call center. Two
pre-trained models based on speech and text were fine-tuned for speech emotion
recognition. Using pre-trained Transformer encoders mitigates our data's
limited and sparse nature. This paper explores the different fusion strategies
of these modality-specific models. In particular, fusions with and without
cross-attention mechanisms were tested to gather the most relevant information
from both the speech and text encoders. We show that multimodal fusion brings
an absolute gain of 4-9% with respect to either single modality and that the
Symmetric multi-headed cross-attention mechanism performed better than late
classical fusion approaches. Our experiments also suggest that for the
real-life CEMO corpus, the audio component encodes more emotive information
than the textual one.
|
2306.07118 | Ibrahim Abualhaol Dr. | Masoud Mehrabi Koushki, Ibrahim AbuAlhaol, Anandharaju Durai Raju,
Yang Zhou, Ronnie Salvador Giagone and Huang Shengqiang | On building machine learning pipelines for Android malware detection: a
procedural survey of practices, challenges and opportunities | file:///C:/Users/ibrahim_abualhaol/Downloads/s42400-022-00119-8.pdf | SpringerOpen Cybersecurity 2022 | null | null | cs.CR cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | As the smartphone market leader, Android has been a prominent target for
malware attacks. The number of malicious applications (apps) identified for it
has increased continually over the past decade, creating an immense challenge
for all parties involved. For market holders and researchers, in particular,
the large number of samples has made manual malware detection unfeasible,
leading to an influx of research that investigate Machine Learning (ML)
approaches to automate this process. However, while some of the proposed
approaches achieve high performance, rapidly evolving Android malware has made
them unable to maintain their accuracy over time. This has created a need in
the community to conduct further research, and build more flexible ML
pipelines. Doing so, however, is currently hindered by a lack of systematic
overview of the existing literature, to learn from and improve upon the
existing solutions. Existing survey papers often focus only on parts of the ML
process (e.g., data collection or model deployment), while omitting other
important stages, such as model evaluation and explanation. In this paper, we
address this problem with a review of 42 highly-cited papers, spanning a decade
of research (from 2011 to 2021). We introduce a novel procedural taxonomy of
the published literature, covering how they have used ML algorithms, what
features they have engineered, which dimensionality reduction techniques they
have employed, what datasets they have employed for training, and what their
evaluation and explanation strategies are. Drawing from this taxonomy, we also
identify gaps in knowledge and provide ideas for improvement and future work.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 13:52:28 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Koushki",
"Masoud Mehrabi",
""
],
[
"AbuAlhaol",
"Ibrahim",
""
],
[
"Raju",
"Anandharaju Durai",
""
],
[
"Zhou",
"Yang",
""
],
[
"Giagone",
"Ronnie Salvador",
""
],
[
"Shengqiang",
"Huang",
""
]
] | TITLE: On building machine learning pipelines for Android malware detection: a
procedural survey of practices, challenges and opportunities
ABSTRACT: As the smartphone market leader, Android has been a prominent target for
malware attacks. The number of malicious applications (apps) identified for it
has increased continually over the past decade, creating an immense challenge
for all parties involved. For market holders and researchers, in particular,
the large number of samples has made manual malware detection unfeasible,
leading to an influx of research that investigate Machine Learning (ML)
approaches to automate this process. However, while some of the proposed
approaches achieve high performance, rapidly evolving Android malware has made
them unable to maintain their accuracy over time. This has created a need in
the community to conduct further research, and build more flexible ML
pipelines. Doing so, however, is currently hindered by a lack of systematic
overview of the existing literature, to learn from and improve upon the
existing solutions. Existing survey papers often focus only on parts of the ML
process (e.g., data collection or model deployment), while omitting other
important stages, such as model evaluation and explanation. In this paper, we
address this problem with a review of 42 highly-cited papers, spanning a decade
of research (from 2011 to 2021). We introduce a novel procedural taxonomy of
the published literature, covering how they have used ML algorithms, what
features they have engineered, which dimensionality reduction techniques they
have employed, what datasets they have employed for training, and what their
evaluation and explanation strategies are. Drawing from this taxonomy, we also
identify gaps in knowledge and provide ideas for improvement and future work.
|
2306.07154 | Jiale Xu | Jiale Xu, Xintao Wang, Yan-Pei Cao, Weihao Cheng, Ying Shan, Shenghua
Gao | InstructP2P: Learning to Edit 3D Point Clouds with Text Instructions | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Enhancing AI systems to perform tasks following human instructions can
significantly boost productivity. In this paper, we present InstructP2P, an
end-to-end framework for 3D shape editing on point clouds, guided by high-level
textual instructions. InstructP2P extends the capabilities of existing methods
by synergizing the strengths of a text-conditioned point cloud diffusion model,
Point-E, and powerful language models, enabling color and geometry editing
using language instructions. To train InstructP2P, we introduce a new shape
editing dataset, constructed by integrating a shape segmentation dataset,
off-the-shelf shape programs, and diverse edit instructions generated by a
large language model, ChatGPT. Our proposed method allows for editing both
color and geometry of specific regions in a single forward pass, while leaving
other regions unaffected. In our experiments, InstructP2P shows generalization
capabilities, adapting to novel shape categories and instructions, despite
being trained on a limited amount of data.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 14:42:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Xu",
"Jiale",
""
],
[
"Wang",
"Xintao",
""
],
[
"Cao",
"Yan-Pei",
""
],
[
"Cheng",
"Weihao",
""
],
[
"Shan",
"Ying",
""
],
[
"Gao",
"Shenghua",
""
]
] | TITLE: InstructP2P: Learning to Edit 3D Point Clouds with Text Instructions
ABSTRACT: Enhancing AI systems to perform tasks following human instructions can
significantly boost productivity. In this paper, we present InstructP2P, an
end-to-end framework for 3D shape editing on point clouds, guided by high-level
textual instructions. InstructP2P extends the capabilities of existing methods
by synergizing the strengths of a text-conditioned point cloud diffusion model,
Point-E, and powerful language models, enabling color and geometry editing
using language instructions. To train InstructP2P, we introduce a new shape
editing dataset, constructed by integrating a shape segmentation dataset,
off-the-shelf shape programs, and diverse edit instructions generated by a
large language model, ChatGPT. Our proposed method allows for editing both
color and geometry of specific regions in a single forward pass, while leaving
other regions unaffected. In our experiments, InstructP2P shows generalization
capabilities, adapting to novel shape categories and instructions, despite
being trained on a limited amount of data.
|
2306.07160 | Alec Reed | Alec Reed, Christoffer Heckman | Looking Around Corners: Generative Methods in Terrain Extension | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we provide an early look at our model for generating terrain
that is occluded in the initial lidar scan or out of range of the sensor. As a
proof of concept, we show that a transformer based framework is able to be
overfit to predict the geometries of unobserved roads around intersections or
corners. We discuss our method for generating training data, as well as a
unique loss function for training our terrain extension network. The framework
is tested on data from the SemanticKitti [1] dataset. Unlabeled point clouds
measured from an onboard lidar are used as input data to generate predicted
road points that are out of range or occluded in the original point-cloud scan.
Then the input pointcloud and predicted terrain are concatenated to the
terrain-extended pointcloud. We show promising qualitative results from these
methods, as well as discussion for potential quantitative metrics to evaluate
the overall success of our framework. Finally, we discuss improvements that can
be made to the framework for successful generalization to test sets.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 14:48:32 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Reed",
"Alec",
""
],
[
"Heckman",
"Christoffer",
""
]
] | TITLE: Looking Around Corners: Generative Methods in Terrain Extension
ABSTRACT: In this paper, we provide an early look at our model for generating terrain
that is occluded in the initial lidar scan or out of range of the sensor. As a
proof of concept, we show that a transformer based framework is able to be
overfit to predict the geometries of unobserved roads around intersections or
corners. We discuss our method for generating training data, as well as a
unique loss function for training our terrain extension network. The framework
is tested on data from the SemanticKitti [1] dataset. Unlabeled point clouds
measured from an onboard lidar are used as input data to generate predicted
road points that are out of range or occluded in the original point-cloud scan.
Then the input pointcloud and predicted terrain are concatenated to the
terrain-extended pointcloud. We show promising qualitative results from these
methods, as well as discussion for potential quantitative metrics to evaluate
the overall success of our framework. Finally, we discuss improvements that can
be made to the framework for successful generalization to test sets.
|
2306.07165 | Abdullah Alharthi | Abdullah Alharthi | Explainable AI and Machine Learning Towards Human Gait Deterioration
Analysis | Book Chapter | null | null | null | eess.SP cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | Gait analysis, an expanding research area, employs non invasive sensors and
machine learning techniques for a range of applicatio ns. In this study, we
concentrate on gait analysis for detecting cognitive decline in Parkinson's
disease (PD) and under dual task conditions. Using convolutional neural
networks (CNNs) and explainable machine learning, we objectively analyze gait
data and associate findings with clinically relevant biomarkers. This is
accomplished by connecting machine learning outputs to decisions based on human
visual observations or derived quantitative gait parameters, which are tested
and routinely implemented in curr ent healthcare practice. Our analysis of gait
deterioration due to cognitive decline in PD enables robust results using the
proposed methods for assessing PD severity from ground reaction force (GRF)
data. We achieved classification accuracies of 98% F1 sc ores for each
PhysioNet.org dataset and 95.5% F1 scores for the combined PhysioNet dataset.
