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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2306.11052 | Maria Grammatikopoulou | Maria Grammatikopoulou, Ricardo Sanchez-Matilla, Felix Bragman, David
Owen, Lucy Culshaw, Karen Kerr, Danail Stoyanov, Imanol Luengo | A spatio-temporal network for video semantic segmentation in surgical
videos | Accepted at IPCAI 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Semantic segmentation in surgical videos has applications in intra-operative
guidance, post-operative analytics and surgical education. Segmentation models
need to provide accurate and consistent predictions since temporally
inconsistent identification of anatomical structures can impair usability and
hinder patient safety. Video information can alleviate these challenges leading
to reliable models suitable for clinical use. We propose a novel architecture
for modelling temporal relationships in videos. The proposed model includes a
spatio-temporal decoder to enable video semantic segmentation by improving
temporal consistency across frames. The encoder processes individual frames
whilst the decoder processes a temporal batch of adjacent frames. The proposed
decoder can be used on top of any segmentation encoder to improve temporal
consistency. Model performance was evaluated on the CholecSeg8k dataset and a
private dataset of robotic Partial Nephrectomy procedures. Segmentation
performance was improved when the temporal decoder was applied across both
datasets. The proposed model also displayed improvements in temporal
consistency.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 16:36:48 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Grammatikopoulou",
"Maria",
""
],
[
"Sanchez-Matilla",
"Ricardo",
""
],
[
"Bragman",
"Felix",
""
],
[
"Owen",
"David",
""
],
[
"Culshaw",
"Lucy",
""
],
[
"Kerr",
"Karen",
""
],
[
"Stoyanov",
"Danail",
""
],
[
"Luengo",
"Imanol",
""
]
] | TITLE: A spatio-temporal network for video semantic segmentation in surgical
videos
ABSTRACT: Semantic segmentation in surgical videos has applications in intra-operative
guidance, post-operative analytics and surgical education. Segmentation models
need to provide accurate and consistent predictions since temporally
inconsistent identification of anatomical structures can impair usability and
hinder patient safety. Video information can alleviate these challenges leading
to reliable models suitable for clinical use. We propose a novel architecture
for modelling temporal relationships in videos. The proposed model includes a
spatio-temporal decoder to enable video semantic segmentation by improving
temporal consistency across frames. The encoder processes individual frames
whilst the decoder processes a temporal batch of adjacent frames. The proposed
decoder can be used on top of any segmentation encoder to improve temporal
consistency. Model performance was evaluated on the CholecSeg8k dataset and a
private dataset of robotic Partial Nephrectomy procedures. Segmentation
performance was improved when the temporal decoder was applied across both
datasets. The proposed model also displayed improvements in temporal
consistency.
|
2306.11056 | Linxin Song | Linxin Song, Jieyu Zhang, Xiaotian Lu, Tianyi Zhou | Taming Small-sample Bias in Low-budget Active Learning | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Active learning (AL) aims to minimize the annotation cost by only querying a
few informative examples for each model training stage. However, training a
model on a few queried examples suffers from the small-sample bias. In this
paper, we address this small-sample bias issue in low-budget AL by exploring a
regularizer called Firth bias reduction, which can provably reduce the bias
during the model training process but might hinder learning if its coefficient
is not adaptive to the learning progress. Instead of tuning the coefficient for
each query round, which is sensitive and time-consuming, we propose the
curriculum Firth bias reduction (CHAIN) that can automatically adjust the
coefficient to be adaptive to the training process. Under both deep learning
and linear model settings, experiments on three benchmark datasets with several
widely used query strategies and hyperparameter searching methods show that
CHAIN can be used to build more efficient AL and can substantially improve the
progress made by each active learning query.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 16:42:11 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Song",
"Linxin",
""
],
[
"Zhang",
"Jieyu",
""
],
[
"Lu",
"Xiaotian",
""
],
[
"Zhou",
"Tianyi",
""
]
] | TITLE: Taming Small-sample Bias in Low-budget Active Learning
ABSTRACT: Active learning (AL) aims to minimize the annotation cost by only querying a
few informative examples for each model training stage. However, training a
model on a few queried examples suffers from the small-sample bias. In this
paper, we address this small-sample bias issue in low-budget AL by exploring a
regularizer called Firth bias reduction, which can provably reduce the bias
during the model training process but might hinder learning if its coefficient
is not adaptive to the learning progress. Instead of tuning the coefficient for
each query round, which is sensitive and time-consuming, we propose the
curriculum Firth bias reduction (CHAIN) that can automatically adjust the
coefficient to be adaptive to the training process. Under both deep learning
and linear model settings, experiments on three benchmark datasets with several
widely used query strategies and hyperparameter searching methods show that
CHAIN can be used to build more efficient AL and can substantially improve the
progress made by each active learning query.
|
2306.11074 | Andres Potapczynski | Shikai Qiu, Andres Potapczynski, Pavel Izmailov and Andrew Gordon
Wilson | Simple and Fast Group Robustness by Automatic Feature Reweighting | ICML 23. Code available at https://github.com/AndPotap/afr | 40th International Conference on Machine Learning 2023 | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A major challenge to out-of-distribution generalization is reliance on
spurious features -- patterns that are predictive of the class label in the
training data distribution, but not causally related to the target. Standard
methods for reducing the reliance on spurious features typically assume that we
know what the spurious feature is, which is rarely true in the real world.
Methods that attempt to alleviate this limitation are complex, hard to tune,
and lead to a significant computational overhead compared to standard training.
In this paper, we propose Automatic Feature Reweighting (AFR), an extremely
simple and fast method for updating the model to reduce the reliance on
spurious features. AFR retrains the last layer of a standard ERM-trained base
model with a weighted loss that emphasizes the examples where the ERM model
predicts poorly, automatically upweighting the minority group without group
labels. With this simple procedure, we improve upon the best reported results
among competing methods trained without spurious attributes on several vision
and natural language classification benchmarks, using only a fraction of their
compute.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 17:19:13 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Qiu",
"Shikai",
""
],
[
"Potapczynski",
"Andres",
""
],
[
"Izmailov",
"Pavel",
""
],
[
"Wilson",
"Andrew Gordon",
""
]
] | TITLE: Simple and Fast Group Robustness by Automatic Feature Reweighting
ABSTRACT: A major challenge to out-of-distribution generalization is reliance on
spurious features -- patterns that are predictive of the class label in the
training data distribution, but not causally related to the target. Standard
methods for reducing the reliance on spurious features typically assume that we
know what the spurious feature is, which is rarely true in the real world.
Methods that attempt to alleviate this limitation are complex, hard to tune,
and lead to a significant computational overhead compared to standard training.
In this paper, we propose Automatic Feature Reweighting (AFR), an extremely
simple and fast method for updating the model to reduce the reliance on
spurious features. AFR retrains the last layer of a standard ERM-trained base
model with a weighted loss that emphasizes the examples where the ERM model
predicts poorly, automatically upweighting the minority group without group
labels. With this simple procedure, we improve upon the best reported results
among competing methods trained without spurious attributes on several vision
and natural language classification benchmarks, using only a fraction of their
compute.
|
2306.11114 | Aleksandr Petrov | Aleksandr V. Petrov and Craig Macdonald | Generative Sequential Recommendation with GPTRec | Accepted at Gen-IR@SIGIR2023 workshop | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Sequential recommendation is an important recommendation task that aims to
predict the next item in a sequence. Recently, adaptations of language models,
particularly Transformer-based models such as SASRec and BERT4Rec, have
achieved state-of-the-art results in sequential recommendation. In these
models, item ids replace tokens in the original language models. However, this
approach has limitations. First, the vocabulary of item ids may be many times
larger than in language models. Second, the classical Top-K recommendation
approach used by these models may not be optimal for complex recommendation
objectives, including auxiliary objectives such as diversity, coverage or
coherence. Recent progress in generative language models inspires us to revisit
generative approaches to address these challenges. This paper presents the
GPTRec sequential recommendation model, which is based on the GPT-2
architecture. GPTRec can address large vocabulary issues by splitting item ids
into sub-id tokens using a novel SVD Tokenisation algorithm based on quantised
item embeddings from an SVD decomposition of the user-item interaction matrix.
The paper also presents a novel Next-K recommendation strategy, which generates
recommendations item-by-item, considering already recommended items. The Next-K
strategy can be used for producing complex interdependent recommendation lists.
We experiment with GPTRec on the MovieLens-1M dataset and show that using
sub-item tokenisation GPTRec can match the quality of SASRec while reducing the
embedding table by 40%. We also show that the recommendations generated by
GPTRec on MovieLens-1M using the Next-K recommendation strategy match the
quality of SASRec in terms of NDCG@10, meaning that the model can serve as a
strong starting point for future research.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 18:27:54 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Petrov",
"Aleksandr V.",
""
],
[
"Macdonald",
"Craig",
""
]
] | TITLE: Generative Sequential Recommendation with GPTRec
ABSTRACT: Sequential recommendation is an important recommendation task that aims to
predict the next item in a sequence. Recently, adaptations of language models,
particularly Transformer-based models such as SASRec and BERT4Rec, have
achieved state-of-the-art results in sequential recommendation. In these
models, item ids replace tokens in the original language models. However, this
approach has limitations. First, the vocabulary of item ids may be many times
larger than in language models. Second, the classical Top-K recommendation
approach used by these models may not be optimal for complex recommendation
objectives, including auxiliary objectives such as diversity, coverage or
coherence. Recent progress in generative language models inspires us to revisit
generative approaches to address these challenges. This paper presents the
GPTRec sequential recommendation model, which is based on the GPT-2
architecture. GPTRec can address large vocabulary issues by splitting item ids
into sub-id tokens using a novel SVD Tokenisation algorithm based on quantised
item embeddings from an SVD decomposition of the user-item interaction matrix.
The paper also presents a novel Next-K recommendation strategy, which generates
recommendations item-by-item, considering already recommended items. The Next-K
strategy can be used for producing complex interdependent recommendation lists.
We experiment with GPTRec on the MovieLens-1M dataset and show that using
sub-item tokenisation GPTRec can match the quality of SASRec while reducing the
embedding table by 40%. We also show that the recommendations generated by
GPTRec on MovieLens-1M using the Next-K recommendation strategy match the
quality of SASRec in terms of NDCG@10, meaning that the model can serve as a
strong starting point for future research.
|
2306.11120 | Annie Chen | Annie S. Chen, Yoonho Lee, Amrith Setlur, Sergey Levine, Chelsea Finn | Confidence-Based Model Selection: When to Take Shortcuts for
Subpopulation Shifts | 15 pages, 5 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Effective machine learning models learn both robust features that directly
determine the outcome of interest (e.g., an object with wheels is more likely
to be a car), and shortcut features (e.g., an object on a road is more likely
to be a car). The latter can be a source of error under distributional shift,
when the correlations change at test-time. The prevailing sentiment in the
robustness literature is to avoid such correlative shortcut features and learn
robust predictors. However, while robust predictors perform better on
worst-case distributional shifts, they often sacrifice accuracy on majority
subpopulations. In this paper, we argue that shortcut features should not be
entirely discarded. Instead, if we can identify the subpopulation to which an
input belongs, we can adaptively choose among models with different strengths
to achieve high performance on both majority and minority subpopulations. We
propose COnfidence-baSed MOdel Selection (CosMoS), where we observe that model
confidence can effectively guide model selection. Notably, CosMoS does not
require any target labels or group annotations, either of which may be
difficult to obtain or unavailable. We evaluate CosMoS on four datasets with
spurious correlations, each with multiple test sets with varying levels of data
distribution shift. We find that CosMoS achieves 2-5% lower average regret
across all subpopulations, compared to using only robust predictors or other
model aggregation methods.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 18:48:15 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Chen",
"Annie S.",
""
],
[
"Lee",
"Yoonho",
""
],
[
"Setlur",
"Amrith",
""
],
[
"Levine",
"Sergey",
""
],
[
"Finn",
"Chelsea",
""
]
] | TITLE: Confidence-Based Model Selection: When to Take Shortcuts for
Subpopulation Shifts
ABSTRACT: Effective machine learning models learn both robust features that directly
determine the outcome of interest (e.g., an object with wheels is more likely
to be a car), and shortcut features (e.g., an object on a road is more likely
to be a car). The latter can be a source of error under distributional shift,
when the correlations change at test-time. The prevailing sentiment in the
robustness literature is to avoid such correlative shortcut features and learn
robust predictors. However, while robust predictors perform better on
worst-case distributional shifts, they often sacrifice accuracy on majority
subpopulations. In this paper, we argue that shortcut features should not be
entirely discarded. Instead, if we can identify the subpopulation to which an
input belongs, we can adaptively choose among models with different strengths
to achieve high performance on both majority and minority subpopulations. We
propose COnfidence-baSed MOdel Selection (CosMoS), where we observe that model
confidence can effectively guide model selection. Notably, CosMoS does not
require any target labels or group annotations, either of which may be
difficult to obtain or unavailable. We evaluate CosMoS on four datasets with
spurious correlations, each with multiple test sets with varying levels of data
distribution shift. We find that CosMoS achieves 2-5% lower average regret
across all subpopulations, compared to using only robust predictors or other
model aggregation methods.
|
2306.11122 | Yuan Chen | Yuan Chen, St\'ephane Marchand-Maillet | Supervised Auto-Encoding Twin-Bottleneck Hashing | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep hashing has shown to be a complexity-efficient solution for the
Approximate Nearest Neighbor search problem in high dimensional space. Many
methods usually build the loss function from pairwise or triplet data points to
capture the local similarity structure. Other existing methods construct the
similarity graph and consider all points simultaneously. Auto-encoding
Twin-bottleneck Hashing is one such method that dynamically builds the graph.
Specifically, each input data is encoded into a binary code and a continuous
variable, or the so-called twin bottlenecks. The similarity graph is then
computed from these binary codes, which get updated consistently during the
training. In this work, we generalize the original model into a supervised deep
hashing network by incorporating the label information. In addition, we examine
the differences of codes structure between these two networks and consider the
class imbalance problem especially in multi-labeled datasets. Experiments on
three datasets yield statistically significant improvement against the original
model. Results are also comparable and competitive to other supervised methods.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 18:50:02 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Chen",
"Yuan",
""
],
[
"Marchand-Maillet",
"Stéphane",
""
]
] | TITLE: Supervised Auto-Encoding Twin-Bottleneck Hashing
ABSTRACT: Deep hashing has shown to be a complexity-efficient solution for the
Approximate Nearest Neighbor search problem in high dimensional space. Many
methods usually build the loss function from pairwise or triplet data points to
capture the local similarity structure. Other existing methods construct the
similarity graph and consider all points simultaneously. Auto-encoding
Twin-bottleneck Hashing is one such method that dynamically builds the graph.
Specifically, each input data is encoded into a binary code and a continuous
variable, or the so-called twin bottlenecks. The similarity graph is then
computed from these binary codes, which get updated consistently during the
training. In this work, we generalize the original model into a supervised deep
hashing network by incorporating the label information. In addition, we examine
the differences of codes structure between these two networks and consider the
class imbalance problem especially in multi-labeled datasets. Experiments on
three datasets yield statistically significant improvement against the original
model. Results are also comparable and competitive to other supervised methods.
|
2306.11132 | Huaisheng Zhu | Huaisheng Zhu, Guoji Fu, Zhimeng Guo, Zhiwei Zhang, Teng Xiao, Suhang
Wang | Fairness-aware Message Passing for Graph Neural Networks | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Graph Neural Networks (GNNs) have shown great power in various domains.
However, their predictions may inherit societal biases on sensitive attributes,
limiting their adoption in real-world applications. Although many efforts have
been taken for fair GNNs, most existing works just adopt widely used fairness
techniques in machine learning to graph domains and ignore or don't have a
thorough understanding of the message passing mechanism with fairness
constraints, which is a distinctive feature of GNNs. To fill the gap, we
propose a novel fairness-aware message passing framework GMMD, which is derived
from an optimization problem that considers both graph smoothness and
representation fairness. GMMD can be intuitively interpreted as encouraging a
node to aggregate representations of other nodes from different sensitive
groups while subtracting representations of other nodes from the same sensitive
group, resulting in fair representations. We also provide a theoretical
analysis to justify that GMMD can guarantee fairness, which leads to a simpler
and theory-guided variant GMMD-S. Extensive experiments on graph benchmarks
show that our proposed framework can significantly improve the fairness of
various backbone GNN models while maintaining high accuracy.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 19:31:35 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Zhu",
"Huaisheng",
""
],
[
"Fu",
"Guoji",
""
],
[
"Guo",
"Zhimeng",
""
],
[
"Zhang",
"Zhiwei",
""
],
[
"Xiao",
"Teng",
""
],
[
"Wang",
"Suhang",
""
]
] | TITLE: Fairness-aware Message Passing for Graph Neural Networks
ABSTRACT: Graph Neural Networks (GNNs) have shown great power in various domains.
However, their predictions may inherit societal biases on sensitive attributes,
limiting their adoption in real-world applications. Although many efforts have
been taken for fair GNNs, most existing works just adopt widely used fairness
techniques in machine learning to graph domains and ignore or don't have a
thorough understanding of the message passing mechanism with fairness
constraints, which is a distinctive feature of GNNs. To fill the gap, we
propose a novel fairness-aware message passing framework GMMD, which is derived
from an optimization problem that considers both graph smoothness and
representation fairness. GMMD can be intuitively interpreted as encouraging a
node to aggregate representations of other nodes from different sensitive
groups while subtracting representations of other nodes from the same sensitive
group, resulting in fair representations. We also provide a theoretical
analysis to justify that GMMD can guarantee fairness, which leads to a simpler
and theory-guided variant GMMD-S. Extensive experiments on graph benchmarks
show that our proposed framework can significantly improve the fairness of
various backbone GNN models while maintaining high accuracy.
|
2306.11135 | Sam Aleyadeh | Sam Aleyadeh, Abbas Javadtalab, Abdallah Shami | Modular Simulation Environment Towards OTN AI-based Solutions | submitted to IDSTA2023 | null | null | null | cs.NI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The current trend for highly dynamic and virtualized networking
infrastructure made automated networking a critical requirement. Multiple
solutions have been proposed to address this, including the most sought-after
machine learning ML-based solutions. However, the main hurdle when developing
Next Generation Network is the availability of large datasets, especially in 5G
and beyond and Optical Transport Networking (OTN) traffic. This need led
researchers to look for viable simulation environments to generate the
necessary volume with highly configurable real-life scenarios, which can be
costly in setup and require subscription-based products and even the purchase
of dedicated hardware, depending on the supplier. We aim to address this issue
by generating high-volume and fidelity datasets by proposing a modular solution
to adapt to the user's available resources. These datasets can be used to
develop better-aforementioned ML solutions resulting in higher accuracy and
adaptation to real-life networking traffic.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 19:38:31 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Aleyadeh",
"Sam",
""
],
[
"Javadtalab",
"Abbas",
""
],
[
"Shami",
"Abdallah",
""
]
] | TITLE: Modular Simulation Environment Towards OTN AI-based Solutions
ABSTRACT: The current trend for highly dynamic and virtualized networking
infrastructure made automated networking a critical requirement. Multiple
solutions have been proposed to address this, including the most sought-after
machine learning ML-based solutions. However, the main hurdle when developing
Next Generation Network is the availability of large datasets, especially in 5G
and beyond and Optical Transport Networking (OTN) traffic. This need led
researchers to look for viable simulation environments to generate the
necessary volume with highly configurable real-life scenarios, which can be
costly in setup and require subscription-based products and even the purchase
of dedicated hardware, depending on the supplier. We aim to address this issue
by generating high-volume and fidelity datasets by proposing a modular solution
to adapt to the user's available resources. These datasets can be used to
develop better-aforementioned ML solutions resulting in higher accuracy and
adaptation to real-life networking traffic.