By linking clinically observable features to the model outputs, we demonstrate
the impact of PD severity on gait. Furthermore, we explore the significance of
cognit ive load in healthy gait analysis, resulting in robust classification
accuracies of 100% F1 scores for subject identity verification. We also
identify weaker features crucial for model predictions using Layer Wise
Relevance Propagation. A notable finding o f this study reveals that cognitive
deterioration's effect on gait influences body balance and foot landing/lifting
dynamics in both classification cases: cognitive load in healthy gait and
cognitive decline in PD gait.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 14:53:00 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Alharthi",
"Abdullah",
""
]
] | TITLE: Explainable AI and Machine Learning Towards Human Gait Deterioration
Analysis
ABSTRACT: Gait analysis, an expanding research area, employs non invasive sensors and
machine learning techniques for a range of applicatio ns. In this study, we
concentrate on gait analysis for detecting cognitive decline in Parkinson's
disease (PD) and under dual task conditions. Using convolutional neural
networks (CNNs) and explainable machine learning, we objectively analyze gait
data and associate findings with clinically relevant biomarkers. This is
accomplished by connecting machine learning outputs to decisions based on human
visual observations or derived quantitative gait parameters, which are tested
and routinely implemented in curr ent healthcare practice. Our analysis of gait
deterioration due to cognitive decline in PD enables robust results using the
proposed methods for assessing PD severity from ground reaction force (GRF)
data. We achieved classification accuracies of 98% F1 sc ores for each
PhysioNet.org dataset and 95.5% F1 scores for the combined PhysioNet dataset.
By linking clinically observable features to the model outputs, we demonstrate
the impact of PD severity on gait. Furthermore, we explore the significance of
cognit ive load in healthy gait analysis, resulting in robust classification
accuracies of 100% F1 scores for subject identity verification. We also
identify weaker features crucial for model predictions using Layer Wise
Relevance Propagation. A notable finding o f this study reveals that cognitive
deterioration's effect on gait influences body balance and foot landing/lifting
dynamics in both classification cases: cognitive load in healthy gait and
cognitive decline in PD gait.
|
2306.07171 | Jinfei Liu | Xiang Li and Haocheng Xia and Jinfei Liu | Shapley Value on Probabilistic Classifiers | null | null | null | null | cs.LG cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data valuation has become an increasingly significant discipline in data
science due to the economic value of data. In the context of machine learning
(ML), data valuation methods aim to equitably measure the contribution of each
data point to the utility of an ML model. One prevalent method is Shapley
value, which helps identify data points that are beneficial or detrimental to
an ML model. However, traditional Shapley-based data valuation methods may not
effectively distinguish between beneficial and detrimental training data points
for probabilistic classifiers. In this paper, we propose Probabilistic Shapley
(P-Shapley) value by constructing a probability-wise utility function that
leverages the predicted class probabilities of probabilistic classifiers rather
than binarized prediction results in the traditional Shapley value. We also
offer several activation functions for confidence calibration to effectively
quantify the marginal contribution of each data point to the probabilistic
classifiers. Extensive experiments on four real-world datasets demonstrate the
effectiveness of our proposed P-Shapley value in evaluating the importance of
data for building a high-usability and trustworthy ML model.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:09:13 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Li",
"Xiang",
""
],
[
"Xia",
"Haocheng",
""
],
[
"Liu",
"Jinfei",
""
]
] | TITLE: Shapley Value on Probabilistic Classifiers
ABSTRACT: Data valuation has become an increasingly significant discipline in data
science due to the economic value of data. In the context of machine learning
(ML), data valuation methods aim to equitably measure the contribution of each
data point to the utility of an ML model. One prevalent method is Shapley
value, which helps identify data points that are beneficial or detrimental to
an ML model. However, traditional Shapley-based data valuation methods may not
effectively distinguish between beneficial and detrimental training data points
for probabilistic classifiers. In this paper, we propose Probabilistic Shapley
(P-Shapley) value by constructing a probability-wise utility function that
leverages the predicted class probabilities of probabilistic classifiers rather
than binarized prediction results in the traditional Shapley value. We also
offer several activation functions for confidence calibration to effectively
quantify the marginal contribution of each data point to the probabilistic
classifiers. Extensive experiments on four real-world datasets demonstrate the
effectiveness of our proposed P-Shapley value in evaluating the importance of
data for building a high-usability and trustworthy ML model.
|
2306.07174 | Weizhi Wang | Weizhi Wang, Li Dong, Hao Cheng, Xiaodong Liu, Xifeng Yan, Jianfeng
Gao, Furu Wei | Augmenting Language Models with Long-Term Memory | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing large language models (LLMs) can only afford fix-sized inputs due to
the input length limit, preventing them from utilizing rich long-context
information from past inputs. To address this, we propose a framework, Language
Models Augmented with Long-Term Memory (LongMem), which enables LLMs to
memorize long history. We design a novel decoupled network architecture with
the original backbone LLM frozen as a memory encoder and an adaptive residual
side-network as a memory retriever and reader. Such a decoupled memory design
can easily cache and update long-term past contexts for memory retrieval
without suffering from memory staleness. Enhanced with memory-augmented
adaptation training, LongMem can thus memorize long past context and use
long-term memory for language modeling. The proposed memory retrieval module
can handle unlimited-length context in its memory bank to benefit various
downstream tasks. Typically, LongMem can enlarge the long-form memory to 65k
tokens and thus cache many-shot extra demonstration examples as long-form
memory for in-context learning. Experiments show that our method outperforms
strong long-context models on ChapterBreak, a challenging long-context modeling
benchmark, and achieves remarkable improvements on memory-augmented in-context
learning over LLMs. The results demonstrate that the proposed method is
effective in helping language models to memorize and utilize long-form
contents. Our code is open-sourced at https://aka.ms/LongMem.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:13:39 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Wang",
"Weizhi",
""
],
[
"Dong",
"Li",
""
],
[
"Cheng",
"Hao",
""
],
[
"Liu",
"Xiaodong",
""
],
[
"Yan",
"Xifeng",
""
],
[
"Gao",
"Jianfeng",
""
],
[
"Wei",
"Furu",
""
]
] | TITLE: Augmenting Language Models with Long-Term Memory
ABSTRACT: Existing large language models (LLMs) can only afford fix-sized inputs due to
the input length limit, preventing them from utilizing rich long-context
information from past inputs. To address this, we propose a framework, Language
Models Augmented with Long-Term Memory (LongMem), which enables LLMs to
memorize long history. We design a novel decoupled network architecture with
the original backbone LLM frozen as a memory encoder and an adaptive residual
side-network as a memory retriever and reader. Such a decoupled memory design
can easily cache and update long-term past contexts for memory retrieval
without suffering from memory staleness. Enhanced with memory-augmented
adaptation training, LongMem can thus memorize long past context and use
long-term memory for language modeling. The proposed memory retrieval module
can handle unlimited-length context in its memory bank to benefit various
downstream tasks. Typically, LongMem can enlarge the long-form memory to 65k
tokens and thus cache many-shot extra demonstration examples as long-form
memory for in-context learning. Experiments show that our method outperforms
strong long-context models on ChapterBreak, a challenging long-context modeling
benchmark, and achieves remarkable improvements on memory-augmented in-context
learning over LLMs. The results demonstrate that the proposed method is
effective in helping language models to memorize and utilize long-form
contents. Our code is open-sourced at https://aka.ms/LongMem.
|
2306.07178 | Harshitha Machiraju | Harshitha Machiraju, Michael H. Herzog, Pascal Frossard | Frequency-Based Vulnerability Analysis of Deep Learning Models against
Image Corruptions | Code: https://github.com/code-Assasin/MUFIACode | null | null | null | cs.CV cs.AI cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning models often face challenges when handling real-world image
corruptions. In response, researchers have developed image corruption datasets
to evaluate the performance of deep neural networks in handling such
corruptions. However, these datasets have a significant limitation: they do not
account for all corruptions encountered in real-life scenarios. To address this
gap, we present MUFIA (Multiplicative Filter Attack), an algorithm designed to
identify the specific types of corruptions that can cause models to fail. Our
algorithm identifies the combination of image frequency components that render
a model susceptible to misclassification while preserving the semantic
similarity to the original image. We find that even state-of-the-art models
trained to be robust against known common corruptions struggle against the low
visibility-based corruptions crafted by MUFIA. This highlights the need for
more comprehensive approaches to enhance model robustness against a wider range
of real-world image corruptions.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:19:13 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Machiraju",
"Harshitha",
""
],
[
"Herzog",
"Michael H.",
""
],
[
"Frossard",
"Pascal",
""
]
] | TITLE: Frequency-Based Vulnerability Analysis of Deep Learning Models against
Image Corruptions
ABSTRACT: Deep learning models often face challenges when handling real-world image
corruptions. In response, researchers have developed image corruption datasets
to evaluate the performance of deep neural networks in handling such
corruptions. However, these datasets have a significant limitation: they do not
account for all corruptions encountered in real-life scenarios. To address this
gap, we present MUFIA (Multiplicative Filter Attack), an algorithm designed to
identify the specific types of corruptions that can cause models to fail. Our
algorithm identifies the combination of image frequency components that render
a model susceptible to misclassification while preserving the semantic
similarity to the original image. We find that even state-of-the-art models
trained to be robust against known common corruptions struggle against the low
visibility-based corruptions crafted by MUFIA. This highlights the need for
more comprehensive approaches to enhance model robustness against a wider range
of real-world image corruptions.
|
2306.07179 | Frank Schneider | George E. Dahl, Frank Schneider, Zachary Nado, Naman Agarwal,
Chandramouli Shama Sastry, Philipp Hennig, Sourabh Medapati, Runa
Eschenhagen, Priya Kasimbeg, Daniel Suo, Juhan Bae, Justin Gilmer, Abel L.
Peirson, Bilal Khan, Rohan Anil, Mike Rabbat, Shankar Krishnan, Daniel
Snider, Ehsan Amid, Kongtao Chen, Chris J. Maddison, Rakshith Vasudev, Michal
Badura, Ankush Garg, Peter Mattson | Benchmarking Neural Network Training Algorithms | 102 pages, 8 figures, 41 tables | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Training algorithms, broadly construed, are an essential part of every deep
learning pipeline. Training algorithm improvements that speed up training
across a wide variety of workloads (e.g., better update rules, tuning
protocols, learning rate schedules, or data selection schemes) could save time,
save computational resources, and lead to better, more accurate, models.
Unfortunately, as a community, we are currently unable to reliably identify
training algorithm improvements, or even determine the state-of-the-art
training algorithm. In this work, using concrete experiments, we argue that
real progress in speeding up training requires new benchmarks that resolve
three basic challenges faced by empirical comparisons of training algorithms:
(1) how to decide when training is complete and precisely measure training
time, (2) how to handle the sensitivity of measurements to exact workload
details, and (3) how to fairly compare algorithms that require hyperparameter
tuning. In order to address these challenges, we introduce a new, competitive,
time-to-result benchmark using multiple workloads running on fixed hardware,
the AlgoPerf: Training Algorithms benchmark. Our benchmark includes a set of
workload variants that make it possible to detect benchmark submissions that
are more robust to workload changes than current widely-used methods. Finally,
we evaluate baseline submissions constructed using various optimizers that
represent current practice, as well as other optimizers that have recently
received attention in the literature. These baseline results collectively
demonstrate the feasibility of our benchmark, show that non-trivial gaps
between methods exist, and set a provisional state-of-the-art for future
benchmark submissions to try and surpass.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:21:02 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Dahl",
"George E.",
""
],
[
"Schneider",
"Frank",
""
],
[
"Nado",
"Zachary",
""
],
[
"Agarwal",
"Naman",
""
],
[
"Sastry",
"Chandramouli Shama",
""
],
[
"Hennig",
"Philipp",
""
],
[
"Medapati",
"Sourabh",
""
],
[
"Eschenhagen",
"Runa",
""
],
[
"Kasimbeg",
"Priya",
""
],
[
"Suo",
"Daniel",
""
],
[
"Bae",
"Juhan",
""
],
[
"Gilmer",
"Justin",
""
],
[
"Peirson",
"Abel L.",
""
],
[
"Khan",
"Bilal",
""
],
[
"Anil",
"Rohan",
""
],
[
"Rabbat",
"Mike",
""
],
[
"Krishnan",
"Shankar",
""
],
[
"Snider",
"Daniel",
""
],
[
"Amid",
"Ehsan",
""
],
[
"Chen",
"Kongtao",
""
],
[
"Maddison",
"Chris J.",
""
],
[
"Vasudev",
"Rakshith",
""
],
[
"Badura",
"Michal",
""
],
[
"Garg",
"Ankush",
""
],
[
"Mattson",
"Peter",
""
]
] | TITLE: Benchmarking Neural Network Training Algorithms
ABSTRACT: Training algorithms, broadly construed, are an essential part of every deep
learning pipeline. Training algorithm improvements that speed up training
across a wide variety of workloads (e.g., better update rules, tuning
protocols, learning rate schedules, or data selection schemes) could save time,
save computational resources, and lead to better, more accurate, models.
Unfortunately, as a community, we are currently unable to reliably identify
training algorithm improvements, or even determine the state-of-the-art
training algorithm. In this work, using concrete experiments, we argue that
real progress in speeding up training requires new benchmarks that resolve
three basic challenges faced by empirical comparisons of training algorithms:
(1) how to decide when training is complete and precisely measure training
time, (2) how to handle the sensitivity of measurements to exact workload
details, and (3) how to fairly compare algorithms that require hyperparameter
tuning. In order to address these challenges, we introduce a new, competitive,
time-to-result benchmark using multiple workloads running on fixed hardware,
the AlgoPerf: Training Algorithms benchmark. Our benchmark includes a set of
workload variants that make it possible to detect benchmark submissions that
are more robust to workload changes than current widely-used methods. Finally,
we evaluate baseline submissions constructed using various optimizers that
represent current practice, as well as other optimizers that have recently
received attention in the literature. These baseline results collectively
demonstrate the feasibility of our benchmark, show that non-trivial gaps
between methods exist, and set a provisional state-of-the-art for future
benchmark submissions to try and surpass.
|
2306.07193 | Yue Yu | Ran Xu, Yue Yu, Joyce C. Ho, Carl Yang | Weakly-Supervised Scientific Document Classification via
Retrieval-Augmented Multi-Stage Training | SIGIR 2023. The code and data will be published to
https://github.com/ritaranx/wander | The 46th International ACM SIGIR Conference on Research and
Development in Information Retrieval (SIGIR 2023) | 10.1145/3539618.3592085 | null | cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scientific document classification is a critical task for a wide range of
applications, but the cost of obtaining massive amounts of human-labeled data
can be prohibitive. To address this challenge, we propose a weakly-supervised
approach for scientific document classification using label names only. In
scientific domains, label names often include domain-specific concepts that may
not appear in the document corpus, making it difficult to match labels and
documents precisely. To tackle this issue, we propose WANDER, which leverages
dense retrieval to perform matching in the embedding space to capture the
semantics of label names. We further design the label name expansion module to
enrich the label name representations. Lastly, a self-training step is used to
refine the predictions. The experiments on three datasets show that WANDER
outperforms the best baseline by 11.9% on average. Our code will be published
at https://github.com/ritaranx/wander.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:50:13 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Xu",
"Ran",
""
],
[
"Yu",
"Yue",
""
],
[
"Ho",
"Joyce C.",
""
],
[
"Yang",
"Carl",
""
]
] | TITLE: Weakly-Supervised Scientific Document Classification via
Retrieval-Augmented Multi-Stage Training
ABSTRACT: Scientific document classification is a critical task for a wide range of
applications, but the cost of obtaining massive amounts of human-labeled data
can be prohibitive. To address this challenge, we propose a weakly-supervised
approach for scientific document classification using label names only. In
scientific domains, label names often include domain-specific concepts that may
not appear in the document corpus, making it difficult to match labels and
documents precisely. To tackle this issue, we propose WANDER, which leverages
dense retrieval to perform matching in the embedding space to capture the
semantics of label names. We further design the label name expansion module to
enrich the label name representations. Lastly, a self-training step is used to
refine the predictions. The experiments on three datasets show that WANDER
outperforms the best baseline by 11.9% on average. Our code will be published
at https://github.com/ritaranx/wander.
|
2306.07198 | Jeremy Gwinnup | Jeremy Gwinnup and Kevin Duh | A Survey of Vision-Language Pre-training from the Lens of Multimodal
Machine Translation | 10 pages | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large language models such as BERT and the GPT series started a paradigm
shift that calls for building general-purpose models via pre-training on large
datasets, followed by fine-tuning on task-specific datasets. There is now a
plethora of large pre-trained models for Natural Language Processing and
Computer Vision. Recently, we have seen rapid developments in the joint
Vision-Language space as well, where pre-trained models such as CLIP (Radford
et al., 2021) have demonstrated improvements in downstream tasks like image
captioning and visual question answering. However, surprisingly there is
comparatively little work on exploring these models for the task of multimodal
machine translation, where the goal is to leverage image/video modality in
text-to-text translation. To fill this gap, this paper surveys the landscape of
language-and-vision pre-training from the lens of multimodal machine
translation. We summarize the common architectures, pre-training objectives,
and datasets from literature and conjecture what further is needed to make
progress on multimodal machine translation.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 15:56:10 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Gwinnup",
"Jeremy",
""
],
[
"Duh",
"Kevin",
""
]
] | TITLE: A Survey of Vision-Language Pre-training from the Lens of Multimodal
Machine Translation
ABSTRACT: Large language models such as BERT and the GPT series started a paradigm
shift that calls for building general-purpose models via pre-training on large
datasets, followed by fine-tuning on task-specific datasets. There is now a
plethora of large pre-trained models for Natural Language Processing and
Computer Vision. Recently, we have seen rapid developments in the joint
Vision-Language space as well, where pre-trained models such as CLIP (Radford
et al., 2021) have demonstrated improvements in downstream tasks like image
captioning and visual question answering. However, surprisingly there is
comparatively little work on exploring these models for the task of multimodal
machine translation, where the goal is to leverage image/video modality in
text-to-text translation. To fill this gap, this paper surveys the landscape of
language-and-vision pre-training from the lens of multimodal machine
translation. We summarize the common architectures, pre-training objectives,
and datasets from literature and conjecture what further is needed to make
progress on multimodal machine translation.