|
2306.11143 | Paolo Bonetti | Paolo Bonetti, Alberto Maria Metelli, Marcello Restelli | Nonlinear Feature Aggregation: Two Algorithms driven by Theory | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Many real-world machine learning applications are characterized by a huge
number of features, leading to computational and memory issues, as well as the
risk of overfitting. Ideally, only relevant and non-redundant features should
be considered to preserve the complete information of the original data and
limit the dimensionality. Dimensionality reduction and feature selection are
common preprocessing techniques addressing the challenge of efficiently dealing
with high-dimensional data. Dimensionality reduction methods control the number
of features in the dataset while preserving its structure and minimizing
information loss. Feature selection aims to identify the most relevant features
for a task, discarding the less informative ones. Previous works have proposed
approaches that aggregate features depending on their correlation without
discarding any of them and preserving their interpretability through
aggregation with the mean. A limitation of methods based on correlation is the
assumption of linearity in the relationship between features and target. In
this paper, we relax such an assumption in two ways. First, we propose a
bias-variance analysis for general models with additive Gaussian noise, leading
to a dimensionality reduction algorithm (NonLinCFA) which aggregates non-linear
transformations of features with a generic aggregation function. Then, we
extend the approach assuming that a generalized linear model regulates the
relationship between features and target. A deviance analysis leads to a second
dimensionality reduction algorithm (GenLinCFA), applicable to a larger class of
regression problems and classification settings. Finally, we test the
algorithms on synthetic and real-world datasets, performing regression and
classification tasks, showing competitive performances.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 19:57:33 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Bonetti",
"Paolo",
""
],
[
"Metelli",
"Alberto Maria",
""
],
[
"Restelli",
"Marcello",
""
]
] | TITLE: Nonlinear Feature Aggregation: Two Algorithms driven by Theory
ABSTRACT: Many real-world machine learning applications are characterized by a huge
number of features, leading to computational and memory issues, as well as the
risk of overfitting. Ideally, only relevant and non-redundant features should
be considered to preserve the complete information of the original data and
limit the dimensionality. Dimensionality reduction and feature selection are
common preprocessing techniques addressing the challenge of efficiently dealing
with high-dimensional data. Dimensionality reduction methods control the number
of features in the dataset while preserving its structure and minimizing
information loss. Feature selection aims to identify the most relevant features
for a task, discarding the less informative ones. Previous works have proposed
approaches that aggregate features depending on their correlation without
discarding any of them and preserving their interpretability through
aggregation with the mean. A limitation of methods based on correlation is the
assumption of linearity in the relationship between features and target. In
this paper, we relax such an assumption in two ways. First, we propose a
bias-variance analysis for general models with additive Gaussian noise, leading
to a dimensionality reduction algorithm (NonLinCFA) which aggregates non-linear
transformations of features with a generic aggregation function. Then, we
extend the approach assuming that a generalized linear model regulates the
relationship between features and target. A deviance analysis leads to a second
dimensionality reduction algorithm (GenLinCFA), applicable to a larger class of
regression problems and classification settings. Finally, we test the
algorithms on synthetic and real-world datasets, performing regression and
classification tasks, showing competitive performances.
|
2306.11152 | Xiaohao Cai | Jiahui Liu, Keqiang Fan, Xiaohao Cai and Mahesan Niranjan | Few-shot Learning for Inference in Medical Imaging with Subspace Feature
Representations | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by/4.0/ | Unlike the field of visual scene recognition where tremendous advances have
taken place due to the availability of very large datasets to train deep neural
networks, inference from medical images is often hampered by the fact that only
small amounts of data may be available. When working with very small dataset
problems, of the order of a few hundred items of data, the power of deep
learning may still be exploited by using a model pre-trained on natural images
as a feature extractor and carrying out classic pattern recognition techniques
in this feature space, the so-called few-shot learning problem. In regimes
where the dimension of this feature space is comparable to or even larger than
the number of items of data, dimensionality reduction is a necessity and is
often achieved by principal component analysis, i.e., singular value
decomposition (SVD). In this paper, noting the inappropriateness of using SVD
for this setting, we usher in and explore two alternatives based on
discriminant analysis and non-negative matrix factorization (NMF). Using 14
different datasets spanning $11$ distinct disease types, we demonstrate that
discriminant subspaces at low dimensions achieve significant improvements over
SVD-based subspaces and the original feature space. We also show that NMF at
modest dimensions is a competitive alternative to SVD in this setting.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 20:18:16 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Liu",
"Jiahui",
""
],
[
"Fan",
"Keqiang",
""
],
[
"Cai",
"Xiaohao",
""
],
[
"Niranjan",
"Mahesan",
""
]
] | TITLE: Few-shot Learning for Inference in Medical Imaging with Subspace Feature
Representations
ABSTRACT: Unlike the field of visual scene recognition where tremendous advances have
taken place due to the availability of very large datasets to train deep neural
networks, inference from medical images is often hampered by the fact that only
small amounts of data may be available. When working with very small dataset
problems, of the order of a few hundred items of data, the power of deep
learning may still be exploited by using a model pre-trained on natural images
as a feature extractor and carrying out classic pattern recognition techniques
in this feature space, the so-called few-shot learning problem. In regimes
where the dimension of this feature space is comparable to or even larger than
the number of items of data, dimensionality reduction is a necessity and is
often achieved by principal component analysis, i.e., singular value
decomposition (SVD). In this paper, noting the inappropriateness of using SVD
for this setting, we usher in and explore two alternatives based on
discriminant analysis and non-negative matrix factorization (NMF). Using 14
different datasets spanning $11$ distinct disease types, we demonstrate that
discriminant subspaces at low dimensions achieve significant improvements over
SVD-based subspaces and the original feature space. We also show that NMF at
modest dimensions is a competitive alternative to SVD in this setting.
|
2306.11159 | Paula Romero Jure | Paula V. Romero Jure and Sergio Masuelli and Juan Bautista Cabral | A labeled dataset of cloud types using data from GOES-16 and CloudSat | 6 pages, 7 figures. Submitted to 2022 IEEE Biennial Congress of
Argentina (ARGENCON). Presented in 2022 IEEE Biennial Congress of Argentina
(ARGENCON), San Juan, Argentina | 2022 IEEE Biennial Congress of Argentina (ARGENCON), 2022, pp. 1-6 | 10.1109/ARGENCON55245.2022.9940053 | null | astro-ph.EP cs.CV cs.LG physics.ao-ph | http://creativecommons.org/licenses/by/4.0/ | In this paper we present the development of a dataset consisting of 91
Multi-band Cloud and Moisture Product Full-Disk (MCMIPF) from the Advanced
Baseline Imager (ABI) on board GOES-16 geostationary satellite with 91
temporally and spatially corresponding CLDCLASS products from the CloudSat
polar satellite. The products are diurnal, corresponding to the months of
January and February 2019 and were chosen such that the products from both
satellites can be co-located over South America. The CLDCLASS product provides
the cloud type observed for each of the orbit's steps and the GOES-16 multiband
images contain pixels that can be co-located with these data. We develop an
algorithm that returns a product in the form of a table that provides pixels
from multiband images labelled with the type of cloud observed in them. These
labelled data conformed in this particular structure are very useful to perform
supervised learning. This was corroborated by training a simple linear
artificial neural network based on the work of Gorooh et al. (2020), which gave
good results, especially for the classification of deep convective clouds.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 20:55:37 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Jure",
"Paula V. Romero",
""
],
[
"Masuelli",
"Sergio",
""
],
[
"Cabral",
"Juan Bautista",
""
]
] | TITLE: A labeled dataset of cloud types using data from GOES-16 and CloudSat
ABSTRACT: In this paper we present the development of a dataset consisting of 91
Multi-band Cloud and Moisture Product Full-Disk (MCMIPF) from the Advanced
Baseline Imager (ABI) on board GOES-16 geostationary satellite with 91
temporally and spatially corresponding CLDCLASS products from the CloudSat
polar satellite. The products are diurnal, corresponding to the months of
January and February 2019 and were chosen such that the products from both
satellites can be co-located over South America. The CLDCLASS product provides
the cloud type observed for each of the orbit's steps and the GOES-16 multiband
images contain pixels that can be co-located with these data. We develop an
algorithm that returns a product in the form of a table that provides pixels
from multiband images labelled with the type of cloud observed in them. These
labelled data conformed in this particular structure are very useful to perform
supervised learning. This was corroborated by training a simple linear
artificial neural network based on the work of Gorooh et al. (2020), which gave
good results, especially for the classification of deep convective clouds.
|
2306.11173 | Ariel Lapid | Ariel Lapid, Idan Achituve, Lior Bracha, Ethan Fetaya | GD-VDM: Generated Depth for better Diffusion-based Video Generation | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The field of generative models has recently witnessed significant progress,
with diffusion models showing remarkable performance in image generation. In
light of this success, there is a growing interest in exploring the application
of diffusion models to other modalities. One such challenge is the generation
of coherent videos of complex scenes, which poses several technical
difficulties, such as capturing temporal dependencies and generating long,
high-resolution videos. This paper proposes GD-VDM, a novel diffusion model for
video generation, demonstrating promising results. GD-VDM is based on a
two-phase generation process involving generating depth videos followed by a
novel diffusion Vid2Vid model that generates a coherent real-world video. We
evaluated GD-VDM on the Cityscapes dataset and found that it generates more
diverse and complex scenes compared to natural baselines, demonstrating the
efficacy of our approach.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 21:32:10 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Lapid",
"Ariel",
""
],
[
"Achituve",
"Idan",
""
],
[
"Bracha",
"Lior",
""
],
[
"Fetaya",
"Ethan",
""
]
] | TITLE: GD-VDM: Generated Depth for better Diffusion-based Video Generation
ABSTRACT: The field of generative models has recently witnessed significant progress,
with diffusion models showing remarkable performance in image generation. In
light of this success, there is a growing interest in exploring the application
of diffusion models to other modalities. One such challenge is the generation
of coherent videos of complex scenes, which poses several technical
difficulties, such as capturing temporal dependencies and generating long,
high-resolution videos. This paper proposes GD-VDM, a novel diffusion model for
video generation, demonstrating promising results. GD-VDM is based on a
two-phase generation process involving generating depth videos followed by a
novel diffusion Vid2Vid model that generates a coherent real-world video. We
evaluated GD-VDM on the Cityscapes dataset and found that it generates more
diverse and complex scenes compared to natural baselines, demonstrating the
efficacy of our approach.
|
2306.11189 | Po-Ting Lai | Po-Ting Lai, Chih-Hsuan Wei, Ling Luo, Qingyu Chen, Zhiyong Lu | BioREx: Improving Biomedical Relation Extraction by Leveraging
Heterogeneous Datasets | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Biomedical relation extraction (RE) is the task of automatically identifying
and characterizing relations between biomedical concepts from free text. RE is
a central task in biomedical natural language processing (NLP) research and
plays a critical role in many downstream applications, such as literature-based
discovery and knowledge graph construction. State-of-the-art methods were used
primarily to train machine learning models on individual RE datasets, such as
protein-protein interaction and chemical-induced disease relation. Manual
dataset annotation, however, is highly expensive and time-consuming, as it
requires domain knowledge. Existing RE datasets are usually domain-specific or
small, which limits the development of generalized and high-performing RE
models. In this work, we present a novel framework for systematically
addressing the data heterogeneity of individual datasets and combining them
into a large dataset. Based on the framework and dataset, we report on BioREx,
a data-centric approach for extracting relations. Our evaluation shows that
BioREx achieves significantly higher performance than the benchmark system
trained on the individual dataset, setting a new SOTA from 74.4% to 79.6% in
F-1 measure on the recently released BioRED corpus. We further demonstrate that
the combined dataset can improve performance for five different RE tasks. In
addition, we show that on average BioREx compares favorably to current
best-performing methods such as transfer learning and multi-task learning.
Finally, we demonstrate BioREx's robustness and generalizability in two
independent RE tasks not previously seen in training data: drug-drug N-ary
combination and document-level gene-disease RE. The integrated dataset and
optimized method have been packaged as a stand-alone tool available at
https://github.com/ncbi/BioREx.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 22:48:18 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Lai",
"Po-Ting",
""
],
[
"Wei",
"Chih-Hsuan",
""
],
[
"Luo",
"Ling",
""
],
[
"Chen",
"Qingyu",
""
],
[
"Lu",
"Zhiyong",
""
]
] | TITLE: BioREx: Improving Biomedical Relation Extraction by Leveraging
Heterogeneous Datasets
ABSTRACT: Biomedical relation extraction (RE) is the task of automatically identifying
and characterizing relations between biomedical concepts from free text. RE is
a central task in biomedical natural language processing (NLP) research and
plays a critical role in many downstream applications, such as literature-based
discovery and knowledge graph construction. State-of-the-art methods were used
primarily to train machine learning models on individual RE datasets, such as
protein-protein interaction and chemical-induced disease relation. Manual
dataset annotation, however, is highly expensive and time-consuming, as it
requires domain knowledge. Existing RE datasets are usually domain-specific or
small, which limits the development of generalized and high-performing RE
models. In this work, we present a novel framework for systematically
addressing the data heterogeneity of individual datasets and combining them
into a large dataset. Based on the framework and dataset, we report on BioREx,
a data-centric approach for extracting relations. Our evaluation shows that
BioREx achieves significantly higher performance than the benchmark system
trained on the individual dataset, setting a new SOTA from 74.4% to 79.6% in
F-1 measure on the recently released BioRED corpus. We further demonstrate that
the combined dataset can improve performance for five different RE tasks. In
addition, we show that on average BioREx compares favorably to current
best-performing methods such as transfer learning and multi-task learning.
Finally, we demonstrate BioREx's robustness and generalizability in two
independent RE tasks not previously seen in training data: drug-drug N-ary
combination and document-level gene-disease RE. The integrated dataset and
optimized method have been packaged as a stand-alone tool available at
https://github.com/ncbi/BioREx.
|
2306.11190 | Penghang Liu | Penghang Liu, A. Erdem Sar{\i}y\"uce | Using Motif Transitions for Temporal Graph Generation | Accepted by 29th ACM SIGKDD Conference on Knowledge Discovery and
Data Mining (KDD2023) | null | 10.1145/3580305.3599540 | null | cs.SI cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph generative models are highly important for sharing surrogate data and
benchmarking purposes. Real-world complex systems often exhibit dynamic nature,
where the interactions among nodes change over time in the form of a temporal
network. Most temporal network generation models extend the static graph
generation models by incorporating temporality in the generation process. More
recently, temporal motifs are used to generate temporal networks with better
success. However, existing models are often restricted to a small set of
predefined motif patterns due to the high computational cost of counting
temporal motifs. In this work, we develop a practical temporal graph generator,
Motif Transition Model (MTM), to generate synthetic temporal networks with
realistic global and local features. Our key idea is modeling the arrival of
new events as temporal motif transition processes. We first calculate the
transition properties from the input graph and then simulate the motif
transition processes based on the transition probabilities and transition
rates. We demonstrate that our model consistently outperforms the baselines
with respect to preserving various global and local temporal graph statistics
and runtime performance.
| [
{
"version": "v1",
"created": "Mon, 19 Jun 2023 22:53:42 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Liu",
"Penghang",
""
],
[
"Sarıyüce",
"A. Erdem",
""
]
] | TITLE: Using Motif Transitions for Temporal Graph Generation
ABSTRACT: Graph generative models are highly important for sharing surrogate data and
benchmarking purposes. Real-world complex systems often exhibit dynamic nature,
where the interactions among nodes change over time in the form of a temporal
network. Most temporal network generation models extend the static graph
generation models by incorporating temporality in the generation process. More
recently, temporal motifs are used to generate temporal networks with better
success. However, existing models are often restricted to a small set of
predefined motif patterns due to the high computational cost of counting
temporal motifs. In this work, we develop a practical temporal graph generator,
Motif Transition Model (MTM), to generate synthetic temporal networks with
realistic global and local features. Our key idea is modeling the arrival of
new events as temporal motif transition processes. We first calculate the
transition properties from the input graph and then simulate the motif
transition processes based on the transition probabilities and transition
rates. We demonstrate that our model consistently outperforms the baselines
with respect to preserving various global and local temporal graph statistics
and runtime performance.
|
2306.11206 | Binbin Zhao | Binbin Zhao, Shouling Ji, Xuhong Zhang, Yuan Tian, Qinying Wang, Yuwen
Pu, Chenyang Lyu, Raheem Beyah | UVSCAN: Detecting Third-Party Component Usage Violations in IoT Firmware | Accepted as a full paper at USENIX Security '23 | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Nowadays, IoT devices integrate a wealth of third-party components (TPCs) in
firmware to shorten the development cycle. TPCs usually have strict usage
specifications, e.g., checking the return value of the function. Violating the
usage specifications of TPCs can cause serious consequences, e.g., NULL pointer
dereference. Therefore, this massive amount of TPC integrations, if not
properly implemented, will lead to pervasive vulnerabilities in IoT devices.
Detecting vulnerabilities automatically in TPC integration is challenging from
several perspectives: (1) There is a gap between the high-level specifications
from TPC documents, and the low-level implementations in the IoT firmware. (2)
IoT firmware is mostly the closed-source binary, which loses a lot of
information when compiling from the source code and has diverse architectures.
To address these challenges, we design and implement UVScan, an automated and
scalable system to detect TPC usage violations in IoT firmware. In UVScan, we
first propose a novel natural language processing (NLP)-based rule extraction
framework, which extracts API specifications from inconsistently formatted TPC
documents. We then design a rule-driven NLP-guided binary analysis engine,
which maps the logical information from the high-level TPC document to the
low-level binary, and detects TPC usage violations in IoT firmware across
different architectures. We evaluate UVScan from four perspectives on four
popular TPCs and six ground-truth datasets. The results show that UVScan
achieves more than 70% precision and recall, and has a significant performance
improvement compared with even the source-level API misuse detectors.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 00:13:33 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Zhao",
"Binbin",
""
],
[
"Ji",
"Shouling",
""
],
[
"Zhang",
"Xuhong",
""
],
[
"Tian",
"Yuan",
""
],
[
"Wang",
"Qinying",
""
],
[
"Pu",
"Yuwen",
""
],
[
"Lyu",
"Chenyang",
""
],
[
"Beyah",
"Raheem",
""
]
] | TITLE: UVSCAN: Detecting Third-Party Component Usage Violations in IoT Firmware
ABSTRACT: Nowadays, IoT devices integrate a wealth of third-party components (TPCs) in
firmware to shorten the development cycle. TPCs usually have strict usage
specifications, e.g., checking the return value of the function. Violating the
usage specifications of TPCs can cause serious consequences, e.g., NULL pointer
dereference. Therefore, this massive amount of TPC integrations, if not
properly implemented, will lead to pervasive vulnerabilities in IoT devices.
Detecting vulnerabilities automatically in TPC integration is challenging from
several perspectives: (1) There is a gap between the high-level specifications
from TPC documents, and the low-level implementations in the IoT firmware. (2)
IoT firmware is mostly the closed-source binary, which loses a lot of
information when compiling from the source code and has diverse architectures.