|
2306.07200 | Joonghyuk Shin | Joonghyuk Shin, Minguk Kang, Jaesik Park | Fill-Up: Balancing Long-Tailed Data with Generative Models | 32 pages, 19 Figures, and 10 Tables. Project webpage at
https://alex4727.github.io/Fill-Up/ | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern text-to-image synthesis models have achieved an exceptional level of
photorealism, generating high-quality images from arbitrary text descriptions.
In light of the impressive synthesis ability, several studies have exhibited
promising results in exploiting generated data for image recognition. However,
directly supplementing data-hungry situations in the real-world (e.g. few-shot
or long-tailed scenarios) with existing approaches result in marginal
performance gains, as they suffer to thoroughly reflect the distribution of the
real data. Through extensive experiments, this paper proposes a new image
synthesis pipeline for long-tailed situations using Textual Inversion. The
study demonstrates that generated images from textual-inverted text tokens
effectively aligns with the real domain, significantly enhancing the
recognition ability of a standard ResNet50 backbone. We also show that
real-world data imbalance scenarios can be successfully mitigated by filling up
the imbalanced data with synthetic images. In conjunction with techniques in
the area of long-tailed recognition, our method achieves state-of-the-art
results on standard long-tailed benchmarks when trained from scratch.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 16:01:20 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Shin",
"Joonghyuk",
""
],
[
"Kang",
"Minguk",
""
],
[
"Park",
"Jaesik",
""
]
] | TITLE: Fill-Up: Balancing Long-Tailed Data with Generative Models
ABSTRACT: Modern text-to-image synthesis models have achieved an exceptional level of
photorealism, generating high-quality images from arbitrary text descriptions.
In light of the impressive synthesis ability, several studies have exhibited
promising results in exploiting generated data for image recognition. However,
directly supplementing data-hungry situations in the real-world (e.g. few-shot
or long-tailed scenarios) with existing approaches result in marginal
performance gains, as they suffer to thoroughly reflect the distribution of the
real data. Through extensive experiments, this paper proposes a new image
synthesis pipeline for long-tailed situations using Textual Inversion. The
study demonstrates that generated images from textual-inverted text tokens
effectively aligns with the real domain, significantly enhancing the
recognition ability of a standard ResNet50 backbone. We also show that
real-world data imbalance scenarios can be successfully mitigated by filling up
the imbalanced data with synthetic images. In conjunction with techniques in
the area of long-tailed recognition, our method achieves state-of-the-art
results on standard long-tailed benchmarks when trained from scratch.
|
2306.07206 | Shuai Liu | Shuai Liu, Hyundong J. Cho, Marjorie Freedman, Xuezhe Ma, Jonathan May | RECAP: Retrieval-Enhanced Context-Aware Prefix Encoder for Personalized
Dialogue Response Generation | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Endowing chatbots with a consistent persona is essential to an engaging
conversation, yet it remains an unresolved challenge. In this work, we propose
a new retrieval-enhanced approach for personalized response generation.
Specifically, we design a hierarchical transformer retriever trained on
dialogue domain data to perform personalized retrieval and a context-aware
prefix encoder that fuses the retrieved information to the decoder more
effectively. Extensive experiments on a real-world dataset demonstrate the
effectiveness of our model at generating more fluent and personalized
responses. We quantitatively evaluate our model's performance under a suite of
human and automatic metrics and find it to be superior compared to
state-of-the-art baselines on English Reddit conversations.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 16:10:21 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Liu",
"Shuai",
""
],
[
"Cho",
"Hyundong J.",
""
],
[
"Freedman",
"Marjorie",
""
],
[
"Ma",
"Xuezhe",
""
],
[
"May",
"Jonathan",
""
]
] | TITLE: RECAP: Retrieval-Enhanced Context-Aware Prefix Encoder for Personalized
Dialogue Response Generation
ABSTRACT: Endowing chatbots with a consistent persona is essential to an engaging
conversation, yet it remains an unresolved challenge. In this work, we propose
a new retrieval-enhanced approach for personalized response generation.
Specifically, we design a hierarchical transformer retriever trained on
dialogue domain data to perform personalized retrieval and a context-aware
prefix encoder that fuses the retrieved information to the decoder more
effectively. Extensive experiments on a real-world dataset demonstrate the
effectiveness of our model at generating more fluent and personalized
responses. We quantitatively evaluate our model's performance under a suite of
human and automatic metrics and find it to be superior compared to
state-of-the-art baselines on English Reddit conversations.
|
2306.07244 | Peter Buckel M.Sc. | Peter Buckel, Timo Oksanen, Thomas Dietmueller | RB-Dust -- A Reference-based Dataset for Vision-based Dust Removal | Accepted by CVPR Workshop NTIRE 2023. Errata: Caption Figure 6
changed | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Dust in the agricultural landscape is a significant challenge and influences,
for example, the environmental perception of autonomous agricultural machines.
Image enhancement algorithms can be used to reduce dust. However, these require
dusty and dust-free images of the same environment for validation. In fact, to
date, there is no dataset that we are aware of that addresses this issue.
Therefore, we present the agriscapes RB-Dust dataset, which is named after its
purpose of reference-based dust removal. It is not possible to take pictures
from the cabin during tillage, as this would cause shifts in the images.
Because of this, we built a setup from which it is possible to take images from
a stationary position close to the passing tractor. The test setup was based on
a half-sided gate through which the tractor could drive. The field tests were
carried out on a farm in Bavaria, Germany, during tillage. During the field
tests, other parameters such as soil moisture and wind speed were controlled,
as these significantly affect dust development. We validated our dataset with
contrast enhancement and image dehazing algorithms and analyzed the
generalizability from recordings from the moving tractor. Finally, we
demonstrate the application of dust removal based on a high-level vision task,
such as person classification. Our empirical study confirms the validity of
RB-Dust for vision-based dust removal in agriculture.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 17:09:24 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Buckel",
"Peter",
""
],
[
"Oksanen",
"Timo",
""
],
[
"Dietmueller",
"Thomas",
""
]
] | TITLE: RB-Dust -- A Reference-based Dataset for Vision-based Dust Removal
ABSTRACT: Dust in the agricultural landscape is a significant challenge and influences,
for example, the environmental perception of autonomous agricultural machines.
Image enhancement algorithms can be used to reduce dust. However, these require
dusty and dust-free images of the same environment for validation. In fact, to
date, there is no dataset that we are aware of that addresses this issue.
Therefore, we present the agriscapes RB-Dust dataset, which is named after its
purpose of reference-based dust removal. It is not possible to take pictures
from the cabin during tillage, as this would cause shifts in the images.
Because of this, we built a setup from which it is possible to take images from
a stationary position close to the passing tractor. The test setup was based on
a half-sided gate through which the tractor could drive. The field tests were
carried out on a farm in Bavaria, Germany, during tillage. During the field
tests, other parameters such as soil moisture and wind speed were controlled,
as these significantly affect dust development. We validated our dataset with
contrast enhancement and image dehazing algorithms and analyzed the
generalizability from recordings from the moving tractor. Finally, we
demonstrate the application of dust removal based on a high-level vision task,
such as person classification. Our empirical study confirms the validity of
RB-Dust for vision-based dust removal in agriculture.
|
2306.07257 | Huiguo He | Junchen Zhu, Huan Yang, Huiguo He, Wenjing Wang, Zixi Tuo, Wen-Huang
Cheng, Lianli Gao, Jingkuan Song, Jianlong Fu | MovieFactory: Automatic Movie Creation from Text using Large Generative
Models for Language and Images | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we present MovieFactory, a powerful framework to generate
cinematic-picture (3072$\times$1280), film-style (multi-scene), and
multi-modality (sounding) movies on the demand of natural languages. As the
first fully automated movie generation model to the best of our knowledge, our
approach empowers users to create captivating movies with smooth transitions
using simple text inputs, surpassing existing methods that produce soundless
videos limited to a single scene of modest quality. To facilitate this
distinctive functionality, we leverage ChatGPT to expand user-provided text
into detailed sequential scripts for movie generation. Then we bring scripts to
life visually and acoustically through vision generation and audio retrieval.