To address these challenges, we design and implement UVScan, an automated and
scalable system to detect TPC usage violations in IoT firmware. In UVScan, we
first propose a novel natural language processing (NLP)-based rule extraction
framework, which extracts API specifications from inconsistently formatted TPC
documents. We then design a rule-driven NLP-guided binary analysis engine,
which maps the logical information from the high-level TPC document to the
low-level binary, and detects TPC usage violations in IoT firmware across
different architectures. We evaluate UVScan from four perspectives on four
popular TPCs and six ground-truth datasets. The results show that UVScan
achieves more than 70% precision and recall, and has a significant performance
improvement compared with even the source-level API misuse detectors.
|
2306.11216 | Song Jiang | Song Jiang, Zijie Huang, Xiao Luo, Yizhou Sun | CF-GODE: Continuous-Time Causal Inference for Multi-Agent Dynamical
Systems | 13 pages, 8 figures | KDD 2023 | null | null | cs.LG stat.ME | http://creativecommons.org/licenses/by/4.0/ | Multi-agent dynamical systems refer to scenarios where multiple units
interact with each other and evolve collectively over time. To make informed
decisions in multi-agent dynamical systems, such as determining the optimal
vaccine distribution plan, it is essential for decision-makers to estimate the
continuous-time counterfactual outcomes. However, existing studies of causal
inference over time rely on the assumption that units are mutually independent,
which is not valid for multi-agent dynamical systems. In this paper, we aim to
bridge this gap and study how to estimate counterfactual outcomes in
multi-agent dynamical systems. Causal inference in a multi-agent dynamical
system has unique challenges: 1) Confounders are time-varying and are present
in both individual unit covariates and those of other units; 2) Units are
affected by not only their own but also others' treatments; 3) The treatments
are naturally dynamic, such as receiving vaccines and boosters in a seasonal
manner. We model a multi-agent dynamical system as a graph and propose
CounterFactual GraphODE (CF-GODE), a causal model that estimates
continuous-time counterfactual outcomes in the presence of inter-dependencies
between units. To facilitate continuous-time estimation, we propose
Treatment-Induced GraphODE, a novel ordinary differential equation based on
GNN, which incorporates dynamical treatments as additional inputs to predict
potential outcomes over time. To remove confounding bias, we propose two domain
adversarial learning based objectives that learn balanced continuous
representation trajectories, which are not predictive of treatments and
interference. We further provide theoretical justification to prove their
effectiveness. Experiments on two semi-synthetic datasets confirm that CF-GODE
outperforms baselines on counterfactual estimation. We also provide extensive
analyses to understand how our model works.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 00:50:09 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Jiang",
"Song",
""
],
[
"Huang",
"Zijie",
""
],
[
"Luo",
"Xiao",
""
],
[
"Sun",
"Yizhou",
""
]
] | TITLE: CF-GODE: Continuous-Time Causal Inference for Multi-Agent Dynamical
Systems
ABSTRACT: Multi-agent dynamical systems refer to scenarios where multiple units
interact with each other and evolve collectively over time. To make informed
decisions in multi-agent dynamical systems, such as determining the optimal
vaccine distribution plan, it is essential for decision-makers to estimate the
continuous-time counterfactual outcomes. However, existing studies of causal
inference over time rely on the assumption that units are mutually independent,
which is not valid for multi-agent dynamical systems. In this paper, we aim to
bridge this gap and study how to estimate counterfactual outcomes in
multi-agent dynamical systems. Causal inference in a multi-agent dynamical
system has unique challenges: 1) Confounders are time-varying and are present
in both individual unit covariates and those of other units; 2) Units are
affected by not only their own but also others' treatments; 3) The treatments
are naturally dynamic, such as receiving vaccines and boosters in a seasonal
manner. We model a multi-agent dynamical system as a graph and propose
CounterFactual GraphODE (CF-GODE), a causal model that estimates
continuous-time counterfactual outcomes in the presence of inter-dependencies
between units. To facilitate continuous-time estimation, we propose
Treatment-Induced GraphODE, a novel ordinary differential equation based on
GNN, which incorporates dynamical treatments as additional inputs to predict
potential outcomes over time. To remove confounding bias, we propose two domain
adversarial learning based objectives that learn balanced continuous
representation trajectories, which are not predictive of treatments and
interference. We further provide theoretical justification to prove their
effectiveness. Experiments on two semi-synthetic datasets confirm that CF-GODE
outperforms baselines on counterfactual estimation. We also provide extensive
analyses to understand how our model works.
|
2306.11247 | Alicia Parrish | Lora Aroyo, Alex S. Taylor, Mark Diaz, Christopher M. Homan, Alicia
Parrish, Greg Serapio-Garcia, Vinodkumar Prabhakaran, Ding Wang | DICES Dataset: Diversity in Conversational AI Evaluation for Safety | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Machine learning approaches often require training and evaluation datasets
with a clear separation between positive and negative examples. This risks
simplifying and even obscuring the inherent subjectivity present in many tasks.
Preserving such variance in content and diversity in datasets is often
expensive and laborious. This is especially troubling when building safety
datasets for conversational AI systems, as safety is both socially and
culturally situated. To demonstrate this crucial aspect of conversational AI
safety, and to facilitate in-depth model performance analyses, we introduce the
DICES (Diversity In Conversational AI Evaluation for Safety) dataset that
contains fine-grained demographic information about raters, high replication of
ratings per item to ensure statistical power for analyses, and encodes rater
votes as distributions across different demographics to allow for in-depth
explorations of different aggregation strategies. In short, the DICES dataset
enables the observation and measurement of variance, ambiguity, and diversity
in the context of conversational AI safety. We also illustrate how the dataset
offers a basis for establishing metrics to show how raters' ratings can
intersects with demographic categories such as racial/ethnic groups, age
groups, and genders. The goal of DICES is to be used as a shared resource and
benchmark that respects diverse perspectives during safety evaluation of
conversational AI systems.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 03:00:12 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Aroyo",
"Lora",
""
],
[
"Taylor",
"Alex S.",
""
],
[
"Diaz",
"Mark",
""
],
[
"Homan",
"Christopher M.",
""
],
[
"Parrish",
"Alicia",
""
],
[
"Serapio-Garcia",
"Greg",
""
],
[
"Prabhakaran",
"Vinodkumar",
""
],
[
"Wang",
"Ding",
""
]
] | TITLE: DICES Dataset: Diversity in Conversational AI Evaluation for Safety
ABSTRACT: Machine learning approaches often require training and evaluation datasets
with a clear separation between positive and negative examples. This risks
simplifying and even obscuring the inherent subjectivity present in many tasks.
Preserving such variance in content and diversity in datasets is often
expensive and laborious. This is especially troubling when building safety
datasets for conversational AI systems, as safety is both socially and
culturally situated. To demonstrate this crucial aspect of conversational AI
safety, and to facilitate in-depth model performance analyses, we introduce the
DICES (Diversity In Conversational AI Evaluation for Safety) dataset that
contains fine-grained demographic information about raters, high replication of
ratings per item to ensure statistical power for analyses, and encodes rater
votes as distributions across different demographics to allow for in-depth
explorations of different aggregation strategies. In short, the DICES dataset
enables the observation and measurement of variance, ambiguity, and diversity
in the context of conversational AI safety. We also illustrate how the dataset
offers a basis for establishing metrics to show how raters' ratings can
intersects with demographic categories such as racial/ethnic groups, age
groups, and genders. The goal of DICES is to be used as a shared resource and
benchmark that respects diverse perspectives during safety evaluation of
conversational AI systems.
|
2306.11256 | Yang Janet Liu | Yang Janet Liu and Amir Zeldes | GUMSum: Multi-Genre Data and Evaluation for English Abstractive
Summarization | Accepted to the Findings of ACL 2023; camera-ready version | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Automatic summarization with pre-trained language models has led to
impressively fluent results, but is prone to 'hallucinations', low performance
on non-news genres, and outputs which are not exactly summaries. Targeting ACL
2023's 'Reality Check' theme, we present GUMSum, a small but carefully crafted
dataset of English summaries in 12 written and spoken genres for evaluation of
abstractive summarization. Summaries are highly constrained, focusing on
substitutive potential, factuality, and faithfulness. We present guidelines and
evaluate human agreement as well as subjective judgments on recent system
outputs, comparing general-domain untuned approaches, a fine-tuned one, and a
prompt-based approach, to human performance. Results show that while GPT3
achieves impressive scores, it still underperforms humans, with varying quality
across genres. Human judgments reveal different types of errors in supervised,
prompted, and human-generated summaries, shedding light on the challenges of
producing a good summary.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 03:21:10 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Liu",
"Yang Janet",
""
],
[
"Zeldes",
"Amir",
""
]
] | TITLE: GUMSum: Multi-Genre Data and Evaluation for English Abstractive
Summarization
ABSTRACT: Automatic summarization with pre-trained language models has led to
impressively fluent results, but is prone to 'hallucinations', low performance
on non-news genres, and outputs which are not exactly summaries. Targeting ACL
2023's 'Reality Check' theme, we present GUMSum, a small but carefully crafted
dataset of English summaries in 12 written and spoken genres for evaluation of
abstractive summarization. Summaries are highly constrained, focusing on
substitutive potential, factuality, and faithfulness. We present guidelines and
evaluate human agreement as well as subjective judgments on recent system
outputs, comparing general-domain untuned approaches, a fine-tuned one, and a
prompt-based approach, to human performance. Results show that while GPT3
achieves impressive scores, it still underperforms humans, with varying quality
across genres. Human judgments reveal different types of errors in supervised,
prompted, and human-generated summaries, shedding light on the challenges of
producing a good summary.
|
2306.11264 | Wentao Zhao | Wentao Zhao, Qitian Wu, Chenxiao Yang and Junchi Yan | GraphGLOW: Universal and Generalizable Structure Learning for Graph
Neural Networks | Published as a conference paper at KDD 2023 | null | 10.1145/3580305.3599373 | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Graph structure learning is a well-established problem that aims at
optimizing graph structures adaptive to specific graph datasets to help message
passing neural networks (i.e., GNNs) to yield effective and robust node
embeddings. However, the common limitation of existing models lies in the
underlying \textit{closed-world assumption}: the testing graph is the same as
the training graph. This premise requires independently training the structure
learning model from scratch for each graph dataset, which leads to prohibitive
computation costs and potential risks for serious over-fitting. To mitigate
these issues, this paper explores a new direction that moves forward to learn a
universal structure learning model that can generalize across graph datasets in
an open world. We first introduce the mathematical definition of this novel
problem setting, and describe the model formulation from a probabilistic
data-generative aspect. Then we devise a general framework that coordinates a
single graph-shared structure learner and multiple graph-specific GNNs to
capture the generalizable patterns of optimal message-passing topology across
datasets. The well-trained structure learner can directly produce adaptive
structures for unseen target graphs without any fine-tuning. Across diverse
datasets and various challenging cross-graph generalization protocols, our
experiments show that even without training on target graphs, the proposed
model i) significantly outperforms expressive GNNs trained on input
(non-optimized) topology, and ii) surprisingly performs on par with
state-of-the-art models that independently optimize adaptive structures for
specific target graphs, with notably orders-of-magnitude acceleration for
training on the target graph.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 03:33:22 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Zhao",
"Wentao",
""
],
[
"Wu",
"Qitian",
""
],
[
"Yang",
"Chenxiao",
""
],
[
"Yan",
"Junchi",
""
]
] | TITLE: GraphGLOW: Universal and Generalizable Structure Learning for Graph
Neural Networks
ABSTRACT: Graph structure learning is a well-established problem that aims at
optimizing graph structures adaptive to specific graph datasets to help message
passing neural networks (i.e., GNNs) to yield effective and robust node
embeddings. However, the common limitation of existing models lies in the
underlying \textit{closed-world assumption}: the testing graph is the same as
the training graph. This premise requires independently training the structure
learning model from scratch for each graph dataset, which leads to prohibitive
computation costs and potential risks for serious over-fitting. To mitigate
these issues, this paper explores a new direction that moves forward to learn a
universal structure learning model that can generalize across graph datasets in
an open world. We first introduce the mathematical definition of this novel
problem setting, and describe the model formulation from a probabilistic
data-generative aspect. Then we devise a general framework that coordinates a
single graph-shared structure learner and multiple graph-specific GNNs to
capture the generalizable patterns of optimal message-passing topology across
datasets. The well-trained structure learner can directly produce adaptive
structures for unseen target graphs without any fine-tuning. Across diverse
datasets and various challenging cross-graph generalization protocols, our
experiments show that even without training on target graphs, the proposed
model i) significantly outperforms expressive GNNs trained on input
(non-optimized) topology, and ii) surprisingly performs on par with
state-of-the-art models that independently optimize adaptive structures for
specific target graphs, with notably orders-of-magnitude acceleration for
training on the target graph.
|
2306.11315 | Xiaojuan Zhang | Jun Fu and Xiaojuan Zhang and Shuang Li and Dali Chen | Variational Disentangled Graph Auto-Encoders for Link Prediction | null | null | null | null | cs.LG cs.AI cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the explosion of graph-structured data, link prediction has emerged as
an increasingly important task. Embedding methods for link prediction utilize
neural networks to generate node embeddings, which are subsequently employed to
predict links between nodes. However, the existing embedding methods typically
take a holistic strategy to learn node embeddings and ignore the entanglement
of latent factors. As a result, entangled embeddings fail to effectively
capture the underlying information and are vulnerable to irrelevant
information, leading to unconvincing and uninterpretable link prediction
results. To address these challenges, this paper proposes a novel framework
with two variants, the disentangled graph auto-encoder (DGAE) and the
variational disentangled graph auto-encoder (VDGAE). Our work provides a
pioneering effort to apply the disentanglement strategy to link prediction. The
proposed framework infers the latent factors that cause edges in the graph and
disentangles the representation into multiple channels corresponding to unique
latent factors, which contributes to improving the performance of link
prediction. To further encourage the embeddings to capture mutually exclusive
latent factors, we introduce mutual information regularization to enhance the
independence among different channels. Extensive experiments on various
real-world benchmarks demonstrate that our proposed methods achieve
state-of-the-art results compared to a variety of strong baselines on link
prediction tasks. Qualitative analysis on the synthetic dataset also
illustrates that the proposed methods can capture distinct latent factors that
cause links, providing empirical evidence that our models are able to explain
the results of link prediction to some extent. All code will be made publicly
available upon publication of the paper.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 06:25:05 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Fu",
"Jun",
""
],
[
"Zhang",
"Xiaojuan",
""
],
[
"Li",
"Shuang",
""
],
[
"Chen",
"Dali",
""
]
] | TITLE: Variational Disentangled Graph Auto-Encoders for Link Prediction
ABSTRACT: With the explosion of graph-structured data, link prediction has emerged as
an increasingly important task. Embedding methods for link prediction utilize
neural networks to generate node embeddings, which are subsequently employed to
predict links between nodes. However, the existing embedding methods typically
take a holistic strategy to learn node embeddings and ignore the entanglement
of latent factors. As a result, entangled embeddings fail to effectively
capture the underlying information and are vulnerable to irrelevant
information, leading to unconvincing and uninterpretable link prediction
results. To address these challenges, this paper proposes a novel framework
with two variants, the disentangled graph auto-encoder (DGAE) and the
variational disentangled graph auto-encoder (VDGAE). Our work provides a
pioneering effort to apply the disentanglement strategy to link prediction. The
proposed framework infers the latent factors that cause edges in the graph and
disentangles the representation into multiple channels corresponding to unique
latent factors, which contributes to improving the performance of link
prediction. To further encourage the embeddings to capture mutually exclusive
latent factors, we introduce mutual information regularization to enhance the
independence among different channels. Extensive experiments on various
real-world benchmarks demonstrate that our proposed methods achieve
state-of-the-art results compared to a variety of strong baselines on link
prediction tasks. Qualitative analysis on the synthetic dataset also
illustrates that the proposed methods can capture distinct latent factors that
cause links, providing empirical evidence that our models are able to explain
the results of link prediction to some extent. All code will be made publicly
available upon publication of the paper.
|
2306.11326 | Mitchell Rogers | Mitchell Rogers, Ga\"el Gendron, David Arturo Soriano Valdez, Mihailo
Azhar, Yang Chen, Shahrokh Heidari, Caleb Perelini, Padriac O'Leary, Kobe
Knowles, Izak Tait, Simon Eyre, Michael Witbrock, and Patrice Delmas | Meerkat Behaviour Recognition Dataset | Presented as a poster for the CV4Animals Workshop, CVPR 2023. For
associated dataset see: https://meerkat-dataset.github.io/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recording animal behaviour is an important step in evaluating the well-being
of animals and further understanding the natural world. Current methods for
documenting animal behaviour within a zoo setting, such as scan sampling,
require excessive human effort, are unfit for around-the-clock monitoring, and
may produce human-biased results. Several animal datasets already exist that
focus predominantly on wildlife interactions, with some extending to action or
behaviour recognition. However, there is limited data in a zoo setting or data
focusing on the group behaviours of social animals. We introduce a large
meerkat (Suricata Suricatta) behaviour recognition video dataset with diverse
annotated behaviours, including group social interactions, tracking of
individuals within the camera view, skewed class distribution, and varying
illumination conditions. This dataset includes videos from two positions within
the meerkat enclosure at the Wellington Zoo (Wellington, New Zealand), with
848,400 annotated frames across 20 videos and 15 unannotated videos.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 06:50:50 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Rogers",
"Mitchell",
""
],
[
"Gendron",
"Gaël",
""
],
[
"Valdez",
"David Arturo Soriano",
""
],
[
"Azhar",
"Mihailo",
""
],
[
"Chen",
"Yang",
""
],
[
"Heidari",
"Shahrokh",
""
],
[
"Perelini",
"Caleb",
""
],
[
"O'Leary",
"Padriac",
""
],
[
"Knowles",
"Kobe",
""
],
[
"Tait",
"Izak",
""
],
[
"Eyre",
"Simon",
""
],
[
"Witbrock",
"Michael",
""
],
[
"Delmas",
"Patrice",
""
]
] | TITLE: Meerkat Behaviour Recognition Dataset
ABSTRACT: Recording animal behaviour is an important step in evaluating the well-being
of animals and further understanding the natural world. Current methods for
documenting animal behaviour within a zoo setting, such as scan sampling,
require excessive human effort, are unfit for around-the-clock monitoring, and
may produce human-biased results. Several animal datasets already exist that
focus predominantly on wildlife interactions, with some extending to action or
behaviour recognition. However, there is limited data in a zoo setting or data
focusing on the group behaviours of social animals. We introduce a large
meerkat (Suricata Suricatta) behaviour recognition video dataset with diverse
annotated behaviours, including group social interactions, tracking of
individuals within the camera view, skewed class distribution, and varying
illumination conditions. This dataset includes videos from two positions within
the meerkat enclosure at the Wellington Zoo (Wellington, New Zealand), with
848,400 annotated frames across 20 videos and 15 unannotated videos.
|
2306.11334 | Yuxin Jin | Yuxin Jin, Ming Qian, Jincheng Xiong, Nan Xue, Gui-Song Xia | Depth and DOF Cues Make A Better Defocus Blur Detector | Code: https://github.com/yuxinjin-whu/D-DFFNet | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Defocus blur detection (DBD) separates in-focus and out-of-focus regions in
an image. Previous approaches mistakenly mistook homogeneous areas in focus for
defocus blur regions, likely due to not considering the internal factors that
cause defocus blur. Inspired by the law of depth, depth of field (DOF), and
defocus, we propose an approach called D-DFFNet, which incorporates depth and
DOF cues in an implicit manner. This allows the model to understand the defocus
phenomenon in a more natural way. Our method proposes a depth feature
distillation strategy to obtain depth knowledge from a pre-trained monocular
depth estimation model and uses a DOF-edge loss to understand the relationship
between DOF and depth. Our approach outperforms state-of-the-art methods on
public benchmarks and a newly collected large benchmark dataset, EBD. Source
codes and EBD dataset are available at: https:github.com/yuxinjin-whu/D-DFFNet.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 07:03:37 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Jin",
"Yuxin",
""
],
[
"Qian",
"Ming",
""
],
[
"Xiong",
"Jincheng",
""
],
[
"Xue",
"Nan",
""
],
[
"Xia",
"Gui-Song",
""
]
] | TITLE: Depth and DOF Cues Make A Better Defocus Blur Detector
ABSTRACT: Defocus blur detection (DBD) separates in-focus and out-of-focus regions in
an image. Previous approaches mistakenly mistook homogeneous areas in focus for
defocus blur regions, likely due to not considering the internal factors that
cause defocus blur. Inspired by the law of depth, depth of field (DOF), and
defocus, we propose an approach called D-DFFNet, which incorporates depth and
DOF cues in an implicit manner. This allows the model to understand the defocus
phenomenon in a more natural way. Our method proposes a depth feature
distillation strategy to obtain depth knowledge from a pre-trained monocular
depth estimation model and uses a DOF-edge loss to understand the relationship
between DOF and depth. Our approach outperforms state-of-the-art methods on
public benchmarks and a newly collected large benchmark dataset, EBD. Source
codes and EBD dataset are available at: https:github.com/yuxinjin-whu/D-DFFNet.
|
2306.11341 | Willy Fitra Hendria | Willy Fitra Hendria | MSVD-Indonesian: A Benchmark for Multimodal Video-Text Tasks in
Indonesian | 13 pages, 5 figures, 5 tables | null | null | null | cs.MM cs.CL cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Multimodal learning on video and text data has been receiving growing
attention from many researchers in various research tasks, including
text-to-video retrieval, video-to-text retrieval, and video captioning.