To generate videos, we extend the capabilities of a pretrained text-to-image
diffusion model through a two-stage process. Firstly, we employ spatial
finetuning to bridge the gap between the pretrained image model and the new
video dataset. Subsequently, we introduce temporal learning to capture object
motion. In terms of audio, we leverage sophisticated retrieval models to select
and align audio elements that correspond to the plot and visual content of the
movie. Extensive experiments demonstrate that our MovieFactory produces movies
with realistic visuals, diverse scenes, and seamlessly fitting audio, offering
users a novel and immersive experience. Generated samples can be found in
YouTube or Bilibili (1080P).
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 17:31:23 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Zhu",
"Junchen",
""
],
[
"Yang",
"Huan",
""
],
[
"He",
"Huiguo",
""
],
[
"Wang",
"Wenjing",
""
],
[
"Tuo",
"Zixi",
""
],
[
"Cheng",
"Wen-Huang",
""
],
[
"Gao",
"Lianli",
""
],
[
"Song",
"Jingkuan",
""
],
[
"Fu",
"Jianlong",
""
]
] | TITLE: MovieFactory: Automatic Movie Creation from Text using Large Generative
Models for Language and Images
ABSTRACT: In this paper, we present MovieFactory, a powerful framework to generate
cinematic-picture (3072$\times$1280), film-style (multi-scene), and
multi-modality (sounding) movies on the demand of natural languages. As the
first fully automated movie generation model to the best of our knowledge, our
approach empowers users to create captivating movies with smooth transitions
using simple text inputs, surpassing existing methods that produce soundless
videos limited to a single scene of modest quality. To facilitate this
distinctive functionality, we leverage ChatGPT to expand user-provided text
into detailed sequential scripts for movie generation. Then we bring scripts to
life visually and acoustically through vision generation and audio retrieval.
To generate videos, we extend the capabilities of a pretrained text-to-image
diffusion model through a two-stage process. Firstly, we employ spatial
finetuning to bridge the gap between the pretrained image model and the new
video dataset. Subsequently, we introduce temporal learning to capture object
motion. In terms of audio, we leverage sophisticated retrieval models to select
and align audio elements that correspond to the plot and visual content of the
movie. Extensive experiments demonstrate that our MovieFactory produces movies
with realistic visuals, diverse scenes, and seamlessly fitting audio, offering
users a novel and immersive experience. Generated samples can be found in
YouTube or Bilibili (1080P).
|
2306.07284 | Tal Reiss | Tal Reiss, Niv Cohen, Yedid Hoshen | No Free Lunch: The Hazards of Over-Expressive Representations in Anomaly
Detection | null | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomaly detection methods, powered by deep learning, have recently been
making significant progress, mostly due to improved representations. It is
tempting to hypothesize that anomaly detection can improve indefinitely by
increasing the scale of our networks, making their representations more
expressive. In this paper, we provide theoretical and empirical evidence to the
contrary. In fact, we empirically show cases where very expressive
representations fail to detect even simple anomalies when evaluated beyond the
well-studied object-centric datasets. To investigate this phenomenon, we begin
by introducing a novel theoretical toy model for anomaly detection performance.
The model uncovers a fundamental trade-off between representation sufficiency
and over-expressivity. It provides evidence for a no-free-lunch theorem in
anomaly detection stating that increasing representation expressivity will
eventually result in performance degradation. Instead, guidance must be
provided to focus the representation on the attributes relevant to the
anomalies of interest. We conduct an extensive empirical investigation
demonstrating that state-of-the-art representations often suffer from
over-expressivity, failing to detect many types of anomalies. Our investigation
demonstrates how this over-expressivity impairs image anomaly detection in
practical settings. We conclude with future directions for mitigating this
issue.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 17:59:50 GMT"
}
] | 2023-06-13T00:00:00 | [
[
"Reiss",
"Tal",
""
],
[
"Cohen",
"Niv",
""
],
[
"Hoshen",
"Yedid",
""
]
] | TITLE: No Free Lunch: The Hazards of Over-Expressive Representations in Anomaly
Detection
ABSTRACT: Anomaly detection methods, powered by deep learning, have recently been
making significant progress, mostly due to improved representations. It is
tempting to hypothesize that anomaly detection can improve indefinitely by
increasing the scale of our networks, making their representations more
expressive. In this paper, we provide theoretical and empirical evidence to the
contrary. In fact, we empirically show cases where very expressive
representations fail to detect even simple anomalies when evaluated beyond the
well-studied object-centric datasets. To investigate this phenomenon, we begin
by introducing a novel theoretical toy model for anomaly detection performance.
The model uncovers a fundamental trade-off between representation sufficiency
and over-expressivity. It provides evidence for a no-free-lunch theorem in
anomaly detection stating that increasing representation expressivity will
eventually result in performance degradation. Instead, guidance must be
provided to focus the representation on the attributes relevant to the
anomalies of interest. We conduct an extensive empirical investigation
demonstrating that state-of-the-art representations often suffer from
over-expressivity, failing to detect many types of anomalies. Our investigation
demonstrates how this over-expressivity impairs image anomaly detection in
practical settings. We conclude with future directions for mitigating this
issue.
|
2008.12243 | Giuseppe Tagliavini | Fabio Montagna, Stefan Mach, Simone Benatti, Angelo Garofalo,
Gianmarco Ottavi, Luca Benini, Davide Rossi, Giuseppe Tagliavini | A transprecision floating-point cluster for efficient near-sensor data
analytics | null | null | 10.1109/TPDS.2021.3101764 | null | cs.DC cs.AR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent applications in the domain of near-sensor computing require the
adoption of floating-point arithmetic to reconcile high precision results with
a wide dynamic range. In this paper, we propose a multi-core computing cluster
that leverages the fined-grained tunable principles of transprecision computing
to provide support to near-sensor applications at a minimum power budget. Our
design - based on the open-source RISC-V architecture - combines
parallelization and sub-word vectorization with near-threshold operation,
leading to a highly scalable and versatile system. We perform an exhaustive
exploration of the design space of the transprecision cluster on a
cycle-accurate FPGA emulator, with the aim to identify the most efficient
configurations in terms of performance, energy efficiency, and area efficiency.
We also provide a full-fledged software stack support, including a parallel
runtime and a compilation toolchain, to enable the development of end-to-end
applications. We perform an experimental assessment of our design on a set of
benchmarks representative of the near-sensor processing domain, complementing
the timing results with a post place-&-route analysis of the power consumption.
Finally, a comparison with the state-of-the-art shows that our solution
outperforms the competitors in energy efficiency, reaching a peak of 97
Gflop/s/W on single-precision scalars and 162 Gflop/s/W on half-precision
vectors.
| [
{
"version": "v1",
"created": "Thu, 27 Aug 2020 16:38:26 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Montagna",
"Fabio",
""
],
[
"Mach",
"Stefan",
""
],
[
"Benatti",
"Simone",
""
],
[
"Garofalo",
"Angelo",
""
],
[
"Ottavi",
"Gianmarco",
""
],
[
"Benini",
"Luca",
""
],
[
"Rossi",
"Davide",
""
],
[
"Tagliavini",
"Giuseppe",
""
]
] | TITLE: A transprecision floating-point cluster for efficient near-sensor data
analytics
ABSTRACT: Recent applications in the domain of near-sensor computing require the
adoption of floating-point arithmetic to reconcile high precision results with
a wide dynamic range. In this paper, we propose a multi-core computing cluster
that leverages the fined-grained tunable principles of transprecision computing
to provide support to near-sensor applications at a minimum power budget. Our
design - based on the open-source RISC-V architecture - combines
parallelization and sub-word vectorization with near-threshold operation,
leading to a highly scalable and versatile system. We perform an exhaustive
exploration of the design space of the transprecision cluster on a
cycle-accurate FPGA emulator, with the aim to identify the most efficient
configurations in terms of performance, energy efficiency, and area efficiency.
We also provide a full-fledged software stack support, including a parallel
runtime and a compilation toolchain, to enable the development of end-to-end
applications. We perform an experimental assessment of our design on a set of
benchmarks representative of the near-sensor processing domain, complementing
the timing results with a post place-&-route analysis of the power consumption.
Finally, a comparison with the state-of-the-art shows that our solution
outperforms the competitors in energy efficiency, reaching a peak of 97
Gflop/s/W on single-precision scalars and 162 Gflop/s/W on half-precision
vectors.
|
2109.08079 | Himanshu Gupta | Himanshu Gupta, Shreyas Verma, Santosh Mashetty, Swaroop Mishra | Context-NER : Contextual Phrase Generation at Scale | 29 pages, 5 Figures, 2 AlgorithmS, 17 Tables. Accepted in NeurIPS
2022 - Efficient Natural Language and Speech Processing (ENLSP) Workshop | null | null | null | cs.IR cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Named Entity Recognition (NER) has seen significant progress in recent years,
with numerous state-of-the-art (SOTA) models achieving high performance.
However, very few studies have focused on the generation of entities' context.
In this paper, we introduce CONTEXT-NER, a task that aims to generate the
relevant context for entities in a sentence, where the context is a phrase
describing the entity but not necessarily present in the sentence. To
facilitate research in this task, we also present the EDGAR10-Q dataset, which
consists of annual and quarterly reports from the top 1500 publicly traded
companies. The dataset is the largest of its kind, containing 1M sentences,
2.8M entities, and an average of 35 tokens per sentence, making it a
challenging dataset. We propose a baseline approach that combines a phrase
generation algorithm with inferencing using a 220M language model, achieving a
ROUGE-L score of 27% on the test split. Additionally, we perform a one-shot
inference with ChatGPT, which obtains a 30% ROUGE-L, highlighting the
difficulty of the dataset. We also evaluate models such as T5 and BART, which
achieve a maximum ROUGE-L of 49% after supervised finetuning on EDGAR10-Q. We
also find that T5-large, when pre-finetuned on EDGAR10-Q, achieve SOTA results
on downstream finance tasks such as Headline, FPB, and FiQA SA, outperforming
vanilla version by 10.81 points. To our surprise, this 66x smaller
pre-finetuned model also surpasses the finance-specific LLM BloombergGPT-50B by
15 points. We hope that our dataset and generated artifacts will encourage
further research in this direction, leading to the development of more
sophisticated language models for financial text analysis
| [
{
"version": "v1",
"created": "Thu, 16 Sep 2021 16:10:05 GMT"
},
{
"version": "v2",
"created": "Thu, 27 Oct 2022 05:33:28 GMT"
},
{
"version": "v3",
"created": "Fri, 28 Oct 2022 04:49:28 GMT"
},
{
"version": "v4",
"created": "Thu, 8 Jun 2023 18:33:01 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Gupta",
"Himanshu",
""
],
[
"Verma",
"Shreyas",
""
],
[
"Mashetty",
"Santosh",
""
],
[
"Mishra",
"Swaroop",
""
]
] | TITLE: Context-NER : Contextual Phrase Generation at Scale
ABSTRACT: Named Entity Recognition (NER) has seen significant progress in recent years,
with numerous state-of-the-art (SOTA) models achieving high performance.
However, very few studies have focused on the generation of entities' context.
In this paper, we introduce CONTEXT-NER, a task that aims to generate the
relevant context for entities in a sentence, where the context is a phrase
describing the entity but not necessarily present in the sentence. To
facilitate research in this task, we also present the EDGAR10-Q dataset, which
consists of annual and quarterly reports from the top 1500 publicly traded
companies. The dataset is the largest of its kind, containing 1M sentences,
2.8M entities, and an average of 35 tokens per sentence, making it a
challenging dataset. We propose a baseline approach that combines a phrase
generation algorithm with inferencing using a 220M language model, achieving a
ROUGE-L score of 27% on the test split. Additionally, we perform a one-shot
inference with ChatGPT, which obtains a 30% ROUGE-L, highlighting the
difficulty of the dataset. We also evaluate models such as T5 and BART, which
achieve a maximum ROUGE-L of 49% after supervised finetuning on EDGAR10-Q. We
also find that T5-large, when pre-finetuned on EDGAR10-Q, achieve SOTA results
on downstream finance tasks such as Headline, FPB, and FiQA SA, outperforming
vanilla version by 10.81 points. To our surprise, this 66x smaller
pre-finetuned model also surpasses the finance-specific LLM BloombergGPT-50B by
15 points. We hope that our dataset and generated artifacts will encourage
further research in this direction, leading to the development of more
sophisticated language models for financial text analysis
|
2110.08627 | Zixin Zhong | Zixin Zhong, Wang Chi Cheung, Vincent Y. F. Tan | Achieving the Pareto Frontier of Regret Minimization and Best Arm
Identification in Multi-Armed Bandits | 43 pages, 10 figures | null | null | null | cs.LG cs.AI cs.IT math.IT stat.ML | http://creativecommons.org/licenses/by/4.0/ | We study the Pareto frontier of two archetypal objectives in multi-armed
bandits, namely, regret minimization (RM) and best arm identification (BAI)
with a fixed horizon. It is folklore that the balance between exploitation and
exploration is crucial for both RM and BAI, but exploration is more critical in
achieving the optimal performance for the latter objective. To this end, we
design and analyze the BoBW-lil'UCB$(\gamma)$ algorithm. Complementarily, by
establishing lower bounds on the regret achievable by any algorithm with a
given BAI failure probability, we show that (i) no algorithm can simultaneously
perform optimally for both the RM and BAI objectives, and (ii)
BoBW-lil'UCB$(\gamma)$ achieves order-wise optimal performance for RM or BAI
under different values of $\gamma$. Our work elucidates the trade-off more
precisely by showing how the constants in previous works depend on certain
hardness parameters. Finally, we show that BoBW-lil'UCB outperforms a close
competitor UCB$_\alpha$ (Degenne et al., 2019) in terms of the time complexity
and the regret on diverse datasets such as MovieLens and Published Kinase
Inhibitor Set.
| [
{
"version": "v1",
"created": "Sat, 16 Oct 2021 17:52:32 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Oct 2022 04:26:14 GMT"
},
{
"version": "v3",
"created": "Fri, 9 Jun 2023 05:14:58 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Zhong",
"Zixin",
""
],
[
"Cheung",
"Wang Chi",
""
],
[
"Tan",
"Vincent Y. F.",
""
]
] | TITLE: Achieving the Pareto Frontier of Regret Minimization and Best Arm
Identification in Multi-Armed Bandits
ABSTRACT: We study the Pareto frontier of two archetypal objectives in multi-armed
bandits, namely, regret minimization (RM) and best arm identification (BAI)
with a fixed horizon. It is folklore that the balance between exploitation and
exploration is crucial for both RM and BAI, but exploration is more critical in
achieving the optimal performance for the latter objective. To this end, we
design and analyze the BoBW-lil'UCB$(\gamma)$ algorithm. Complementarily, by
establishing lower bounds on the regret achievable by any algorithm with a
given BAI failure probability, we show that (i) no algorithm can simultaneously
perform optimally for both the RM and BAI objectives, and (ii)
BoBW-lil'UCB$(\gamma)$ achieves order-wise optimal performance for RM or BAI
under different values of $\gamma$. Our work elucidates the trade-off more
precisely by showing how the constants in previous works depend on certain
hardness parameters. Finally, we show that BoBW-lil'UCB outperforms a close
competitor UCB$_\alpha$ (Degenne et al., 2019) in terms of the time complexity
and the regret on diverse datasets such as MovieLens and Published Kinase
Inhibitor Set.
|
2111.10056 | Zhihong Lin | Zhihong Lin, Donghao Zhang, Qingyi Tao, Danli Shi, Gholamreza Haffari,
Qi Wu, Mingguang He, and Zongyuan Ge | Medical Visual Question Answering: A Survey | null | null | 10.1016/j.artmed.2023.102611 | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Medical Visual Question Answering~(VQA) is a combination of medical
artificial intelligence and popular VQA challenges. Given a medical image and a
clinically relevant question in natural language, the medical VQA system is
expected to predict a plausible and convincing answer. Although the
general-domain VQA has been extensively studied, the medical VQA still needs
specific investigation and exploration due to its task features. In the first
part of this survey, we collect and discuss the publicly available medical VQA
datasets up-to-date about the data source, data quantity, and task feature. In
the second part, we review the approaches used in medical VQA tasks. We
summarize and discuss their techniques, innovations, and potential
improvements. In the last part, we analyze some medical-specific challenges for
the field and discuss future research directions. Our goal is to provide
comprehensive and helpful information for researchers interested in the medical
visual question answering field and encourage them to conduct further research
in this field.
| [
{
"version": "v1",
"created": "Fri, 19 Nov 2021 05:55:15 GMT"
},
{
"version": "v2",
"created": "Wed, 19 Jan 2022 03:46:33 GMT"
},
{
"version": "v3",
"created": "Wed, 7 Jun 2023 00:37:15 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Lin",
"Zhihong",
""
],
[
"Zhang",
"Donghao",
""
],
[
"Tao",
"Qingyi",
""
],
[
"Shi",
"Danli",
""
],
[
"Haffari",
"Gholamreza",
""
],
[
"Wu",
"Qi",
""
],
[
"He",
"Mingguang",
""
],
[
"Ge",
"Zongyuan",
""
]
] | TITLE: Medical Visual Question Answering: A Survey
ABSTRACT: Medical Visual Question Answering~(VQA) is a combination of medical
artificial intelligence and popular VQA challenges. Given a medical image and a
clinically relevant question in natural language, the medical VQA system is
expected to predict a plausible and convincing answer. Although the
general-domain VQA has been extensively studied, the medical VQA still needs
specific investigation and exploration due to its task features. In the first
part of this survey, we collect and discuss the publicly available medical VQA
datasets up-to-date about the data source, data quantity, and task feature. In
the second part, we review the approaches used in medical VQA tasks. We
summarize and discuss their techniques, innovations, and potential
improvements. In the last part, we analyze some medical-specific challenges for
the field and discuss future research directions. Our goal is to provide
comprehensive and helpful information for researchers interested in the medical
visual question answering field and encourage them to conduct further research
in this field.
|
2202.06268 | Nannan Li | Nannan Li, Yaran Chen, Weifan Li, Zixiang Ding, Dongbin Zhao | BViT: Broad Attention based Vision Transformer | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent works have demonstrated that transformer can achieve promising
performance in computer vision, by exploiting the relationship among image
patches with self-attention. While they only consider the attention in a single
feature layer, but ignore the complementarity of attention in different levels.