Although many algorithms have been proposed for those challenging tasks, most
of them are developed on English language datasets. Despite Indonesian being
one of the most spoken languages in the world, the research progress on the
multimodal video-text with Indonesian sentences is still under-explored, likely
due to the absence of the public benchmark dataset. To address this issue, we
construct the first public Indonesian video-text dataset by translating English
sentences from the MSVD dataset to Indonesian sentences. Using our dataset, we
then train neural network models which were developed for the English
video-text dataset on three tasks, i.e., text-to-video retrieval, video-to-text
retrieval, and video captioning. The recent neural network-based approaches to
video-text tasks often utilized a feature extractor that is primarily
pretrained on an English vision-language dataset. Since the availability of the
pretraining resources with Indonesian sentences is relatively limited, the
applicability of those approaches to our dataset is still questionable. To
overcome the lack of pretraining resources, we apply cross-lingual transfer
learning by utilizing the feature extractors pretrained on the English dataset,
and we then fine-tune the models on our Indonesian dataset. Our experimental
results show that this approach can help to improve the performance for the
three tasks on all metrics. Finally, we discuss potential future works using
our dataset, inspiring further research in the Indonesian multimodal video-text
tasks. We believe that our dataset and our experimental results could provide
valuable contributions to the community. Our dataset is available on GitHub.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 07:19:36 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Hendria",
"Willy Fitra",
""
]
] | TITLE: MSVD-Indonesian: A Benchmark for Multimodal Video-Text Tasks in
Indonesian
ABSTRACT: Multimodal learning on video and text data has been receiving growing
attention from many researchers in various research tasks, including
text-to-video retrieval, video-to-text retrieval, and video captioning.
Although many algorithms have been proposed for those challenging tasks, most
of them are developed on English language datasets. Despite Indonesian being
one of the most spoken languages in the world, the research progress on the
multimodal video-text with Indonesian sentences is still under-explored, likely
due to the absence of the public benchmark dataset. To address this issue, we
construct the first public Indonesian video-text dataset by translating English
sentences from the MSVD dataset to Indonesian sentences. Using our dataset, we
then train neural network models which were developed for the English
video-text dataset on three tasks, i.e., text-to-video retrieval, video-to-text
retrieval, and video captioning. The recent neural network-based approaches to
video-text tasks often utilized a feature extractor that is primarily
pretrained on an English vision-language dataset. Since the availability of the
pretraining resources with Indonesian sentences is relatively limited, the
applicability of those approaches to our dataset is still questionable. To
overcome the lack of pretraining resources, we apply cross-lingual transfer
learning by utilizing the feature extractors pretrained on the English dataset,
and we then fine-tune the models on our Indonesian dataset. Our experimental
results show that this approach can help to improve the performance for the
three tasks on all metrics. Finally, we discuss potential future works using
our dataset, inspiring further research in the Indonesian multimodal video-text
tasks. We believe that our dataset and our experimental results could provide
valuable contributions to the community. Our dataset is available on GitHub.
|
2306.11345 | Zhongzhen Huang | Zhongzhen Huang, Xiaofan Zhang, Shaoting Zhang | KiUT: Knowledge-injected U-Transformer for Radiology Report Generation | null | null | null | null | cs.CV cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Radiology report generation aims to automatically generate a clinically
accurate and coherent paragraph from the X-ray image, which could relieve
radiologists from the heavy burden of report writing. Although various image
caption methods have shown remarkable performance in the natural image field,
generating accurate reports for medical images requires knowledge of multiple
modalities, including vision, language, and medical terminology. We propose a
Knowledge-injected U-Transformer (KiUT) to learn multi-level visual
representation and adaptively distill the information with contextual and
clinical knowledge for word prediction. In detail, a U-connection schema
between the encoder and decoder is designed to model interactions between
different modalities. And a symptom graph and an injected knowledge distiller
are developed to assist the report generation. Experimentally, we outperform
state-of-the-art methods on two widely used benchmark datasets: IU-Xray and
MIMIC-CXR. Further experimental results prove the advantages of our
architecture and the complementary benefits of the injected knowledge.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 07:27:28 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Huang",
"Zhongzhen",
""
],
[
"Zhang",
"Xiaofan",
""
],
[
"Zhang",
"Shaoting",
""
]
] | TITLE: KiUT: Knowledge-injected U-Transformer for Radiology Report Generation
ABSTRACT: Radiology report generation aims to automatically generate a clinically
accurate and coherent paragraph from the X-ray image, which could relieve
radiologists from the heavy burden of report writing. Although various image
caption methods have shown remarkable performance in the natural image field,
generating accurate reports for medical images requires knowledge of multiple
modalities, including vision, language, and medical terminology. We propose a
Knowledge-injected U-Transformer (KiUT) to learn multi-level visual
representation and adaptively distill the information with contextual and
clinical knowledge for word prediction. In detail, a U-connection schema
between the encoder and decoder is designed to model interactions between
different modalities. And a symptom graph and an injected knowledge distiller
are developed to assist the report generation. Experimentally, we outperform
state-of-the-art methods on two widely used benchmark datasets: IU-Xray and
MIMIC-CXR. Further experimental results prove the advantages of our
architecture and the complementary benefits of the injected knowledge.
|
2306.11346 | Yu Zheng | Guangming Wang, Yu Zheng, Yanfeng Guo, Zhe Liu, Yixiang Zhu, Wolfram
Burgard, and Hesheng Wang | End-to-end 2D-3D Registration between Image and LiDAR Point Cloud for
Vehicle Localization | 18 pages, 14 figures, under review | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Robot localization using a previously built map is essential for a variety of
tasks including highly accurate navigation and mobile manipulation. A popular
approach to robot localization is based on image-to-point cloud registration,
which combines illumination-invariant LiDAR-based mapping with economical
image-based localization. However, the recent works for image-to-point cloud
registration either divide the registration into separate modules or project
the point cloud to the depth image to register the RGB and depth images. In
this paper, we present I2PNet, a novel end-to-end 2D-3D registration network.
I2PNet directly registers the raw 3D point cloud with the 2D RGB image using
differential modules with a unique target. The 2D-3D cost volume module for
differential 2D-3D association is proposed to bridge feature extraction and
pose regression. 2D-3D cost volume module implicitly constructs the soft
point-to-pixel correspondence on the intrinsic-independent normalized plane of
the pinhole camera model. Moreover, we introduce an outlier mask prediction
module to filter the outliers in the 2D-3D association before pose regression.
Furthermore, we propose the coarse-to-fine 2D-3D registration architecture to
increase localization accuracy. We conduct extensive localization experiments
on the KITTI Odometry and nuScenes datasets. The results demonstrate that
I2PNet outperforms the state-of-the-art by a large margin. In addition, I2PNet
has a higher efficiency than the previous works and can perform the
localization in real-time. Moreover, we extend the application of I2PNet to the
camera-LiDAR online calibration and demonstrate that I2PNet outperforms recent
approaches on the online calibration task.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 07:28:40 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Wang",
"Guangming",
""
],
[
"Zheng",
"Yu",
""
],
[
"Guo",
"Yanfeng",
""
],
[
"Liu",
"Zhe",
""
],
[
"Zhu",
"Yixiang",
""
],
[
"Burgard",
"Wolfram",
""
],
[
"Wang",
"Hesheng",
""
]
] | TITLE: End-to-end 2D-3D Registration between Image and LiDAR Point Cloud for
Vehicle Localization
ABSTRACT: Robot localization using a previously built map is essential for a variety of
tasks including highly accurate navigation and mobile manipulation. A popular
approach to robot localization is based on image-to-point cloud registration,
which combines illumination-invariant LiDAR-based mapping with economical
image-based localization. However, the recent works for image-to-point cloud
registration either divide the registration into separate modules or project
the point cloud to the depth image to register the RGB and depth images. In
this paper, we present I2PNet, a novel end-to-end 2D-3D registration network.
I2PNet directly registers the raw 3D point cloud with the 2D RGB image using
differential modules with a unique target. The 2D-3D cost volume module for
differential 2D-3D association is proposed to bridge feature extraction and
pose regression. 2D-3D cost volume module implicitly constructs the soft
point-to-pixel correspondence on the intrinsic-independent normalized plane of
the pinhole camera model. Moreover, we introduce an outlier mask prediction
module to filter the outliers in the 2D-3D association before pose regression.
Furthermore, we propose the coarse-to-fine 2D-3D registration architecture to
increase localization accuracy. We conduct extensive localization experiments
on the KITTI Odometry and nuScenes datasets. The results demonstrate that
I2PNet outperforms the state-of-the-art by a large margin. In addition, I2PNet
has a higher efficiency than the previous works and can perform the
localization in real-time. Moreover, we extend the application of I2PNet to the
camera-LiDAR online calibration and demonstrate that I2PNet outperforms recent
approaches on the online calibration task.
|
2306.11378 | Gongshu Wang | Ning Jiang, Gongshu Wang, and Tianyi Yan | Multi-task Collaborative Pre-training and Individual-adaptive-tokens
Fine-tuning: A Unified Framework for Brain Representation Learning | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Structural magnetic resonance imaging (sMRI) provides accurate estimates of
the brain's structural organization and learning invariant brain
representations from sMRI is an enduring issue in neuroscience. Previous deep
representation learning models ignore the fact that the brain, as the core of
human cognitive activity, is distinct from other organs whose primary attribute
is anatomy. Therefore, capturing the semantic structure that dominates
interindividual cognitive variability is key to accurately representing the
brain. Given that this high-level semantic information is subtle, distributed,
and interdependently latent in the brain structure, sMRI-based models need to
capture fine-grained details and understand how they relate to the overall
global structure. However, existing models are optimized by simple objectives,
making features collapse into homogeneity and worsening simultaneous
representation of fine-grained information and holistic semantics, causing a
lack of biological plausibility and interpretation of cognition. Here, we
propose MCIAT, a unified framework that combines Multi-task Collaborative
pre-training and Individual-Adaptive-Tokens fine-tuning. Specifically, we first
synthesize restorative learning, age prediction auxiliary learning and
adversarial learning as a joint proxy task for deep semantic representation
learning. Then, a mutual-attention-based token selection method is proposed to
highlight discriminative features. The proposed MCIAT achieves state-of-the-art
diagnosis performance on the ADHD-200 dataset compared with several sMRI-based
approaches and shows superior generalization on the MCIC and OASIS datasets.
Moreover, we studied 12 behavioral tasks and found significant associations
between cognitive functions and MCIAT-established representations, which
verifies the interpretability of our proposed framework.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 08:38:17 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Jiang",
"Ning",
""
],
[
"Wang",
"Gongshu",
""
],
[
"Yan",
"Tianyi",
""
]
] | TITLE: Multi-task Collaborative Pre-training and Individual-adaptive-tokens
Fine-tuning: A Unified Framework for Brain Representation Learning
ABSTRACT: Structural magnetic resonance imaging (sMRI) provides accurate estimates of
the brain's structural organization and learning invariant brain
representations from sMRI is an enduring issue in neuroscience. Previous deep
representation learning models ignore the fact that the brain, as the core of
human cognitive activity, is distinct from other organs whose primary attribute
is anatomy. Therefore, capturing the semantic structure that dominates
interindividual cognitive variability is key to accurately representing the
brain. Given that this high-level semantic information is subtle, distributed,
and interdependently latent in the brain structure, sMRI-based models need to
capture fine-grained details and understand how they relate to the overall
global structure. However, existing models are optimized by simple objectives,
making features collapse into homogeneity and worsening simultaneous
representation of fine-grained information and holistic semantics, causing a
lack of biological plausibility and interpretation of cognition. Here, we
propose MCIAT, a unified framework that combines Multi-task Collaborative
pre-training and Individual-Adaptive-Tokens fine-tuning. Specifically, we first
synthesize restorative learning, age prediction auxiliary learning and
adversarial learning as a joint proxy task for deep semantic representation
learning. Then, a mutual-attention-based token selection method is proposed to
highlight discriminative features. The proposed MCIAT achieves state-of-the-art
diagnosis performance on the ADHD-200 dataset compared with several sMRI-based
approaches and shows superior generalization on the MCIC and OASIS datasets.
Moreover, we studied 12 behavioral tasks and found significant associations
between cognitive functions and MCIAT-established representations, which
verifies the interpretability of our proposed framework.
|
2306.11389 | Rodrigo Diaz | Teresa Pelinski, Rodrigo Diaz, Ad\'an L. Benito Temprano and Andrew
McPherson | Pipeline for recording datasets and running neural networks on the Bela
embedded hardware platform | null | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Deploying deep learning models on embedded devices is an arduous task:
oftentimes, there exist no platform-specific instructions, and compilation
times can be considerably large due to the limited computational resources
available on-device. Moreover, many music-making applications demand real-time
inference. Embedded hardware platforms for audio, such as Bela, offer an entry
point for beginners into physical audio computing; however, the need for
cross-compilation environments and low-level software development tools for
deploying embedded deep learning models imposes high entry barriers on
non-expert users. We present a pipeline for deploying neural networks in the
Bela embedded hardware platform. In our pipeline, we include a tool to record a
multichannel dataset of sensor signals. Additionally, we provide a dockerised
cross-compilation environment for faster compilation. With this pipeline, we
aim to provide a template for programmers and makers to prototype and
experiment with neural networks for real-time embedded musical applications.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 08:55:04 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Pelinski",
"Teresa",
""
],
[
"Diaz",
"Rodrigo",
""
],
[
"Temprano",
"Adán L. Benito",
""
],
[
"McPherson",
"Andrew",
""
]
] | TITLE: Pipeline for recording datasets and running neural networks on the Bela
embedded hardware platform
ABSTRACT: Deploying deep learning models on embedded devices is an arduous task:
oftentimes, there exist no platform-specific instructions, and compilation
times can be considerably large due to the limited computational resources
available on-device. Moreover, many music-making applications demand real-time
inference. Embedded hardware platforms for audio, such as Bela, offer an entry
point for beginners into physical audio computing; however, the need for
cross-compilation environments and low-level software development tools for
deploying embedded deep learning models imposes high entry barriers on
non-expert users. We present a pipeline for deploying neural networks in the
Bela embedded hardware platform. In our pipeline, we include a tool to record a
multichannel dataset of sensor signals. Additionally, we provide a dockerised
cross-compilation environment for faster compilation. With this pipeline, we
aim to provide a template for programmers and makers to prototype and
experiment with neural networks for real-time embedded musical applications.
|
2306.11391 | Romain Lefeuvre | Romain Lefeuvre, Jessie Galasso, Benoit Combemale, Houari Sahraoui and
Stefano Zacchiroli | Fingerprinting and Building Large Reproducible Datasets | null | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Obtaining a relevant dataset is central to conducting empirical studies in
software engineering. However, in the context of mining software repositories,
the lack of appropriate tooling for large scale mining tasks hinders the
creation of new datasets. Moreover, limitations related to data sources that
change over time (e.g., code bases) and the lack of documentation of extraction
processes make it difficult to reproduce datasets over time. This threatens the
quality and reproducibility of empirical studies.
In this paper, we propose a tool-supported approach facilitating the creation
of large tailored datasets while ensuring their reproducibility. We leveraged
all the sources feeding the Software Heritage append-only archive which are
accessible through a unified programming interface to outline a reproducible
and generic extraction process. We propose a way to define a unique fingerprint
to characterize a dataset which, when provided to the extraction process,
ensures that the same dataset will be extracted.
We demonstrate the feasibility of our approach by implementing a prototype.
We show how it can help reduce the limitations researchers face when creating
or reproducing datasets.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 08:59:33 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Lefeuvre",
"Romain",
""
],
[
"Galasso",
"Jessie",
""
],
[
"Combemale",
"Benoit",
""
],
[
"Sahraoui",
"Houari",
""
],
[
"Zacchiroli",
"Stefano",
""
]
] | TITLE: Fingerprinting and Building Large Reproducible Datasets
ABSTRACT: Obtaining a relevant dataset is central to conducting empirical studies in
software engineering. However, in the context of mining software repositories,
the lack of appropriate tooling for large scale mining tasks hinders the
creation of new datasets. Moreover, limitations related to data sources that
change over time (e.g., code bases) and the lack of documentation of extraction
processes make it difficult to reproduce datasets over time. This threatens the
quality and reproducibility of empirical studies.
In this paper, we propose a tool-supported approach facilitating the creation
of large tailored datasets while ensuring their reproducibility. We leveraged
all the sources feeding the Software Heritage append-only archive which are
accessible through a unified programming interface to outline a reproducible
and generic extraction process. We propose a way to define a unique fingerprint
to characterize a dataset which, when provided to the extraction process,
ensures that the same dataset will be extracted.
We demonstrate the feasibility of our approach by implementing a prototype.
We show how it can help reduce the limitations researchers face when creating
or reproducing datasets.
|
2306.11397 | Thong Nguyen | Thong Nguyen, Andrew Yates | Generative Retrieval as Dense Retrieval | GenIR@SIGIR2023 | GenIR@SIGIR2023 | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Generative retrieval is a promising new neural retrieval paradigm that aims
to optimize the retrieval pipeline by performing both indexing and retrieval
with a single transformer model. However, this new paradigm faces challenges
with updating the index and scaling to large collections. In this paper, we
analyze two prominent variants of generative retrieval and show that they can
be conceptually viewed as bi-encoders for dense retrieval. Specifically, we
analytically demonstrate that the generative retrieval process can be
decomposed into dot products between query and document vectors, similar to
dense retrieval. This analysis leads us to propose a new variant of generative
retrieval, called Tied-Atomic, which addresses the updating and scaling issues
by incorporating techniques from dense retrieval. In experiments on two
datasets, NQ320k and the full MSMARCO, we confirm that this approach does not
reduce retrieval effectiveness while enabling the model to scale to large
collections.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 09:04:06 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Nguyen",
"Thong",
""
],
[
"Yates",
"Andrew",
""
]
] | TITLE: Generative Retrieval as Dense Retrieval
ABSTRACT: Generative retrieval is a promising new neural retrieval paradigm that aims
to optimize the retrieval pipeline by performing both indexing and retrieval
with a single transformer model. However, this new paradigm faces challenges
with updating the index and scaling to large collections. In this paper, we
analyze two prominent variants of generative retrieval and show that they can
be conceptually viewed as bi-encoders for dense retrieval. Specifically, we
analytically demonstrate that the generative retrieval process can be
decomposed into dot products between query and document vectors, similar to
dense retrieval. This analysis leads us to propose a new variant of generative
retrieval, called Tied-Atomic, which addresses the updating and scaling issues
by incorporating techniques from dense retrieval. In experiments on two
datasets, NQ320k and the full MSMARCO, we confirm that this approach does not
reduce retrieval effectiveness while enabling the model to scale to large
collections.
|
2306.11417 | Chenghao Liu | Chenghao Liu, Wenzhuo Yang, Himanshu Mittal, Manpreet Singh, Doyen
Sahoo, Steven C. H. Hoi | PyRCA: A Library for Metric-based Root Cause Analysis | Github repo: https://github.com/salesforce/PyRCA | null | null | null | cs.AI cs.LG cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce PyRCA, an open-source Python machine learning library of Root
Cause Analysis (RCA) for Artificial Intelligence for IT Operations (AIOps). It
provides a holistic framework to uncover the complicated metric causal
dependencies and automatically locate root causes of incidents. It offers a
unified interface for multiple commonly used RCA models, encompassing both
graph construction and scoring tasks. This library aims to provide IT
operations staff, data scientists, and researchers a one-step solution to rapid
model development, model evaluation and deployment to online applications. In
particular, our library includes various causal discovery methods to support
causal graph construction, and multiple types of root cause scoring methods
inspired by Bayesian analysis, graph analysis and causal analysis, etc. Our GUI
dashboard offers practitioners an intuitive point-and-click interface,
empowering them to easily inject expert knowledge through human interaction.
With the ability to visualize causal graphs and the root cause of incidents,
practitioners can quickly gain insights and improve their workflow efficiency.
This technical report introduces PyRCA's architecture and major
functionalities, while also presenting benchmark performance numbers in
comparison to various baseline models. Additionally, we demonstrate PyRCA's
capabilities through several example use cases.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 09:55:10 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Liu",
"Chenghao",
""
],
[
"Yang",
"Wenzhuo",
""
],
[
"Mittal",
"Himanshu",
""
],
[
"Singh",
"Manpreet",
""
],
[
"Sahoo",
"Doyen",
""
],
[
"Hoi",
"Steven C. H.",
""
]
] | TITLE: PyRCA: A Library for Metric-based Root Cause Analysis
ABSTRACT: We introduce PyRCA, an open-source Python machine learning library of Root
Cause Analysis (RCA) for Artificial Intelligence for IT Operations (AIOps). It
provides a holistic framework to uncover the complicated metric causal
dependencies and automatically locate root causes of incidents. It offers a
unified interface for multiple commonly used RCA models, encompassing both
graph construction and scoring tasks. This library aims to provide IT
operations staff, data scientists, and researchers a one-step solution to rapid
model development, model evaluation and deployment to online applications. In
particular, our library includes various causal discovery methods to support
causal graph construction, and multiple types of root cause scoring methods
inspired by Bayesian analysis, graph analysis and causal analysis, etc. Our GUI
dashboard offers practitioners an intuitive point-and-click interface,
empowering them to easily inject expert knowledge through human interaction.
With the ability to visualize causal graphs and the root cause of incidents,
practitioners can quickly gain insights and improve their workflow efficiency.
This technical report introduces PyRCA's architecture and major
functionalities, while also presenting benchmark performance numbers in
comparison to various baseline models. Additionally, we demonstrate PyRCA's
capabilities through several example use cases.
|
2306.11420 | Zi Wang | Zi Wang and Daniel Hershcovich | On Evaluating Multilingual Compositional Generalization with Translated
Datasets | ACL 2023 long paper | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Compositional generalization allows efficient learning and human-like
inductive biases. Since most research investigating compositional
generalization in NLP is done on English, important questions remain
underexplored. Do the necessary compositional generalization abilities differ
across languages? Can models compositionally generalize cross-lingually? As a
first step to answering these questions, recent work used neural machine
translation to translate datasets for evaluating compositional generalization
in semantic parsing. However, we show that this entails critical semantic
distortion. To address this limitation, we craft a faithful rule-based
translation of the MCWQ dataset from English to Chinese and Japanese. Even with
the resulting robust benchmark, which we call MCWQ-R, we show that the
distribution of compositions still suffers due to linguistic divergences, and
that multilingual models still struggle with cross-lingual compositional
generalization. Our dataset and methodology will be useful resources for the
study of cross-lingual compositional generalization in other tasks.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 10:03:57 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Wang",
"Zi",
""
],
[
"Hershcovich",
"Daniel",
""
]
] | TITLE: On Evaluating Multilingual Compositional Generalization with Translated
Datasets
ABSTRACT: Compositional generalization allows efficient learning and human-like
inductive biases. Since most research investigating compositional
generalization in NLP is done on English, important questions remain
underexplored. Do the necessary compositional generalization abilities differ
across languages? Can models compositionally generalize cross-lingually? As a
first step to answering these questions, recent work used neural machine
translation to translate datasets for evaluating compositional generalization
in semantic parsing. However, we show that this entails critical semantic
distortion. To address this limitation, we craft a faithful rule-based
translation of the MCWQ dataset from English to Chinese and Japanese. Even with
the resulting robust benchmark, which we call MCWQ-R, we show that the
distribution of compositions still suffers due to linguistic divergences, and
that multilingual models still struggle with cross-lingual compositional
generalization. Our dataset and methodology will be useful resources for the
study of cross-lingual compositional generalization in other tasks.
|
2306.11435 | N M Anoop Krishnan | Suresh Bishnoi, Jayadeva, Sayan Ranu, N. M. Anoop Krishnan | Graph Neural Stochastic Differential Equations for Learning Brownian
Dynamics | null | null | null | null | cs.LG cond-mat.dis-nn | http://creativecommons.org/licenses/by-sa/4.0/ | Neural networks (NNs) that exploit strong inductive biases based on physical
laws and symmetries have shown remarkable success in learning the dynamics of
physical systems directly from their trajectory. However, these works focus
only on the systems that follow deterministic dynamics, for instance, Newtonian
or Hamiltonian dynamics. Here, we propose a framework, namely Brownian graph
neural networks (BROGNET), combining stochastic differential equations (SDEs)
and GNNs to learn Brownian dynamics directly from the trajectory. We
theoretically show that BROGNET conserves the linear momentum of the system,
which in turn, provides superior performance on learning dynamics as revealed
empirically. We demonstrate this approach on several systems, namely, linear
spring, linear spring with binary particle types, and non-linear spring
systems, all following Brownian dynamics at finite temperatures. We show that
BROGNET significantly outperforms proposed baselines across all the benchmarked
Brownian systems. In addition, we demonstrate zero-shot generalizability of
BROGNET to simulate unseen system sizes that are two orders of magnitude larger
and to different temperatures than those used during training. Altogether, our
study contributes to advancing the understanding of the intricate dynamics of
Brownian motion and demonstrates the effectiveness of graph neural networks in
modeling such complex systems.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 10:30:46 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Bishnoi",
"Suresh",
""
],
[
"Jayadeva",
"",
""
],
[
"Ranu",
"Sayan",
""
],
[
"Krishnan",
"N. M. Anoop",
""
]
] | TITLE: Graph Neural Stochastic Differential Equations for Learning Brownian
Dynamics
ABSTRACT: Neural networks (NNs) that exploit strong inductive biases based on physical
laws and symmetries have shown remarkable success in learning the dynamics of
physical systems directly from their trajectory. However, these works focus
only on the systems that follow deterministic dynamics, for instance, Newtonian
or Hamiltonian dynamics. Here, we propose a framework, namely Brownian graph
neural networks (BROGNET), combining stochastic differential equations (SDEs)
and GNNs to learn Brownian dynamics directly from the trajectory. We
theoretically show that BROGNET conserves the linear momentum of the system,
which in turn, provides superior performance on learning dynamics as revealed
empirically. We demonstrate this approach on several systems, namely, linear
spring, linear spring with binary particle types, and non-linear spring
systems, all following Brownian dynamics at finite temperatures. We show that
BROGNET significantly outperforms proposed baselines across all the benchmarked
Brownian systems. In addition, we demonstrate zero-shot generalizability of
BROGNET to simulate unseen system sizes that are two orders of magnitude larger
and to different temperatures than those used during training. Altogether, our
study contributes to advancing the understanding of the intricate dynamics of
Brownian motion and demonstrates the effectiveness of graph neural networks in
modeling such complex systems.
|
2306.11444 | Paola Merlo | Paola Merlo | Blackbird language matrices (BLM), a new task for rule-like
generalization in neural networks: Motivations and Formal Specifications | 7pages, 6 figures. arXiv admin note: text overlap with
arXiv:2205.10866 | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We motivate and formally define a new task for fine-tuning rule-like
generalization in large language models. It is conjectured that the
shortcomings of current LLMs are due to a lack of ability to generalize. It has
been argued that, instead, humans are better at generalization because they
have a tendency at extracting rules from complex data. We try to recreate this
tendency to rule-based generalization. When exposed to tests of analytic
intelligence, for example, the visual RAVEN IQ test, human problem-solvers
identify the relevant objects in the picture and their relevant attributes and
reason based on rules applied to these objects and attributes. Based on the
induced rules, they are able to provide a solution to the test. We propose a
task that translates this IQ task into language. In this paper, we provide the
formal specification for the task and the generative process of its datasets.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 10:45:56 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Merlo",
"Paola",
""
]
] | TITLE: Blackbird language matrices (BLM), a new task for rule-like
generalization in neural networks: Motivations and Formal Specifications
ABSTRACT: We motivate and formally define a new task for fine-tuning rule-like
generalization in large language models. It is conjectured that the
shortcomings of current LLMs are due to a lack of ability to generalize. It has
been argued that, instead, humans are better at generalization because they
have a tendency at extracting rules from complex data. We try to recreate this
tendency to rule-based generalization. When exposed to tests of analytic
intelligence, for example, the visual RAVEN IQ test, human problem-solvers
identify the relevant objects in the picture and their relevant attributes and
reason based on rules applied to these objects and attributes. Based on the
induced rules, they are able to provide a solution to the test. We propose a
task that translates this IQ task into language. In this paper, we provide the
formal specification for the task and the generative process of its datasets.
|
2306.11472 | Pratik Nag | Pratik Nag, Ying Sun, Brian J Reich | Spatio-temporal DeepKriging for Interpolation and Probabilistic
Forecasting | null | null | null | null | stat.ML cs.LG stat.CO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gaussian processes (GP) and Kriging are widely used in traditional
spatio-temporal mod-elling and prediction. These techniques typically
presuppose that the data are observed from a stationary GP with parametric
covariance structure. However, processes in real-world applications often
exhibit non-Gaussianity and nonstationarity. Moreover, likelihood-based
inference for GPs is computationally expensive and thus prohibitive for large
datasets. In this paper we propose a deep neural network (DNN) based two-stage
model for spatio-temporal interpolation and forecasting. Interpolation is
performed in the first step, which utilizes a dependent DNN with the embedding
layer constructed with spatio-temporal basis functions. For the second stage,
we use Long-Short Term Memory (LSTM) and convolutional LSTM to forecast future
observations at a given location. We adopt the quantile-based loss function in
the DNN to provide probabilistic forecasting. Compared to Kriging, the proposed
method does not require specifying covariance functions or making stationarity
assumption, and is computationally efficient. Therefore, it is suitable for
large-scale prediction of complex spatio-temporal processes. We apply our
method to monthly $PM_{2.5}$ data at more than $200,000$ space-time locations
from January 1999 to December 2022 for fast imputation of missing values and
forecasts with uncertainties.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 11:51:44 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Nag",
"Pratik",
""
],
[
"Sun",
"Ying",
""
],
[
"Reich",
"Brian J",
""
]
] | TITLE: Spatio-temporal DeepKriging for Interpolation and Probabilistic
Forecasting
ABSTRACT: Gaussian processes (GP) and Kriging are widely used in traditional
spatio-temporal mod-elling and prediction. These techniques typically
presuppose that the data are observed from a stationary GP with parametric
covariance structure. However, processes in real-world applications often
exhibit non-Gaussianity and nonstationarity. Moreover, likelihood-based
inference for GPs is computationally expensive and thus prohibitive for large
datasets. In this paper we propose a deep neural network (DNN) based two-stage
model for spatio-temporal interpolation and forecasting. Interpolation is
performed in the first step, which utilizes a dependent DNN with the embedding
layer constructed with spatio-temporal basis functions. For the second stage,
we use Long-Short Term Memory (LSTM) and convolutional LSTM to forecast future
observations at a given location. We adopt the quantile-based loss function in
the DNN to provide probabilistic forecasting. Compared to Kriging, the proposed
method does not require specifying covariance functions or making stationarity
assumption, and is computationally efficient. Therefore, it is suitable for
large-scale prediction of complex spatio-temporal processes. We apply our
method to monthly $PM_{2.5}$ data at more than $200,000$ space-time locations
from January 1999 to December 2022 for fast imputation of missing values and
forecasts with uncertainties.
|
2306.11477 | Liang Li | Liang Li, Ruiying Geng, Chengyang Fang, Bing Li, Can Ma, Rongyu Cao,
Binhua Li, Fei Huang, Yongbin Li | CATS: A Pragmatic Chinese Answer-to-Sequence Dataset with Large Scale
and High Quality | ACL 2023 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | There are three problems existing in the popular data-to-text datasets.
First, the large-scale datasets either contain noise or lack real application
scenarios. Second, the datasets close to real applications are relatively small
in size. Last, current datasets bias in the English language while leaving
other languages underexplored. To alleviate these limitations, in this paper,
we present CATS, a pragmatic Chinese answer-to-sequence dataset with large
scale and high quality. The dataset aims to generate textual descriptions for
the answer in the practical TableQA system. Further, to bridge the structural
gap between the input SQL and table and establish better semantic alignments,
we propose a Unified Graph Transformation approach to establish a joint
encoding space for the two hybrid knowledge resources and convert this task to
a graph-to-text problem. The experiment results demonstrate the effectiveness
of our proposed method. Further analysis on CATS attests to both the high
quality and challenges of the dataset.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 12:02:26 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Li",
"Liang",
""
],
[
"Geng",
"Ruiying",
""
],
[
"Fang",
"Chengyang",
""
],
[
"Li",
"Bing",
""
],
[
"Ma",
"Can",
""
],
[
"Cao",
"Rongyu",
""
],
[
"Li",
"Binhua",
""
],
[
"Huang",
"Fei",
""
],
[
"Li",
"Yongbin",
""
]
] | TITLE: CATS: A Pragmatic Chinese Answer-to-Sequence Dataset with Large Scale
and High Quality
ABSTRACT: There are three problems existing in the popular data-to-text datasets.
First, the large-scale datasets either contain noise or lack real application
scenarios. Second, the datasets close to real applications are relatively small
in size. Last, current datasets bias in the English language while leaving
other languages underexplored. To alleviate these limitations, in this paper,
we present CATS, a pragmatic Chinese answer-to-sequence dataset with large
scale and high quality. The dataset aims to generate textual descriptions for
the answer in the practical TableQA system. Further, to bridge the structural
gap between the input SQL and table and establish better semantic alignments,
we propose a Unified Graph Transformation approach to establish a joint
encoding space for the two hybrid knowledge resources and convert this task to
a graph-to-text problem. The experiment results demonstrate the effectiveness
of our proposed method. Further analysis on CATS attests to both the high
quality and challenges of the dataset.
|
2306.11487 | Pratik Nag | Pratik Nag, Yiping Hong, Sameh Abdulah, Ghulam A. Qadir, Marc G.
Genton, and Ying Sun | Efficient Large-scale Nonstationary Spatial Covariance Function
Estimation Using Convolutional Neural Networks | null | null | null | null | stat.ML cs.LG stat.CO | http://creativecommons.org/licenses/by/4.0/ | Spatial processes observed in various fields, such as climate and
environmental science, often occur on a large scale and demonstrate spatial
nonstationarity. Fitting a Gaussian process with a nonstationary Mat\'ern
covariance is challenging. Previous studies in the literature have tackled this
challenge by employing spatial partitioning techniques to estimate the
parameters that vary spatially in the covariance function. The selection of
partitions is an important consideration, but it is often subjective and lacks
a data-driven approach. To address this issue, in this study, we utilize the
power of Convolutional Neural Networks (ConvNets) to derive subregions from the
nonstationary data. We employ a selection mechanism to identify subregions that
exhibit similar behavior to stationary fields. In order to distinguish between
stationary and nonstationary random fields, we conducted training on ConvNet
using various simulated data. These simulations are generated from Gaussian
processes with Mat\'ern covariance models under a wide range of parameter
settings, ensuring adequate representation of both stationary and nonstationary
spatial data. We assess the performance of the proposed method with synthetic
and real datasets at a large scale. The results revealed enhanced accuracy in
parameter estimations when relying on ConvNet-based partition compared to
traditional user-defined approaches.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 12:17:46 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Nag",
"Pratik",
""
],
[
"Hong",
"Yiping",
""
],
[
"Abdulah",
"Sameh",
""
],
[
"Qadir",
"Ghulam A.",
""
],
[
"Genton",
"Marc G.",
""
],
[
"Sun",
"Ying",
""
]
] | TITLE: Efficient Large-scale Nonstationary Spatial Covariance Function
Estimation Using Convolutional Neural Networks
ABSTRACT: Spatial processes observed in various fields, such as climate and
environmental science, often occur on a large scale and demonstrate spatial
nonstationarity. Fitting a Gaussian process with a nonstationary Mat\'ern
covariance is challenging. Previous studies in the literature have tackled this
challenge by employing spatial partitioning techniques to estimate the
parameters that vary spatially in the covariance function. The selection of
partitions is an important consideration, but it is often subjective and lacks
a data-driven approach. To address this issue, in this study, we utilize the
power of Convolutional Neural Networks (ConvNets) to derive subregions from the
nonstationary data. We employ a selection mechanism to identify subregions that
exhibit similar behavior to stationary fields. In order to distinguish between
stationary and nonstationary random fields, we conducted training on ConvNet
using various simulated data. These simulations are generated from Gaussian
processes with Mat\'ern covariance models under a wide range of parameter
settings, ensuring adequate representation of both stationary and nonstationary
spatial data. We assess the performance of the proposed method with synthetic
and real datasets at a large scale. The results revealed enhanced accuracy in
parameter estimations when relying on ConvNet-based partition compared to
traditional user-defined approaches.
|
2306.11504 | Zeyue Xue | Yue Yang, Kaipeng Zhang, Yuying Ge, Wenqi Shao, Zeyue Xue, Yu Qiao,
Ping Luo | Align, Adapt and Inject: Sound-guided Unified Image Generation | Tech Report | null | null | null | cs.GR cs.CV cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Text-guided image generation has witnessed unprecedented progress due to the
development of diffusion models. Beyond text and image, sound is a vital
element within the sphere of human perception, offering vivid representations
and naturally coinciding with corresponding scenes. Taking advantage of sound
therefore presents a promising avenue for exploration within image generation
research. However, the relationship between audio and image supervision remains
significantly underdeveloped, and the scarcity of related, high-quality
datasets brings further obstacles. In this paper, we propose a unified
framework 'Align, Adapt, and Inject' (AAI) for sound-guided image generation,
editing, and stylization. In particular, our method adapts input sound into a
sound token, like an ordinary word, which can plug and play with existing
powerful diffusion-based Text-to-Image (T2I) models. Specifically, we first
train a multi-modal encoder to align audio representation with the pre-trained
textual manifold and visual manifold, respectively. Then, we propose the audio
adapter to adapt audio representation into an audio token enriched with
specific semantics, which can be injected into a frozen T2I model flexibly. In
this way, we are able to extract the dynamic information of varied sounds,
while utilizing the formidable capability of existing T2I models to facilitate
sound-guided image generation, editing, and stylization in a convenient and
cost-effective manner. The experiment results confirm that our proposed AAI
outperforms other text and sound-guided state-of-the-art methods. And our
aligned multi-modal encoder is also competitive with other approaches in the
audio-visual retrieval and audio-text retrieval tasks.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 12:50:49 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Yang",
"Yue",
""
],
[
"Zhang",
"Kaipeng",
""
],
[
"Ge",
"Yuying",
""
],
[
"Shao",
"Wenqi",
""
],
[
"Xue",
"Zeyue",
""
],
[
"Qiao",
"Yu",
""
],
[
"Luo",
"Ping",
""
]
] | TITLE: Align, Adapt and Inject: Sound-guided Unified Image Generation
ABSTRACT: Text-guided image generation has witnessed unprecedented progress due to the
development of diffusion models. Beyond text and image, sound is a vital
element within the sphere of human perception, offering vivid representations
and naturally coinciding with corresponding scenes. Taking advantage of sound
therefore presents a promising avenue for exploration within image generation
research. However, the relationship between audio and image supervision remains
significantly underdeveloped, and the scarcity of related, high-quality
datasets brings further obstacles. In this paper, we propose a unified
framework 'Align, Adapt, and Inject' (AAI) for sound-guided image generation,
editing, and stylization. In particular, our method adapts input sound into a
sound token, like an ordinary word, which can plug and play with existing
powerful diffusion-based Text-to-Image (T2I) models. Specifically, we first
train a multi-modal encoder to align audio representation with the pre-trained
textual manifold and visual manifold, respectively. Then, we propose the audio
adapter to adapt audio representation into an audio token enriched with
specific semantics, which can be injected into a frozen T2I model flexibly. In
this way, we are able to extract the dynamic information of varied sounds,
while utilizing the formidable capability of existing T2I models to facilitate
sound-guided image generation, editing, and stylization in a convenient and
cost-effective manner. The experiment results confirm that our proposed AAI
outperforms other text and sound-guided state-of-the-art methods. And our
aligned multi-modal encoder is also competitive with other approaches in the
audio-visual retrieval and audio-text retrieval tasks.
|
2306.11507 | Yue Huang | Yue Huang and Qihui Zhang and Philip S. Y and Lichao Sun | TrustGPT: A Benchmark for Trustworthy and Responsible Large Language
Models | We are currently expanding this work and welcome collaborators! | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large Language Models (LLMs) such as ChatGPT, have gained significant
attention due to their impressive natural language processing capabilities. It
is crucial to prioritize human-centered principles when utilizing these models.
Safeguarding the ethical and moral compliance of LLMs is of utmost importance.
However, individual ethical issues have not been well studied on the latest
LLMs. Therefore, this study aims to address these gaps by introducing a new
benchmark -- TrustGPT. TrustGPT provides a comprehensive evaluation of LLMs in
three crucial areas: toxicity, bias, and value-alignment. Initially, TrustGPT
examines toxicity in language models by employing toxic prompt templates
derived from social norms. It then quantifies the extent of bias in models by
measuring quantifiable toxicity values across different groups. Lastly,
TrustGPT assesses the value of conversation generation models from both active
value-alignment and passive value-alignment tasks. Through the implementation
of TrustGPT, this research aims to enhance our understanding of the performance
of conversation generation models and promote the development of language
models that are more ethical and socially responsible.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 12:53:39 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Huang",
"Yue",
""
],
[
"Zhang",
"Qihui",
""
],
[
"Y",
"Philip S.",
""
],
[
"Sun",
"Lichao",
""
]
] | TITLE: TrustGPT: A Benchmark for Trustworthy and Responsible Large Language
Models
ABSTRACT: Large Language Models (LLMs) such as ChatGPT, have gained significant
attention due to their impressive natural language processing capabilities. It
is crucial to prioritize human-centered principles when utilizing these models.