In this paper, we propose the broad attention to improve the performance by
incorporating the attention relationship of different layers for vision
transformer, which is called BViT. The broad attention is implemented by broad
connection and parameter-free attention. Broad connection of each transformer
layer promotes the transmission and integration of information for BViT.
Without introducing additional trainable parameters, parameter-free attention
jointly focuses on the already available attention information in different
layers for extracting useful information and building their relationship.
Experiments on image classification tasks demonstrate that BViT delivers
state-of-the-art accuracy of 74.8\%/81.6\% top-1 accuracy on ImageNet with
5M/22M parameters. Moreover, we transfer BViT to downstream object recognition
benchmarks to achieve 98.9\% and 89.9\% on CIFAR10 and CIFAR100 respectively
that exceed ViT with fewer parameters. For the generalization test, the broad
attention in Swin Transformer and T2T-ViT also bring an improvement of more
than 1\%. To sum up, broad attention is promising to promote the performance of
attention based models. Code and pre-trained models are available at
https://github.com/DRL-CASIA/Broad_ViT.
| [
{
"version": "v1",
"created": "Sun, 13 Feb 2022 09:23:29 GMT"
},
{
"version": "v2",
"created": "Fri, 9 Jun 2023 06:08:37 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Li",
"Nannan",
""
],
[
"Chen",
"Yaran",
""
],
[
"Li",
"Weifan",
""
],
[
"Ding",
"Zixiang",
""
],
[
"Zhao",
"Dongbin",
""
]
] | TITLE: BViT: Broad Attention based Vision Transformer
ABSTRACT: Recent works have demonstrated that transformer can achieve promising
performance in computer vision, by exploiting the relationship among image
patches with self-attention. While they only consider the attention in a single
feature layer, but ignore the complementarity of attention in different levels.
In this paper, we propose the broad attention to improve the performance by
incorporating the attention relationship of different layers for vision
transformer, which is called BViT. The broad attention is implemented by broad
connection and parameter-free attention. Broad connection of each transformer
layer promotes the transmission and integration of information for BViT.
Without introducing additional trainable parameters, parameter-free attention
jointly focuses on the already available attention information in different
layers for extracting useful information and building their relationship.
Experiments on image classification tasks demonstrate that BViT delivers
state-of-the-art accuracy of 74.8\%/81.6\% top-1 accuracy on ImageNet with
5M/22M parameters. Moreover, we transfer BViT to downstream object recognition
benchmarks to achieve 98.9\% and 89.9\% on CIFAR10 and CIFAR100 respectively
that exceed ViT with fewer parameters. For the generalization test, the broad
attention in Swin Transformer and T2T-ViT also bring an improvement of more
than 1\%. To sum up, broad attention is promising to promote the performance of
attention based models. Code and pre-trained models are available at
https://github.com/DRL-CASIA/Broad_ViT.
|
2203.13617 | Zheng Lian | Haiyang Sun, Zheng Lian, Bin Liu, Ying Li, Licai Sun, Cong Cai,
Jianhua Tao, Meng Wang, Yuan Cheng | EmotionNAS: Two-stream Neural Architecture Search for Speech Emotion
Recognition | Accepted to Interspeech 2023 | null | null | null | eess.AS cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech emotion recognition (SER) is an important research topic in
human-computer interaction. Existing works mainly rely on human expertise to
design models. Despite their success, different datasets often require distinct
structures and hyperparameters. Searching for an optimal model for each dataset
is time-consuming and labor-intensive. To address this problem, we propose a
two-stream neural architecture search (NAS) based framework, called
\enquote{EmotionNAS}. Specifically, we take two-stream features (i.e.,
handcrafted and deep features) as the inputs, followed by NAS to search for the
optimal structure for each stream. Furthermore, we incorporate complementary
information in different streams through an efficient information supplement
module. Experimental results demonstrate that our method outperforms existing
manually-designed and NAS-based models, setting the new state-of-the-art
record.
| [
{
"version": "v1",
"created": "Fri, 25 Mar 2022 12:35:44 GMT"
},
{
"version": "v2",
"created": "Fri, 9 Jun 2023 14:45:18 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Sun",
"Haiyang",
""
],
[
"Lian",
"Zheng",
""
],
[
"Liu",
"Bin",
""
],
[
"Li",
"Ying",
""
],
[
"Sun",
"Licai",
""
],
[
"Cai",
"Cong",
""
],
[
"Tao",
"Jianhua",
""
],
[
"Wang",
"Meng",
""
],
[
"Cheng",
"Yuan",
""
]
] | TITLE: EmotionNAS: Two-stream Neural Architecture Search for Speech Emotion
Recognition
ABSTRACT: Speech emotion recognition (SER) is an important research topic in
human-computer interaction. Existing works mainly rely on human expertise to
design models. Despite their success, different datasets often require distinct
structures and hyperparameters. Searching for an optimal model for each dataset
is time-consuming and labor-intensive. To address this problem, we propose a
two-stream neural architecture search (NAS) based framework, called
\enquote{EmotionNAS}. Specifically, we take two-stream features (i.e.,
handcrafted and deep features) as the inputs, followed by NAS to search for the
optimal structure for each stream. Furthermore, we incorporate complementary
information in different streams through an efficient information supplement
module. Experimental results demonstrate that our method outperforms existing
manually-designed and NAS-based models, setting the new state-of-the-art
record.
|
2205.06733 | Dominic Petrak | Dominic Petrak, Nafise Sadat Moosavi, Iryna Gurevych | Arithmetic-Based Pretraining -- Improving Numeracy of Pretrained
Language Models | Published at StarSEM2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | State-of-the-art pretrained language models tend to perform below their
capabilities when applied out-of-the-box on tasks that require understanding
and working with numbers. Recent work suggests two main reasons for this: (1)
popular tokenisation algorithms have limited expressiveness for numbers, and
(2) common pretraining objectives do not target numeracy. Approaches that
address these shortcomings usually require architectural changes or pretraining
from scratch. In this paper, we propose a new extended pretraining approach
called Arithmetic-Based Pretraining that jointly addresses both in one extended
pretraining step without requiring architectural changes or pretraining from
scratch. Arithmetic-Based Pretraining combines contrastive learning to improve
the number representation, and a novel extended pretraining objective called
Inferable Number Prediction Task to improve numeracy. Our experiments show the
effectiveness of Arithmetic-Based Pretraining in three different tasks that
require improved numeracy, i.e., reading comprehension in the DROP dataset,
inference-on-tables in the InfoTabs dataset, and table-to-text generation in
the WikiBio and SciGen datasets.
| [
{
"version": "v1",
"created": "Fri, 13 May 2022 16:10:13 GMT"
},
{
"version": "v2",
"created": "Fri, 9 Jun 2023 08:04:40 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Petrak",
"Dominic",
""
],
[
"Moosavi",
"Nafise Sadat",
""
],
[
"Gurevych",
"Iryna",
""
]
] | TITLE: Arithmetic-Based Pretraining -- Improving Numeracy of Pretrained
Language Models
ABSTRACT: State-of-the-art pretrained language models tend to perform below their
capabilities when applied out-of-the-box on tasks that require understanding
and working with numbers. Recent work suggests two main reasons for this: (1)
popular tokenisation algorithms have limited expressiveness for numbers, and
(2) common pretraining objectives do not target numeracy. Approaches that
address these shortcomings usually require architectural changes or pretraining
from scratch. In this paper, we propose a new extended pretraining approach
called Arithmetic-Based Pretraining that jointly addresses both in one extended
pretraining step without requiring architectural changes or pretraining from
scratch. Arithmetic-Based Pretraining combines contrastive learning to improve
the number representation, and a novel extended pretraining objective called
Inferable Number Prediction Task to improve numeracy. Our experiments show the
effectiveness of Arithmetic-Based Pretraining in three different tasks that
require improved numeracy, i.e., reading comprehension in the DROP dataset,
inference-on-tables in the InfoTabs dataset, and table-to-text generation in
the WikiBio and SciGen datasets.