Safeguarding the ethical and moral compliance of LLMs is of utmost importance.
However, individual ethical issues have not been well studied on the latest
LLMs. Therefore, this study aims to address these gaps by introducing a new
benchmark -- TrustGPT. TrustGPT provides a comprehensive evaluation of LLMs in
three crucial areas: toxicity, bias, and value-alignment. Initially, TrustGPT
examines toxicity in language models by employing toxic prompt templates
derived from social norms. It then quantifies the extent of bias in models by
measuring quantifiable toxicity values across different groups. Lastly,
TrustGPT assesses the value of conversation generation models from both active
value-alignment and passive value-alignment tasks. Through the implementation
of TrustGPT, this research aims to enhance our understanding of the performance
of conversation generation models and promote the development of language
models that are more ethical and socially responsible.
|
2306.11534 | Mat\'u\v{s} Sul\'ir | Mat\'u\v{s} Sul\'ir, Marcel Regeci | Software Engineers' Questions and Answers on Stack Exchange | null | 2022 IEEE 16th International Scientific Conference on Informatics,
IEEE, 2022, pp. 304-310 | 10.1109/Informatics57926.2022.10083403 | null | cs.SE cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There exists a large number of research works analyzing questions and answers
on the popular Stack Overflow website. However, other sub-sites of the Stack
Exchange platform are studied rarely. In this paper, we analyze the questions
and answers on the Software Engineering Stack Exchange site that encompasses a
broader set of areas, such as testing or software processes. Topics and
quantities of the questions, historical trends, and the authors' sentiment were
analyzed using downloaded datasets. We found that the asked questions are most
frequently related to database systems, quality assurance, and agile software
development. The most attractive topics were career and teamwork problems, and
the least attractive ones were network programming and software modeling.
Historically, the topic of domain-driven design recorded the highest rise, and
jobs and career the most significant fall. The number of new questions dropped,
while the portion of unanswered ones increased.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 13:39:49 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Sulír",
"Matúš",
""
],
[
"Regeci",
"Marcel",
""
]
] | TITLE: Software Engineers' Questions and Answers on Stack Exchange
ABSTRACT: There exists a large number of research works analyzing questions and answers
on the popular Stack Overflow website. However, other sub-sites of the Stack
Exchange platform are studied rarely. In this paper, we analyze the questions
and answers on the Software Engineering Stack Exchange site that encompasses a
broader set of areas, such as testing or software processes. Topics and
quantities of the questions, historical trends, and the authors' sentiment were
analyzed using downloaded datasets. We found that the asked questions are most
frequently related to database systems, quality assurance, and agile software
development. The most attractive topics were career and teamwork problems, and
the least attractive ones were network programming and software modeling.
Historically, the topic of domain-driven design recorded the highest rise, and
jobs and career the most significant fall. The number of new questions dropped,
while the portion of unanswered ones increased.
|
2306.11536 | Yu Takagi | Yu Takagi, Shinji Nishimoto | Improving visual image reconstruction from human brain activity using
latent diffusion models via multiple decoded inputs | null | null | null | null | q-bio.NC cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | The integration of deep learning and neuroscience has been advancing rapidly,
which has led to improvements in the analysis of brain activity and the
understanding of deep learning models from a neuroscientific perspective. The
reconstruction of visual experience from human brain activity is an area that
has particularly benefited: the use of deep learning models trained on large
amounts of natural images has greatly improved its quality, and approaches that
combine the diverse information contained in visual experiences have
proliferated rapidly in recent years. In this technical paper, by taking
advantage of the simple and generic framework that we proposed (Takagi and
Nishimoto, CVPR 2023), we examine the extent to which various additional
decoding techniques affect the performance of visual experience reconstruction.
Specifically, we combined our earlier work with the following three techniques:
using decoded text from brain activity, nonlinear optimization for structural
image reconstruction, and using decoded depth information from brain activity.
We confirmed that these techniques contributed to improving accuracy over the
baseline. We also discuss what researchers should consider when performing
visual reconstruction using deep generative models trained on large datasets.
Please check our webpage at
https://sites.google.com/view/stablediffusion-with-brain/. Code is also
available at https://github.com/yu-takagi/StableDiffusionReconstruction.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 13:48:02 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Takagi",
"Yu",
""
],
[
"Nishimoto",
"Shinji",
""
]
] | TITLE: Improving visual image reconstruction from human brain activity using
latent diffusion models via multiple decoded inputs
ABSTRACT: The integration of deep learning and neuroscience has been advancing rapidly,
which has led to improvements in the analysis of brain activity and the
understanding of deep learning models from a neuroscientific perspective. The
reconstruction of visual experience from human brain activity is an area that
has particularly benefited: the use of deep learning models trained on large
amounts of natural images has greatly improved its quality, and approaches that
combine the diverse information contained in visual experiences have
proliferated rapidly in recent years. In this technical paper, by taking
advantage of the simple and generic framework that we proposed (Takagi and
Nishimoto, CVPR 2023), we examine the extent to which various additional
decoding techniques affect the performance of visual experience reconstruction.
Specifically, we combined our earlier work with the following three techniques:
using decoded text from brain activity, nonlinear optimization for structural
image reconstruction, and using decoded depth information from brain activity.
We confirmed that these techniques contributed to improving accuracy over the
baseline. We also discuss what researchers should consider when performing
visual reconstruction using deep generative models trained on large datasets.
Please check our webpage at
https://sites.google.com/view/stablediffusion-with-brain/. Code is also
available at https://github.com/yu-takagi/StableDiffusionReconstruction.
|
2306.11541 | Liying Lu | Liying Lu, Tianke Zhang, Yunfei Liu, Xuangeng Chu, Yu Li | Audio-Driven 3D Facial Animation from In-the-Wild Videos | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given an arbitrary audio clip, audio-driven 3D facial animation aims to
generate lifelike lip motions and facial expressions for a 3D head. Existing
methods typically rely on training their models using limited public 3D
datasets that contain a restricted number of audio-3D scan pairs. Consequently,
their generalization capability remains limited. In this paper, we propose a
novel method that leverages in-the-wild 2D talking-head videos to train our 3D
facial animation model. The abundance of easily accessible 2D talking-head
videos equips our model with a robust generalization capability. By combining
these videos with existing 3D face reconstruction methods, our model excels in
generating consistent and high-fidelity lip synchronization. Additionally, our
model proficiently captures the speaking styles of different individuals,
allowing it to generate 3D talking-heads with distinct personal styles.
Extensive qualitative and quantitative experimental results demonstrate the
superiority of our method.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 13:53:05 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Lu",
"Liying",
""
],
[
"Zhang",
"Tianke",
""
],
[
"Liu",
"Yunfei",
""
],
[
"Chu",
"Xuangeng",
""
],
[
"Li",
"Yu",
""
]
] | TITLE: Audio-Driven 3D Facial Animation from In-the-Wild Videos
ABSTRACT: Given an arbitrary audio clip, audio-driven 3D facial animation aims to
generate lifelike lip motions and facial expressions for a 3D head. Existing
methods typically rely on training their models using limited public 3D
datasets that contain a restricted number of audio-3D scan pairs. Consequently,
their generalization capability remains limited. In this paper, we propose a
novel method that leverages in-the-wild 2D talking-head videos to train our 3D
facial animation model. The abundance of easily accessible 2D talking-head
videos equips our model with a robust generalization capability. By combining
these videos with existing 3D face reconstruction methods, our model excels in
generating consistent and high-fidelity lip synchronization. Additionally, our
model proficiently captures the speaking styles of different individuals,
allowing it to generate 3D talking-heads with distinct personal styles.
Extensive qualitative and quantitative experimental results demonstrate the
superiority of our method.
|
2306.11543 | Iasson Karafyllis | Iasson Karafyllis and Miroslav Krstic | Stabilization and Spill-Free Transfer of Viscous Liquid in a Tank | 40 pages, submitted for possible publication | null | null | null | math.OC cs.SY eess.SY math.AP physics.class-ph physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | Flow control occupies a special place in the fields of partial differential
equations (PDEs) and control theory, where the complex behavior of solutions of
nonlinear dynamics in very high dimension is not just to be understood but also
to be assigned specific desired properties, by feedback control. Among several
benchmark problems in flow control, the liquid-tank problem is particularly
attractive as a research topic. In the liquid-tank problem the objective is to
move a tank filled with liquid, suppress the nonlinear oscillations of the
liquid in the process, bring the tank and liquid to rest, and avoid liquid
spillage in the process. In other words, this is a problem of nonlinear PDE
stabilization subject to state constraints.
This review article focuses only on recent results on liquid-tank
stabilization for viscous liquids. All possible cases are studied: with and
without friction from the tank walls, with and without surface tension.
Moreover, novel results are provided for the linearization of the tank-liquid
system. The linearization of the tank-liquid system gives a high-order PDE
which is a combination of a wave equation with Kelvin-Voigt damping and an
Euler-Bernoulli beam equation.
The feedback design methodology presented in the article is based on Control
Lyapunov Functionals (CLFs), suitably extended from the CLF methodology for
ODEs to the infinite-dimensional case. The CLFs proposed are modifications and
augmentations of the total energy functionals for the tank-liquid system, so
that the dissipative effects of viscosity, friction, and surface tension are
captured and additional dissipation by feedback is made relatively easy.
The article closes with an extensive list of open problems.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 13:54:26 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Karafyllis",
"Iasson",
""
],
[
"Krstic",
"Miroslav",
""
]
] | TITLE: Stabilization and Spill-Free Transfer of Viscous Liquid in a Tank
ABSTRACT: Flow control occupies a special place in the fields of partial differential
equations (PDEs) and control theory, where the complex behavior of solutions of
nonlinear dynamics in very high dimension is not just to be understood but also
to be assigned specific desired properties, by feedback control. Among several
benchmark problems in flow control, the liquid-tank problem is particularly
attractive as a research topic. In the liquid-tank problem the objective is to
move a tank filled with liquid, suppress the nonlinear oscillations of the
liquid in the process, bring the tank and liquid to rest, and avoid liquid
spillage in the process. In other words, this is a problem of nonlinear PDE
stabilization subject to state constraints.
This review article focuses only on recent results on liquid-tank
stabilization for viscous liquids. All possible cases are studied: with and
without friction from the tank walls, with and without surface tension.
Moreover, novel results are provided for the linearization of the tank-liquid
system. The linearization of the tank-liquid system gives a high-order PDE
which is a combination of a wave equation with Kelvin-Voigt damping and an
Euler-Bernoulli beam equation.
The feedback design methodology presented in the article is based on Control
Lyapunov Functionals (CLFs), suitably extended from the CLF methodology for
ODEs to the infinite-dimensional case. The CLFs proposed are modifications and
augmentations of the total energy functionals for the tank-liquid system, so
that the dissipative effects of viscosity, friction, and surface tension are
captured and additional dissipation by feedback is made relatively easy.
The article closes with an extensive list of open problems.
|
2306.11550 | Yuxuan Wang | Yuxuan Wang, Hong Lyu | Query Encoder Distillation via Embedding Alignment is a Strong Baseline
Method to Boost Dense Retriever Online Efficiency | Accepted by the 4th SustaiNLP workshop at ACL 2023 | null | null | null | cs.IR cs.AI | http://creativecommons.org/licenses/by/4.0/ | The information retrieval community has made significant progress in
improving the efficiency of Dual Encoder (DE) dense passage retrieval systems,
making them suitable for latency-sensitive settings. However, many proposed
procedures are often too complex or resource-intensive, which makes it
difficult for practitioners to adopt them or identify sources of empirical
gains. Therefore, in this work, we propose a trivially simple recipe to serve
as a baseline method for boosting the efficiency of DE retrievers leveraging an
asymmetric architecture. Our results demonstrate that even a 2-layer,
BERT-based query encoder can still retain 92.5% of the full DE performance on
the BEIR benchmark via unsupervised distillation and proper student
initialization. We hope that our findings will encourage the community to
re-evaluate the trade-offs between method complexity and performance
improvements.
| [
{
"version": "v1",
"created": "Mon, 5 Jun 2023 06:53:55 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Wang",
"Yuxuan",
""
],
[
"Lyu",
"Hong",
""
]
] | TITLE: Query Encoder Distillation via Embedding Alignment is a Strong Baseline
Method to Boost Dense Retriever Online Efficiency
ABSTRACT: The information retrieval community has made significant progress in
improving the efficiency of Dual Encoder (DE) dense passage retrieval systems,
making them suitable for latency-sensitive settings. However, many proposed
procedures are often too complex or resource-intensive, which makes it
difficult for practitioners to adopt them or identify sources of empirical
gains. Therefore, in this work, we propose a trivially simple recipe to serve
as a baseline method for boosting the efficiency of DE retrievers leveraging an
asymmetric architecture. Our results demonstrate that even a 2-layer,
BERT-based query encoder can still retain 92.5% of the full DE performance on
the BEIR benchmark via unsupervised distillation and proper student
initialization. We hope that our findings will encourage the community to
re-evaluate the trade-offs between method complexity and performance
improvements.
|
2306.11576 | Gorana Goji\'c | Gorana Goji\'c, Veljko Petrovi\'c, Radovan Turovi\'c, Dinu Dragan, Ana
Oros, Du\v{s}an Gaji\'c, Neboj\v{s}a Horvat | Deep Learning Methods for Retinal Blood Vessel Segmentation: Evaluation
on Images with Retinopathy of Prematurity | null | Proceedings of 18th International Symposium on Intelligent Systems
and Informatics (SISY), IEEE, 2020, pp. 131-136 | 10.1109/SISY50555.2020.9217082 | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Automatic blood vessel segmentation from retinal images plays an important
role in the diagnosis of many systemic and eye diseases, including retinopathy
of prematurity. Current state-of-the-art research in blood vessel segmentation
from retinal images is based on convolutional neural networks. The solutions
proposed so far are trained and tested on images from a few available retinal
blood vessel segmentation datasets, which might limit their performance when
given an image with retinopathy of prematurity signs. In this paper, we
evaluate the performance of three high-performing convolutional neural networks
for retinal blood vessel segmentation in the context of blood vessel
segmentation on retinopathy of prematurity retinal images. The main motive
behind the study is to test if existing public datasets suffice to develop a
high-performing predictor that could assist an ophthalmologist in retinopathy
of prematurity diagnosis. To do so, we create a dataset consisting solely of
retinopathy of prematurity images with retinal blood vessel annotations
manually labeled by two observers, where one is the ophthalmologist experienced
in retinopathy of prematurity treatment. Experimental results show that all
three solutions have difficulties in detecting the retinal blood vessels of
infants due to a lower contrast compared to images from public datasets as
demonstrated by a significant drop in classification sensitivity. All three
solutions segment alongside retinal also choroidal blood vessels which are not
used to diagnose retinopathy of prematurity, but instead represent noise and
are confused with retinal blood vessels. By visual and numerical observations,
we observe that existing solutions for retinal blood vessel segmentation need
improvement toward more detailed datasets or deeper models in order to assist
the ophthalmologist in retinopathy of prematurity diagnosis.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 14:46:26 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Gojić",
"Gorana",
""
],
[
"Petrović",
"Veljko",
""
],
[
"Turović",
"Radovan",
""
],
[
"Dragan",
"Dinu",
""
],
[
"Oros",
"Ana",
""
],
[
"Gajić",
"Dušan",
""
],
[
"Horvat",
"Nebojša",
""
]
] | TITLE: Deep Learning Methods for Retinal Blood Vessel Segmentation: Evaluation
on Images with Retinopathy of Prematurity
ABSTRACT: Automatic blood vessel segmentation from retinal images plays an important
role in the diagnosis of many systemic and eye diseases, including retinopathy
of prematurity. Current state-of-the-art research in blood vessel segmentation
from retinal images is based on convolutional neural networks. The solutions
proposed so far are trained and tested on images from a few available retinal
blood vessel segmentation datasets, which might limit their performance when
given an image with retinopathy of prematurity signs. In this paper, we
evaluate the performance of three high-performing convolutional neural networks
for retinal blood vessel segmentation in the context of blood vessel
segmentation on retinopathy of prematurity retinal images. The main motive
behind the study is to test if existing public datasets suffice to develop a
high-performing predictor that could assist an ophthalmologist in retinopathy
of prematurity diagnosis. To do so, we create a dataset consisting solely of
retinopathy of prematurity images with retinal blood vessel annotations
manually labeled by two observers, where one is the ophthalmologist experienced
in retinopathy of prematurity treatment. Experimental results show that all
three solutions have difficulties in detecting the retinal blood vessels of
infants due to a lower contrast compared to images from public datasets as
demonstrated by a significant drop in classification sensitivity. All three
solutions segment alongside retinal also choroidal blood vessels which are not
used to diagnose retinopathy of prematurity, but instead represent noise and
are confused with retinal blood vessels. By visual and numerical observations,
we observe that existing solutions for retinal blood vessel segmentation need
improvement toward more detailed datasets or deeper models in order to assist
the ophthalmologist in retinopathy of prematurity diagnosis.
|
2306.11592 | Zhao Ling | Ling Zhao, Yunpeng Ma, Shanxiong Chen, Jun Zhou | Deep Double Self-Expressive Subspace Clustering | 5 pagees,4 figures,ICASSP2023, revised version | null | 10.1109/ICASSP49357.2023.10095270 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep subspace clustering based on auto-encoder has received wide attention.
However, most subspace clustering based on auto-encoder does not utilize the
structural information in the self-expressive coefficient matrix, which limits
the clustering performance. In this paper, we propose a double self-expressive
subspace clustering algorithm. The key idea of our solution is to view the
self-expressive coefficient as a feature representation of the example to get
another coefficient matrix. Then, we use the two coefficient matrices to
construct the affinity matrix for spectral clustering. We find that it can
reduce the subspace-preserving representation error and improve connectivity.
To further enhance the clustering performance, we proposed a self-supervised
module based on contrastive learning, which can further improve the performance
of the trained network. Experiments on several benchmark datasets demonstrate
that the proposed algorithm can achieve better clustering than state-of-the-art
methods.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:10:35 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Zhao",
"Ling",
""
],
[
"Ma",
"Yunpeng",
""
],
[
"Chen",
"Shanxiong",
""
],
[
"Zhou",
"Jun",
""
]
] | TITLE: Deep Double Self-Expressive Subspace Clustering
ABSTRACT: Deep subspace clustering based on auto-encoder has received wide attention.
However, most subspace clustering based on auto-encoder does not utilize the
structural information in the self-expressive coefficient matrix, which limits
the clustering performance. In this paper, we propose a double self-expressive
subspace clustering algorithm. The key idea of our solution is to view the
self-expressive coefficient as a feature representation of the example to get
another coefficient matrix. Then, we use the two coefficient matrices to
construct the affinity matrix for spectral clustering. We find that it can
reduce the subspace-preserving representation error and improve connectivity.
To further enhance the clustering performance, we proposed a self-supervised
module based on contrastive learning, which can further improve the performance
of the trained network. Experiments on several benchmark datasets demonstrate
that the proposed algorithm can achieve better clustering than state-of-the-art
methods.
|
2306.11593 | Luigi Celona | Simone Bianco and Luigi Celona and Marco Donzella and Paolo Napoletano | Improving Image Captioning Descriptiveness by Ranking and LLM-based
Fusion | null | null | null | null | cs.CV cs.AI cs.CL cs.DB cs.LG | http://creativecommons.org/licenses/by/4.0/ | State-of-The-Art (SoTA) image captioning models often rely on the Microsoft
COCO (MS-COCO) dataset for training. This dataset contains annotations provided
by human annotators, who typically produce captions averaging around ten
tokens. However, this constraint presents a challenge in effectively capturing
complex scenes and conveying detailed information. Furthermore, captioning
models tend to exhibit bias towards the ``average'' caption, which captures
only the more general aspects. What would happen if we were able to
automatically generate longer captions, thereby making them more detailed?
Would these captions, evaluated by humans, be more or less representative of
the image content compared to the original MS-COCO captions? In this paper, we
present a novel approach to address previous challenges by showcasing how
captions generated from different SoTA models can be effectively fused,
resulting in richer captions. Our proposed method leverages existing models
from the literature, eliminating the need for additional training. Instead, it
utilizes an image-text based metric to rank the captions generated by SoTA
models for a given image. Subsequently, the top two captions are fused using a
Large Language Model (LLM). Experimental results demonstrate the effectiveness
of our approach, as the captions generated by our model exhibit higher
consistency with human judgment when evaluated on the MS-COCO test set. By
combining the strengths of various SoTA models, our method enhances the quality
and appeal of image captions, bridging the gap between automated systems and
the rich, informative nature of human-generated descriptions. This advance
opens up new possibilities for generating captions that are more suitable for
the training of both vision-language and captioning models.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:13:02 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Bianco",
"Simone",
""
],
[
"Celona",
"Luigi",
""
],
[
"Donzella",
"Marco",
""
],
[
"Napoletano",
"Paolo",
""
]
] | TITLE: Improving Image Captioning Descriptiveness by Ranking and LLM-based
Fusion
ABSTRACT: State-of-The-Art (SoTA) image captioning models often rely on the Microsoft
COCO (MS-COCO) dataset for training. This dataset contains annotations provided
by human annotators, who typically produce captions averaging around ten
tokens. However, this constraint presents a challenge in effectively capturing
complex scenes and conveying detailed information. Furthermore, captioning
models tend to exhibit bias towards the ``average'' caption, which captures
only the more general aspects. What would happen if we were able to
automatically generate longer captions, thereby making them more detailed?
Would these captions, evaluated by humans, be more or less representative of
the image content compared to the original MS-COCO captions? In this paper, we
present a novel approach to address previous challenges by showcasing how
captions generated from different SoTA models can be effectively fused,
resulting in richer captions. Our proposed method leverages existing models
from the literature, eliminating the need for additional training. Instead, it
utilizes an image-text based metric to rank the captions generated by SoTA
models for a given image. Subsequently, the top two captions are fused using a
Large Language Model (LLM). Experimental results demonstrate the effectiveness
of our approach, as the captions generated by our model exhibit higher
consistency with human judgment when evaluated on the MS-COCO test set. By
combining the strengths of various SoTA models, our method enhances the quality
and appeal of image captions, bridging the gap between automated systems and
the rich, informative nature of human-generated descriptions. This advance
opens up new possibilities for generating captions that are more suitable for
the training of both vision-language and captioning models.
|
2306.11598 | Yucheng Mao | Yucheng Mao, Ruowen Zhao, Tianbao Zhang and Hang Zhao | BEVScope: Enhancing Self-Supervised Depth Estimation Leveraging
Bird's-Eye-View in Dynamic Scenarios | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Depth estimation is a cornerstone of perception in autonomous driving and
robotic systems. The considerable cost and relatively sparse data acquisition
of LiDAR systems have led to the exploration of cost-effective alternatives,
notably, self-supervised depth estimation. Nevertheless, current
self-supervised depth estimation methods grapple with several limitations: (1)
the failure to adequately leverage informative multi-camera views. (2) the
limited capacity to handle dynamic objects effectively. To address these
challenges, we present BEVScope, an innovative approach to self-supervised
depth estimation that harnesses Bird's-Eye-View (BEV) features. Concurrently,
we propose an adaptive loss function, specifically designed to mitigate the
complexities associated with moving objects. Empirical evaluations conducted on
the Nuscenes dataset validate our approach, demonstrating competitive
performance. Code will be released at https://github.com/myc634/BEVScope.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:16:35 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Mao",
"Yucheng",
""
],
[
"Zhao",
"Ruowen",
""
],
[
"Zhang",
"Tianbao",
""
],
[
"Zhao",
"Hang",
""
]
] | TITLE: BEVScope: Enhancing Self-Supervised Depth Estimation Leveraging
Bird's-Eye-View in Dynamic Scenarios
ABSTRACT: Depth estimation is a cornerstone of perception in autonomous driving and
robotic systems. The considerable cost and relatively sparse data acquisition
of LiDAR systems have led to the exploration of cost-effective alternatives,
notably, self-supervised depth estimation. Nevertheless, current
self-supervised depth estimation methods grapple with several limitations: (1)
the failure to adequately leverage informative multi-camera views. (2) the
limited capacity to handle dynamic objects effectively. To address these
challenges, we present BEVScope, an innovative approach to self-supervised
depth estimation that harnesses Bird's-Eye-View (BEV) features. Concurrently,
we propose an adaptive loss function, specifically designed to mitigate the
complexities associated with moving objects. Empirical evaluations conducted on
the Nuscenes dataset validate our approach, demonstrating competitive
performance. Code will be released at https://github.com/myc634/BEVScope.
|
2306.11610 | Kai Ouyang | Kai Ouyang, Xianghong Xu, Miaoxin Chen, Zuotong Xie, Hai-Tao Zheng,
Shuangyong Song, Yu Zhao | Mining Interest Trends and Adaptively Assigning SampleWeight for
Session-based Recommendation | This work has been accepted by SIGIR 2023 | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Session-based Recommendation (SR) aims to predict users' next click based on
their behavior within a short period, which is crucial for online platforms.
However, most existing SR methods somewhat ignore the fact that user preference
is not necessarily strongly related to the order of interactions. Moreover,
they ignore the differences in importance between different samples, which
limits the model-fitting performance. To tackle these issues, we put forward
the method, Mining Interest Trends and Adaptively Assigning Sample Weight,
abbreviated as MTAW. Specifically, we model users' instant interest based on
their present behavior and all their previous behaviors. Meanwhile, we
discriminatively integrate instant interests to capture the changing trend of
user interest to make more personalized recommendations. Furthermore, we devise
a novel loss function that dynamically weights the samples according to their
prediction difficulty in the current epoch. Extensive experimental results on
two benchmark datasets demonstrate the effectiveness and superiority of our
method.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:39:24 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Ouyang",
"Kai",
""
],
[
"Xu",
"Xianghong",
""
],
[
"Chen",
"Miaoxin",
""
],
[
"Xie",
"Zuotong",
""
],
[
"Zheng",
"Hai-Tao",
""
],
[
"Song",
"Shuangyong",
""
],
[
"Zhao",
"Yu",
""
]
] | TITLE: Mining Interest Trends and Adaptively Assigning SampleWeight for
Session-based Recommendation
ABSTRACT: Session-based Recommendation (SR) aims to predict users' next click based on
their behavior within a short period, which is crucial for online platforms.
However, most existing SR methods somewhat ignore the fact that user preference
is not necessarily strongly related to the order of interactions. Moreover,
they ignore the differences in importance between different samples, which
limits the model-fitting performance. To tackle these issues, we put forward
the method, Mining Interest Trends and Adaptively Assigning Sample Weight,
abbreviated as MTAW. Specifically, we model users' instant interest based on
their present behavior and all their previous behaviors. Meanwhile, we
discriminatively integrate instant interests to capture the changing trend of
user interest to make more personalized recommendations. Furthermore, we devise
a novel loss function that dynamically weights the samples according to their
prediction difficulty in the current epoch. Extensive experimental results on
two benchmark datasets demonstrate the effectiveness and superiority of our
method.
|
2306.11615 | Zhao Zhang | Ed Karrels, Lei Huang, Yuhong Kan, Ishank Arora, Yinzhi Wang, Daniel
S. Katz, William D. Gropp, Zhao Zhang | Fine-grained Policy-driven I/O Sharing for Burst Buffers | null | null | null | null | cs.DC | http://creativecommons.org/licenses/by-nc-nd/4.0/ | A burst buffer is a common method to bridge the performance gap between the
I/O needs of modern supercomputing applications and the performance of the
shared file system on large-scale supercomputers. However, existing I/O sharing
methods require resource isolation, offline profiling, or repeated execution
that significantly limit the utilization and applicability of these systems.
Here we present ThemisIO, a policy-driven I/O sharing framework for a
remote-shared burst buffer: a dedicated group of I/O nodes, each with a local
storage device. ThemisIO preserves high utilization by implementing opportunity
fairness so that it can reallocate unused I/O resources to other applications.
ThemisIO accurately and efficiently allocates I/O cycles among applications,
purely based on real-time I/O behavior without requiring user-supplied
information or offline-profiled application characteristics. ThemisIO supports
a variety of fair sharing policies, such as user-fair, size-fair, as well as
composite policies, e.g., group-then-user-fair. All these features are enabled
by its statistical token design. ThemisIO can alter the execution order of
incoming I/O requests based on assigned tokens to precisely balance I/O cycles
between applications via time slicing, thereby enforcing processing isolation.
Experiments using I/O benchmarks show that ThemisIO sustains 13.5-13.7% higher
I/O throughput and 19.5-40.4% lower performance variation than existing
algorithms. For real applications, ThemisIO significantly reduces the slowdown
by 59.1-99.8% caused by I/O interference.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:44:58 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Karrels",
"Ed",
""
],
[
"Huang",
"Lei",
""
],
[
"Kan",
"Yuhong",
""
],
[
"Arora",
"Ishank",
""
],
[
"Wang",
"Yinzhi",
""
],
[
"Katz",
"Daniel S.",
""
],
[
"Gropp",
"William D.",
""
],
[
"Zhang",
"Zhao",
""
]
] | TITLE: Fine-grained Policy-driven I/O Sharing for Burst Buffers
ABSTRACT: A burst buffer is a common method to bridge the performance gap between the
I/O needs of modern supercomputing applications and the performance of the
shared file system on large-scale supercomputers. However, existing I/O sharing
methods require resource isolation, offline profiling, or repeated execution
that significantly limit the utilization and applicability of these systems.
Here we present ThemisIO, a policy-driven I/O sharing framework for a
remote-shared burst buffer: a dedicated group of I/O nodes, each with a local
storage device. ThemisIO preserves high utilization by implementing opportunity
fairness so that it can reallocate unused I/O resources to other applications.
ThemisIO accurately and efficiently allocates I/O cycles among applications,
purely based on real-time I/O behavior without requiring user-supplied
information or offline-profiled application characteristics. ThemisIO supports
a variety of fair sharing policies, such as user-fair, size-fair, as well as
composite policies, e.g., group-then-user-fair. All these features are enabled
by its statistical token design. ThemisIO can alter the execution order of
incoming I/O requests based on assigned tokens to precisely balance I/O cycles
between applications via time slicing, thereby enforcing processing isolation.
Experiments using I/O benchmarks show that ThemisIO sustains 13.5-13.7% higher
I/O throughput and 19.5-40.4% lower performance variation than existing
algorithms. For real applications, ThemisIO significantly reduces the slowdown
by 59.1-99.8% caused by I/O interference.
|
2306.11629 | Jong-Yun Park | Jong-Yun Park, Mitsuaki Tsukamoto, Misato Tanaka, Yukiyasu Kamitani | Sound reconstruction from human brain activity via a generative model
with brain-like auditory features | null | null | null | null | cs.SD cs.HC eess.AS | http://creativecommons.org/licenses/by/4.0/ | The successful reconstruction of perceptual experiences from human brain
activity has provided insights into the neural representations of sensory
experiences. However, reconstructing arbitrary sounds has been avoided due to
the complexity of temporal sequences in sounds and the limited resolution of
neuroimaging modalities. To overcome these challenges, leveraging the
hierarchical nature of brain auditory processing could provide a path toward
reconstructing arbitrary sounds. Previous studies have indicated a hierarchical
homology between the human auditory system and deep neural network (DNN)
models. Furthermore, advancements in audio-generative models enable to
transform compressed representations back into high-resolution sounds. In this
study, we introduce a novel sound reconstruction method that combines brain
decoding of auditory features with an audio-generative model. Using fMRI
responses to natural sounds, we found that the hierarchical sound features of a
DNN model could be better decoded than spectrotemporal features. We then
reconstructed the sound using an audio transformer that disentangled compressed
temporal information in the decoded DNN features. Our method shows
unconstrained sounds reconstruction capturing sound perceptual contents and
quality and generalizability by reconstructing sound categories not included in
the training dataset. Reconstructions from different auditory regions remain
similar to actual sounds, highlighting the distributed nature of auditory
representations. To see whether the reconstructions mirrored actual subjective
perceptual experiences, we performed an experiment involving selective auditory
attention to one of overlapping sounds. The results tended to resemble the
attended sound than the unattended. These findings demonstrate that our
proposed model provides a means to externalize experienced auditory contents
from human brain activity.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 15:59:09 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Park",
"Jong-Yun",
""
],
[
"Tsukamoto",
"Mitsuaki",
""
],
[
"Tanaka",
"Misato",
""
],
[
"Kamitani",
"Yukiyasu",
""
]
] | TITLE: Sound reconstruction from human brain activity via a generative model
with brain-like auditory features
ABSTRACT: The successful reconstruction of perceptual experiences from human brain
activity has provided insights into the neural representations of sensory
experiences. However, reconstructing arbitrary sounds has been avoided due to
the complexity of temporal sequences in sounds and the limited resolution of
neuroimaging modalities. To overcome these challenges, leveraging the
hierarchical nature of brain auditory processing could provide a path toward
reconstructing arbitrary sounds. Previous studies have indicated a hierarchical
homology between the human auditory system and deep neural network (DNN)
models. Furthermore, advancements in audio-generative models enable to
transform compressed representations back into high-resolution sounds. In this
study, we introduce a novel sound reconstruction method that combines brain
decoding of auditory features with an audio-generative model. Using fMRI
responses to natural sounds, we found that the hierarchical sound features of a
DNN model could be better decoded than spectrotemporal features. We then
reconstructed the sound using an audio transformer that disentangled compressed
temporal information in the decoded DNN features. Our method shows
unconstrained sounds reconstruction capturing sound perceptual contents and
quality and generalizability by reconstructing sound categories not included in
the training dataset. Reconstructions from different auditory regions remain
similar to actual sounds, highlighting the distributed nature of auditory
representations. To see whether the reconstructions mirrored actual subjective
perceptual experiences, we performed an experiment involving selective auditory
attention to one of overlapping sounds. The results tended to resemble the
attended sound than the unattended. These findings demonstrate that our
proposed model provides a means to externalize experienced auditory contents
from human brain activity.
|
2306.11636 | Katarzyna Wo\'znica | Katarzyna Wo\'znica, Piotr Wilczy\'nski, Przemys{\l}aw Biecek | SeFNet: Bridging Tabular Datasets with Semantic Feature Nets | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Machine learning applications cover a wide range of predictive tasks in which
tabular datasets play a significant role. However, although they often address
similar problems, tabular datasets are typically treated as standalone tasks.
The possibilities of using previously solved problems are limited due to the
lack of structured contextual information about their features and the lack of
understanding of the relations between them. To overcome this limitation, we
propose a new approach called Semantic Feature Net (SeFNet), capturing the
semantic meaning of the analyzed tabular features. By leveraging existing
ontologies and domain knowledge, SeFNet opens up new opportunities for sharing
insights between diverse predictive tasks. One such opportunity is the Dataset
Ontology-based Semantic Similarity (DOSS) measure, which quantifies the
similarity between datasets using relations across their features. In this
paper, we present an example of SeFNet prepared for a collection of predictive
tasks in healthcare, with the features' relations derived from the SNOMED-CT
ontology. The proposed SeFNet framework and the accompanying DOSS measure
address the issue of limited contextual information in tabular datasets. By
incorporating domain knowledge and establishing semantic relations between
features, we enhance the potential for meta-learning and enable valuable
insights to be shared across different predictive tasks.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 16:02:56 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Woźnica",
"Katarzyna",
""
],
[
"Wilczyński",
"Piotr",
""
],
[
"Biecek",
"Przemysław",
""
]
] | TITLE: SeFNet: Bridging Tabular Datasets with Semantic Feature Nets
ABSTRACT: Machine learning applications cover a wide range of predictive tasks in which
tabular datasets play a significant role. However, although they often address
similar problems, tabular datasets are typically treated as standalone tasks.
The possibilities of using previously solved problems are limited due to the
lack of structured contextual information about their features and the lack of
understanding of the relations between them. To overcome this limitation, we
propose a new approach called Semantic Feature Net (SeFNet), capturing the
semantic meaning of the analyzed tabular features. By leveraging existing
ontologies and domain knowledge, SeFNet opens up new opportunities for sharing
insights between diverse predictive tasks. One such opportunity is the Dataset
Ontology-based Semantic Similarity (DOSS) measure, which quantifies the
similarity between datasets using relations across their features. In this
paper, we present an example of SeFNet prepared for a collection of predictive
tasks in healthcare, with the features' relations derived from the SNOMED-CT
ontology. The proposed SeFNet framework and the accompanying DOSS measure
address the issue of limited contextual information in tabular datasets. By
incorporating domain knowledge and establishing semantic relations between
features, we enhance the potential for meta-learning and enable valuable
insights to be shared across different predictive tasks.
|
2306.11640 | Shubham Patil | Shubham Patil, Anand Sharma, Gaurav R, Abhishek Kadam, Ajay Kumar
Singh, Sandip Lashkare, Nihar Ranjan Mohapatra, and Udayan Ganguly | Process Voltage Temperature Variability Estimation of Tunneling Current
for Band-to-Band-Tunneling based Neuron | null | null | null | null | physics.app-ph cs.ET | http://creativecommons.org/publicdomain/zero/1.0/ | Compact and energy-efficient Synapse and Neurons are essential to realize the
full potential of neuromorphic computing. In addition, a low variability is
indeed needed for neurons in Deep neural networks for higher accuracy. Further,
process (P), voltage (V), and temperature (T) variation (PVT) are essential
considerations for low-power circuits as performance impact and compensation
complexities are added costs. Recently, band-to-band tunneling (BTBT) neuron
has been demonstrated to operate successfully in a network to enable a Liquid
State Machine. A comparison of the PVT with competing modes of operation (e.g.,
BTBT vs. sub-threshold and above threshold) of the same transistor is a
critical factor in assessing performance. In this work, we demonstrate the PVT
variation impact in the BTBT regime and benchmark the operation against the
subthreshold slope (SS) and ON-regime (ION) of partially depleted-Silicon on
Insulator MOSFET. It is shown that the On-state regime offers the lowest
variability but dissipates higher power. Hence, not usable for low-power
sources. Among the BTBT and SS regimes, which can enable the low-power neuron,
the BTBT regime has shown ~3x variability reduction ({\sigma}_I_D/{\mu}_I_D)
than the SS regime, considering the cumulative PVT variability. The improvement
is due to the well-known weaker P, V, and T dependence of BTBT vs. SS. We show
that the BTBT variation is uncorrelated with mutually correlated SS & ION
operation - indicating its different origin from the mechanism and location
perspectives. Hence, the BTBT regime is promising for low-current, low-power,
and low device-to-device variability neuron operation.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 16:11:17 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Patil",
"Shubham",
""
],
[
"Sharma",
"Anand",
""
],
[
"R",
"Gaurav",
""
],
[
"Kadam",
"Abhishek",
""
],
[
"Singh",
"Ajay Kumar",
""
],
[
"Lashkare",
"Sandip",
""
],
[
"Mohapatra",
"Nihar Ranjan",
""
],
[
"Ganguly",
"Udayan",
""
]
] | TITLE: Process Voltage Temperature Variability Estimation of Tunneling Current
for Band-to-Band-Tunneling based Neuron
ABSTRACT: Compact and energy-efficient Synapse and Neurons are essential to realize the
full potential of neuromorphic computing. In addition, a low variability is
indeed needed for neurons in Deep neural networks for higher accuracy. Further,
process (P), voltage (V), and temperature (T) variation (PVT) are essential
considerations for low-power circuits as performance impact and compensation
complexities are added costs. Recently, band-to-band tunneling (BTBT) neuron
has been demonstrated to operate successfully in a network to enable a Liquid
State Machine. A comparison of the PVT with competing modes of operation (e.g.,
BTBT vs. sub-threshold and above threshold) of the same transistor is a
critical factor in assessing performance. In this work, we demonstrate the PVT
variation impact in the BTBT regime and benchmark the operation against the
subthreshold slope (SS) and ON-regime (ION) of partially depleted-Silicon on
Insulator MOSFET. It is shown that the On-state regime offers the lowest
variability but dissipates higher power. Hence, not usable for low-power
sources. Among the BTBT and SS regimes, which can enable the low-power neuron,
the BTBT regime has shown ~3x variability reduction ({\sigma}_I_D/{\mu}_I_D)
than the SS regime, considering the cumulative PVT variability. The improvement
is due to the well-known weaker P, V, and T dependence of BTBT vs. SS. We show
that the BTBT variation is uncorrelated with mutually correlated SS & ION
operation - indicating its different origin from the mechanism and location
perspectives. Hence, the BTBT regime is promising for low-current, low-power,
and low device-to-device variability neuron operation.
|
2306.11651 | Walter Boscheri Prof. | Walter Boscheri and Michael Dumbser and Pierre-Henri Maire | A new thermodynamically compatible finite volume scheme for Lagrangian
gas dynamics | null | null | null | null | math.NA cs.NA | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The equations of Lagrangian gas dynamics fall into the larger class of
overdetermined hyperbolic and thermodynamically compatible (HTC) systems of
partial differential equations. They satisfy an entropy inequality (second
principle of thermodynamics) and conserve total energy (first principle of
thermodynamics). The aim of this work is to construct a novel thermodynamically
compatible cell-centered Lagrangian finite volume scheme on unstructured
meshes. Unlike in existing schemes, we choose to directly discretize the
entropy inequality, hence obtaining total energy conservation as a consequence
of the new thermodynamically compatible discretization of the other equations.