|
2205.15239 | Julian Lienen | Julian Lienen, Caglar Demir, Eyke H\"ullermeier | Conformal Credal Self-Supervised Learning | 26 pages, 5 figures, 10 tables, to be published at the 12th Symposium
on Conformal and Probabilistic Prediction with Applications (COPA 2023) | null | null | null | stat.ML cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In semi-supervised learning, the paradigm of self-training refers to the idea
of learning from pseudo-labels suggested by the learner itself. Across various
domains, corresponding methods have proven effective and achieve
state-of-the-art performance. However, pseudo-labels typically stem from ad-hoc
heuristics, relying on the quality of the predictions though without
guaranteeing their validity. One such method, so-called credal self-supervised
learning, maintains pseudo-supervision in the form of sets of (instead of
single) probability distributions over labels, thereby allowing for a flexible
yet uncertainty-aware labeling. Again, however, there is no justification
beyond empirical effectiveness. To address this deficiency, we make use of
conformal prediction, an approach that comes with guarantees on the validity of
set-valued predictions. As a result, the construction of credal sets of labels
is supported by a rigorous theoretical foundation, leading to better calibrated
and less error-prone supervision for unlabeled data. Along with this, we
present effective algorithms for learning from credal self-supervision. An
empirical study demonstrates excellent calibration properties of the
pseudo-supervision, as well as the competitiveness of our method on several
benchmark datasets.
| [
{
"version": "v1",
"created": "Mon, 30 May 2022 16:53:16 GMT"
},
{
"version": "v2",
"created": "Fri, 9 Jun 2023 13:30:44 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Lienen",
"Julian",
""
],
[
"Demir",
"Caglar",
""
],
[
"Hüllermeier",
"Eyke",
""
]
] | TITLE: Conformal Credal Self-Supervised Learning
ABSTRACT: In semi-supervised learning, the paradigm of self-training refers to the idea
of learning from pseudo-labels suggested by the learner itself. Across various
domains, corresponding methods have proven effective and achieve
state-of-the-art performance. However, pseudo-labels typically stem from ad-hoc
heuristics, relying on the quality of the predictions though without
guaranteeing their validity. One such method, so-called credal self-supervised
learning, maintains pseudo-supervision in the form of sets of (instead of
single) probability distributions over labels, thereby allowing for a flexible
yet uncertainty-aware labeling. Again, however, there is no justification
beyond empirical effectiveness. To address this deficiency, we make use of
conformal prediction, an approach that comes with guarantees on the validity of
set-valued predictions. As a result, the construction of credal sets of labels
is supported by a rigorous theoretical foundation, leading to better calibrated
and less error-prone supervision for unlabeled data. Along with this, we
present effective algorithms for learning from credal self-supervision. An
empirical study demonstrates excellent calibration properties of the
pseudo-supervision, as well as the competitiveness of our method on several
benchmark datasets.
|
2206.06602 | Hongzuo Xu | Hongzuo Xu and Guansong Pang and Yijie Wang and Yongjun Wang | Deep Isolation Forest for Anomaly Detection | Accepted by IEEE Transactions on Knowledge and Data Engineering
(TKDE) | null | 10.1109/TKDE.2023.3270293 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Isolation forest (iForest) has been emerging as arguably the most popular
anomaly detector in recent years due to its general effectiveness across
different benchmarks and strong scalability. Nevertheless, its linear
axis-parallel isolation method often leads to (i) failure in detecting hard
anomalies that are difficult to isolate in
high-dimensional/non-linear-separable data space, and (ii) notorious
algorithmic bias that assigns unexpectedly lower anomaly scores to artefact
regions. These issues contribute to high false negative errors. Several iForest
extensions are introduced, but they essentially still employ shallow, linear
data partition, restricting their power in isolating true anomalies. Therefore,
this paper proposes deep isolation forest. We introduce a new representation
scheme that utilises casually initialised neural networks to map original data
into random representation ensembles, where random axis-parallel cuts are
subsequently applied to perform the data partition. This representation scheme
facilitates high freedom of the partition in the original data space
(equivalent to non-linear partition on subspaces of varying sizes), encouraging
a unique synergy between random representations and random partition-based
isolation. Extensive experiments show that our model achieves significant
improvement over state-of-the-art isolation-based methods and deep detectors on
tabular, graph and time series datasets; our model also inherits desired
scalability from iForest.
| [
{
"version": "v1",
"created": "Tue, 14 Jun 2022 05:47:07 GMT"
},
{
"version": "v2",
"created": "Thu, 29 Sep 2022 02:50:25 GMT"
},
{
"version": "v3",
"created": "Sun, 23 Apr 2023 04:45:05 GMT"
},
{
"version": "v4",
"created": "Fri, 9 Jun 2023 01:19:55 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Xu",
"Hongzuo",
""
],
[
"Pang",
"Guansong",
""
],
[
"Wang",
"Yijie",
""
],
[
"Wang",
"Yongjun",
""
]
] | TITLE: Deep Isolation Forest for Anomaly Detection
ABSTRACT: Isolation forest (iForest) has been emerging as arguably the most popular
anomaly detector in recent years due to its general effectiveness across
different benchmarks and strong scalability. Nevertheless, its linear
axis-parallel isolation method often leads to (i) failure in detecting hard
anomalies that are difficult to isolate in
high-dimensional/non-linear-separable data space, and (ii) notorious
algorithmic bias that assigns unexpectedly lower anomaly scores to artefact
regions. These issues contribute to high false negative errors. Several iForest
extensions are introduced, but they essentially still employ shallow, linear
data partition, restricting their power in isolating true anomalies. Therefore,
this paper proposes deep isolation forest. We introduce a new representation
scheme that utilises casually initialised neural networks to map original data
into random representation ensembles, where random axis-parallel cuts are
subsequently applied to perform the data partition. This representation scheme
facilitates high freedom of the partition in the original data space
(equivalent to non-linear partition on subspaces of varying sizes), encouraging
a unique synergy between random representations and random partition-based
isolation. Extensive experiments show that our model achieves significant
improvement over state-of-the-art isolation-based methods and deep detectors on
tabular, graph and time series datasets; our model also inherits desired
scalability from iForest.
|
2206.09280 | Namyong Park | Namyong Park, Ryan Rossi, Nesreen Ahmed, Christos Faloutsos | MetaGL: Evaluation-Free Selection of Graph Learning Models via
Meta-Learning | ICLR 2023 | null | null | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given a graph learning task, such as link prediction, on a new graph, how can
we select the best method as well as its hyperparameters (collectively called a
model) without having to train or evaluate any model on the new graph? Model
selection for graph learning has been largely ad hoc. A typical approach has
been to apply popular methods to new datasets, but this is often suboptimal. On
the other hand, systematically comparing models on the new graph quickly
becomes too costly, or even impractical. In this work, we develop the first
meta-learning approach for evaluation-free graph learning model selection,
called MetaGL, which utilizes the prior performances of existing methods on
various benchmark graph datasets to automatically select an effective model for
the new graph, without any model training or evaluations. To quantify
similarities across a wide variety of graphs, we introduce specialized
meta-graph features that capture the structural characteristics of a graph.
Then we design G-M network, which represents the relations among graphs and
models, and develop a graph-based meta-learner operating on this G-M network,
which estimates the relevance of each model to different graphs. Extensive
experiments show that using MetaGL to select a model for the new graph greatly
outperforms several existing meta-learning techniques tailored for graph
learning model selection (up to 47% better), while being extremely fast at test
time (~1 sec).
| [
{
"version": "v1",
"created": "Sat, 18 Jun 2022 20:43:38 GMT"
},
{
"version": "v2",
"created": "Tue, 23 Aug 2022 04:50:40 GMT"
},
{
"version": "v3",
"created": "Thu, 8 Jun 2023 23:11:48 GMT"
}
] | 2023-06-12T00:00:00 | [
[
"Park",
"Namyong",
""
],
[
"Rossi",
"Ryan",
""
],
[
"Ahmed",
"Nesreen",
""
],
[
"Faloutsos",
"Christos",
""
]
] | TITLE: MetaGL: Evaluation-Free Selection of Graph Learning Models via
Meta-Learning
ABSTRACT: Given a graph learning task, such as link prediction, on a new graph, how can
we select the best method as well as its hyperparameters (collectively called a
model) without having to train or evaluate any model on the new graph? Model
selection for graph learning has been largely ad hoc. A typical approach has
been to apply popular methods to new datasets, but this is often suboptimal. On
the other hand, systematically comparing models on the new graph quickly
becomes too costly, or even impractical. In this work, we develop the first
meta-learning approach for evaluation-free graph learning model selection,
called MetaGL, which utilizes the prior performances of existing methods on
various benchmark graph datasets to automatically select an effective model for
the new graph, without any model training or evaluations. To quantify
similarities across a wide variety of graphs, we introduce specialized
meta-graph features that capture the structural characteristics of a graph.
Then we design G-M network, which represents the relations among graphs and
models, and develop a graph-based meta-learner operating on this G-M network,
which estimates the relevance of each model to different graphs. Extensive
experiments show that using MetaGL to select a model for the new graph greatly
outperforms several existing meta-learning techniques tailored for graph
learning model selection (up to 47% better), while being extremely fast at test
time (~1 sec).
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.