First, the governing equations are written in fluctuation form. Next, the
non-compatible centered numerical fluxes are corrected according to the
approach recently introduced by Abgrall et al., using a scalar correction
factor that is defined at the nodes of the grid. This perfectly fits into the
formalism of nodal solvers which is typically adopted in cell-centered
Lagrangian finite volume methods. Semi-discrete entropy conservative and
entropy stable Lagrangian schemes are devised, and they are adequately blended
together via a convex combination based on either a priori or a posteriori
detectors of discontinuous solutions. The nonlinear stability in the energy
norm is rigorously demonstrated and the new schemes are provably positivity
preserving for density and pressure. Furthermore, they exhibit zero numerical
diffusion for isentropic flows while being still nonlinearly stable. The new
schemes are tested against classical benchmarks for Lagrangian hydrodynamics,
assessing their convergence and robustness and comparing their numerical
dissipation with classical Lagrangian finite volume methods.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 16:19:19 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Boscheri",
"Walter",
""
],
[
"Dumbser",
"Michael",
""
],
[
"Maire",
"Pierre-Henri",
""
]
] | TITLE: A new thermodynamically compatible finite volume scheme for Lagrangian
gas dynamics
ABSTRACT: The equations of Lagrangian gas dynamics fall into the larger class of
overdetermined hyperbolic and thermodynamically compatible (HTC) systems of
partial differential equations. They satisfy an entropy inequality (second
principle of thermodynamics) and conserve total energy (first principle of
thermodynamics). The aim of this work is to construct a novel thermodynamically
compatible cell-centered Lagrangian finite volume scheme on unstructured
meshes. Unlike in existing schemes, we choose to directly discretize the
entropy inequality, hence obtaining total energy conservation as a consequence
of the new thermodynamically compatible discretization of the other equations.
First, the governing equations are written in fluctuation form. Next, the
non-compatible centered numerical fluxes are corrected according to the
approach recently introduced by Abgrall et al., using a scalar correction
factor that is defined at the nodes of the grid. This perfectly fits into the
formalism of nodal solvers which is typically adopted in cell-centered
Lagrangian finite volume methods. Semi-discrete entropy conservative and
entropy stable Lagrangian schemes are devised, and they are adequately blended
together via a convex combination based on either a priori or a posteriori
detectors of discontinuous solutions. The nonlinear stability in the energy
norm is rigorously demonstrated and the new schemes are provably positivity
preserving for density and pressure. Furthermore, they exhibit zero numerical
diffusion for isentropic flows while being still nonlinearly stable. The new
schemes are tested against classical benchmarks for Lagrangian hydrodynamics,
assessing their convergence and robustness and comparing their numerical
dissipation with classical Lagrangian finite volume methods.
|
2306.11673 | Nima Shiri Harzevili | Nima Shiri Harzevili, Alvine Boaye Belle, Junjie Wang, Song Wang, Zhen
Ming (Jack) Jiang, Nachiappan Nagappan | A Survey on Automated Software Vulnerability Detection Using Machine
Learning and Deep Learning | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Software vulnerability detection is critical in software security because it
identifies potential bugs in software systems, enabling immediate remediation
and mitigation measures to be implemented before they may be exploited.
Automatic vulnerability identification is important because it can evaluate
large codebases more efficiently than manual code auditing. Many Machine
Learning (ML) and Deep Learning (DL) based models for detecting vulnerabilities
in source code have been presented in recent years. However, a survey that
summarises, classifies, and analyses the application of ML/DL models for
vulnerability detection is missing. It may be difficult to discover gaps in
existing research and potential for future improvement without a comprehensive
survey. This could result in essential areas of research being overlooked or
under-represented, leading to a skewed understanding of the state of the art in
vulnerability detection. This work address that gap by presenting a systematic
survey to characterize various features of ML/DL-based source code level
software vulnerability detection approaches via five primary research questions
(RQs). Specifically, our RQ1 examines the trend of publications that leverage
ML/DL for vulnerability detection, including the evolution of research and the
distribution of publication venues. RQ2 describes vulnerability datasets used
by existing ML/DL-based models, including their sources, types, and
representations, as well as analyses of the embedding techniques used by these
approaches. RQ3 explores the model architectures and design assumptions of
ML/DL-based vulnerability detection approaches. RQ4 summarises the type and
frequency of vulnerabilities that are covered by existing studies. Lastly, RQ5
presents a list of current challenges to be researched and an outline of a
potential research roadmap that highlights crucial opportunities for future
work.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 16:51:59 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Harzevili",
"Nima Shiri",
"",
"Jack"
],
[
"Belle",
"Alvine Boaye",
"",
"Jack"
],
[
"Wang",
"Junjie",
"",
"Jack"
],
[
"Wang",
"Song",
"",
"Jack"
],
[
"Ming",
"Zhen",
"",
"Jack"
],
[
"Jiang",
"",
""
],
[
"Nagappan",
"Nachiappan",
""
]
] | TITLE: A Survey on Automated Software Vulnerability Detection Using Machine
Learning and Deep Learning
ABSTRACT: Software vulnerability detection is critical in software security because it
identifies potential bugs in software systems, enabling immediate remediation
and mitigation measures to be implemented before they may be exploited.
Automatic vulnerability identification is important because it can evaluate
large codebases more efficiently than manual code auditing. Many Machine
Learning (ML) and Deep Learning (DL) based models for detecting vulnerabilities
in source code have been presented in recent years. However, a survey that
summarises, classifies, and analyses the application of ML/DL models for
vulnerability detection is missing. It may be difficult to discover gaps in
existing research and potential for future improvement without a comprehensive
survey. This could result in essential areas of research being overlooked or
under-represented, leading to a skewed understanding of the state of the art in
vulnerability detection. This work address that gap by presenting a systematic
survey to characterize various features of ML/DL-based source code level
software vulnerability detection approaches via five primary research questions
(RQs). Specifically, our RQ1 examines the trend of publications that leverage
ML/DL for vulnerability detection, including the evolution of research and the
distribution of publication venues. RQ2 describes vulnerability datasets used
by existing ML/DL-based models, including their sources, types, and
representations, as well as analyses of the embedding techniques used by these
approaches. RQ3 explores the model architectures and design assumptions of
ML/DL-based vulnerability detection approaches. RQ4 summarises the type and
frequency of vulnerabilities that are covered by existing studies. Lastly, RQ5
presents a list of current challenges to be researched and an outline of a
potential research roadmap that highlights crucial opportunities for future
work.
|
2306.11682 | Yuhao Nie | Yuhao Nie, Eric Zelikman, Andea Scott, Quentin Paletta, Adam Brandt | SkyGPT: Probabilistic Short-term Solar Forecasting Using Synthetic Sky
Videos from Physics-constrained VideoGPT | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, deep learning-based solar forecasting using all-sky images
has emerged as a promising approach for alleviating uncertainty in PV power
generation. However, the stochastic nature of cloud movement remains a major
challenge for accurate and reliable solar forecasting. With the recent advances
in generative artificial intelligence, the synthesis of visually plausible yet
diversified sky videos has potential for aiding in forecasts. In this study, we
introduce \emph{SkyGPT}, a physics-informed stochastic video prediction model
that is able to generate multiple possible future images of the sky with
diverse cloud motion patterns, by using past sky image sequences as input.
Extensive experiments and comparison with benchmark video prediction models
demonstrate the effectiveness of the proposed model in capturing cloud dynamics
and generating future sky images with high realism and diversity. Furthermore,
we feed the generated future sky images from the video prediction models for
15-minute-ahead probabilistic solar forecasting for a 30-kW roof-top PV system,
and compare it with an end-to-end deep learning baseline model SUNSET and a
smart persistence model. Better PV output prediction reliability and sharpness
is observed by using the predicted sky images generated with SkyGPT compared
with other benchmark models, achieving a continuous ranked probability score
(CRPS) of 2.81 (13\% better than SUNSET and 23\% better than smart persistence)
and a Winkler score of 26.70 for the test set. Although an arbitrary number of
futures can be generated from a historical sky image sequence, the results
suggest that 10 future scenarios is a good choice that balances probabilistic
solar forecasting performance and computational cost.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 16:58:51 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Nie",
"Yuhao",
""
],
[
"Zelikman",
"Eric",
""
],
[
"Scott",
"Andea",
""
],
[
"Paletta",
"Quentin",
""
],
[
"Brandt",
"Adam",
""
]
] | TITLE: SkyGPT: Probabilistic Short-term Solar Forecasting Using Synthetic Sky
Videos from Physics-constrained VideoGPT
ABSTRACT: In recent years, deep learning-based solar forecasting using all-sky images
has emerged as a promising approach for alleviating uncertainty in PV power
generation. However, the stochastic nature of cloud movement remains a major
challenge for accurate and reliable solar forecasting. With the recent advances
in generative artificial intelligence, the synthesis of visually plausible yet
diversified sky videos has potential for aiding in forecasts. In this study, we
introduce \emph{SkyGPT}, a physics-informed stochastic video prediction model
that is able to generate multiple possible future images of the sky with
diverse cloud motion patterns, by using past sky image sequences as input.
Extensive experiments and comparison with benchmark video prediction models
demonstrate the effectiveness of the proposed model in capturing cloud dynamics
and generating future sky images with high realism and diversity. Furthermore,
we feed the generated future sky images from the video prediction models for
15-minute-ahead probabilistic solar forecasting for a 30-kW roof-top PV system,
and compare it with an end-to-end deep learning baseline model SUNSET and a
smart persistence model. Better PV output prediction reliability and sharpness
is observed by using the predicted sky images generated with SkyGPT compared
with other benchmark models, achieving a continuous ranked probability score
(CRPS) of 2.81 (13\% better than SUNSET and 23\% better than smart persistence)
and a Winkler score of 26.70 for the test set. Although an arbitrary number of
futures can be generated from a historical sky image sequence, the results
suggest that 10 future scenarios is a good choice that balances probabilistic
solar forecasting performance and computational cost.
|
2306.11699 | Brendan Artley Mr | Brendan Artley | GenPlot: Increasing the Scale and Diversity of Chart Derendering Data | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Vertical bars, horizontal bars, dot, scatter, and line plots provide a
diverse set of visualizations to represent data. To understand these plots, one
must be able to recognize textual components, locate data points in a plot, and
process diverse visual contexts to extract information. In recent works such as
Pix2Struct, Matcha, and Deplot, OCR-free chart-to-text translation has achieved
state-of-the-art results on visual language tasks. These results outline the
importance of chart-derendering as a pre-training objective, yet existing
datasets provide a fixed set of training examples. In this paper, we propose
GenPlot; a plot generator that can generate billions of additional plots for
chart-derendering using synthetic data.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 17:25:53 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Artley",
"Brendan",
""
]
] | TITLE: GenPlot: Increasing the Scale and Diversity of Chart Derendering Data
ABSTRACT: Vertical bars, horizontal bars, dot, scatter, and line plots provide a
diverse set of visualizations to represent data. To understand these plots, one
must be able to recognize textual components, locate data points in a plot, and
process diverse visual contexts to extract information. In recent works such as
Pix2Struct, Matcha, and Deplot, OCR-free chart-to-text translation has achieved
state-of-the-art results on visual language tasks. These results outline the
importance of chart-derendering as a pre-training objective, yet existing
datasets provide a fixed set of training examples. In this paper, we propose
GenPlot; a plot generator that can generate billions of additional plots for
chart-derendering using synthetic data.
|
2306.11714 | Sovesh Mohapatra | Sovesh Mohapatra, Advait Gosai, Anant Shinde, Aleksei Rutkovskii,
Sirisha Nouduri, Gottfried Schlaug | Meta-Analysis of Transfer Learning for Segmentation of Brain Lesions | 13 Pages, 5 figures, 3 tables | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | A major challenge in stroke research and stroke recovery predictions is the
determination of a stroke lesion's extent and its impact on relevant brain
systems. Manual segmentation of stroke lesions from 3D magnetic resonance (MR)
imaging volumes, the current gold standard, is not only very time-consuming,
but its accuracy highly depends on the operator's experience. As a result,
there is a need for a fully automated segmentation method that can efficiently
and objectively measure lesion extent and the impact of each lesion to predict
impairment and recovery potential which might be beneficial for clinical,
translational, and research settings. We have implemented and tested a fully
automatic method for stroke lesion segmentation which was developed using eight
different 2D-model architectures trained via transfer learning (TL) and mixed
data approaches. Additionally, the final prediction was made using a novel
ensemble method involving stacking and agreement window. Our novel method was
evaluated in a novel in-house dataset containing 22 T1w brain MR images, which
were challenging in various perspectives, but mostly because they included T1w
MR images from the subacute (which typically less well defined T1 lesions) and
chronic stroke phase (which typically means well defined T1-lesions).
Cross-validation results indicate that our new method can efficiently and
automatically segment lesions fast and with high accuracy compared to ground
truth. In addition to segmentation, we provide lesion volume and weighted
lesion load of relevant brain systems based on the lesions' overlap with a
canonical structural motor system that stretches from the cortical motor region
to the lowest end of the brain stem.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 17:42:30 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Mohapatra",
"Sovesh",
""
],
[
"Gosai",
"Advait",
""
],
[
"Shinde",
"Anant",
""
],
[
"Rutkovskii",
"Aleksei",
""
],
[
"Nouduri",
"Sirisha",
""
],
[
"Schlaug",
"Gottfried",
""
]
] | TITLE: Meta-Analysis of Transfer Learning for Segmentation of Brain Lesions
ABSTRACT: A major challenge in stroke research and stroke recovery predictions is the
determination of a stroke lesion's extent and its impact on relevant brain
systems. Manual segmentation of stroke lesions from 3D magnetic resonance (MR)
imaging volumes, the current gold standard, is not only very time-consuming,
but its accuracy highly depends on the operator's experience. As a result,
there is a need for a fully automated segmentation method that can efficiently
and objectively measure lesion extent and the impact of each lesion to predict
impairment and recovery potential which might be beneficial for clinical,
translational, and research settings. We have implemented and tested a fully
automatic method for stroke lesion segmentation which was developed using eight
different 2D-model architectures trained via transfer learning (TL) and mixed
data approaches. Additionally, the final prediction was made using a novel
ensemble method involving stacking and agreement window. Our novel method was
evaluated in a novel in-house dataset containing 22 T1w brain MR images, which
were challenging in various perspectives, but mostly because they included T1w
MR images from the subacute (which typically less well defined T1 lesions) and
chronic stroke phase (which typically means well defined T1-lesions).
Cross-validation results indicate that our new method can efficiently and
automatically segment lesions fast and with high accuracy compared to ground
truth. In addition to segmentation, we provide lesion volume and weighted
lesion load of relevant brain systems based on the lesions' overlap with a
canonical structural motor system that stretches from the cortical motor region
to the lowest end of the brain stem.
|
2306.11726 | Xingyi Zhou | Xingyi Zhou, Anurag Arnab, Chen Sun, Cordelia Schmid | How can objects help action recognition? | CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current state-of-the-art video models process a video clip as a long sequence
of spatio-temporal tokens. However, they do not explicitly model objects, their
interactions across the video, and instead process all the tokens in the video.
In this paper, we investigate how we can use knowledge of objects to design
better video models, namely to process fewer tokens and to improve recognition
accuracy. This is in contrast to prior works which either drop tokens at the
cost of accuracy, or increase accuracy whilst also increasing the computation
required. First, we propose an object-guided token sampling strategy that
enables us to retain a small fraction of the input tokens with minimal impact
on accuracy. And second, we propose an object-aware attention module that
enriches our feature representation with object information and improves
overall accuracy. Our resulting framework achieves better performance when
using fewer tokens than strong baselines. In particular, we match our baseline
with 30%, 40%, and 60% of the input tokens on SomethingElse,
Something-something v2, and Epic-Kitchens, respectively. When we use our model
to process the same number of tokens as our baseline, we improve by 0.6 to 4.2
points on these datasets.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 17:56:16 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Zhou",
"Xingyi",
""
],
[
"Arnab",
"Anurag",
""
],
[
"Sun",
"Chen",
""
],
[
"Schmid",
"Cordelia",
""
]
] | TITLE: How can objects help action recognition?
ABSTRACT: Current state-of-the-art video models process a video clip as a long sequence
of spatio-temporal tokens. However, they do not explicitly model objects, their
interactions across the video, and instead process all the tokens in the video.
In this paper, we investigate how we can use knowledge of objects to design
better video models, namely to process fewer tokens and to improve recognition
accuracy. This is in contrast to prior works which either drop tokens at the
cost of accuracy, or increase accuracy whilst also increasing the computation
required. First, we propose an object-guided token sampling strategy that
enables us to retain a small fraction of the input tokens with minimal impact
on accuracy. And second, we propose an object-aware attention module that
enriches our feature representation with object information and improves
overall accuracy. Our resulting framework achieves better performance when
using fewer tokens than strong baselines. In particular, we match our baseline
with 30%, 40%, and 60% of the input tokens on SomethingElse,
Something-something v2, and Epic-Kitchens, respectively. When we use our model
to process the same number of tokens as our baseline, we improve by 0.6 to 4.2
points on these datasets.
|
2306.11732 | Junting Pan | Junting Pan, Ziyi Lin, Yuying Ge, Xiatian Zhu, Renrui Zhang, Yi Wang,
Yu Qiao, Hongsheng Li | Retrieving-to-Answer: Zero-Shot Video Question Answering with Frozen
Large Language Models | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Video Question Answering (VideoQA) has been significantly advanced from the
scaling of recent Large Language Models (LLMs). The key idea is to convert the
visual information into the language feature space so that the capacity of LLMs
can be fully exploited. Existing VideoQA methods typically take two paradigms:
(1) learning cross-modal alignment, and (2) using an off-the-shelf captioning
model to describe the visual data. However, the first design needs costly
training on many extra multi-modal data, whilst the second is further limited
by limited domain generalization. To address these limitations, a simple yet
effective Retrieving-to-Answer (R2A) framework is proposed.Given an input
video, R2A first retrieves a set of semantically similar texts from a generic
text corpus using a pre-trained multi-modal model (e.g., CLIP). With both the
question and the retrieved texts, a LLM (e.g., DeBERTa) can be directly used to
yield a desired answer. Without the need for cross-modal fine-tuning, R2A
allows for all the key components (e.g., LLM, retrieval model, and text corpus)
to plug-and-play. Extensive experiments on several VideoQA benchmarks show that
despite with 1.3B parameters and no fine-tuning, our R2A can outperform the 61
times larger Flamingo-80B model even additionally trained on nearly 2.1B
multi-modal data.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 20:56:20 GMT"
}
] | 2023-06-21T00:00:00 | [
[
"Pan",
"Junting",
""
],
[
"Lin",
"Ziyi",
""
],
[
"Ge",
"Yuying",
""
],
[
"Zhu",
"Xiatian",
""
],
[
"Zhang",
"Renrui",
""
],
[
"Wang",
"Yi",
""
],
[
"Qiao",
"Yu",
""
],
[
"Li",
"Hongsheng",
""
]
] | TITLE: Retrieving-to-Answer: Zero-Shot Video Question Answering with Frozen
Large Language Models
ABSTRACT: Video Question Answering (VideoQA) has been significantly advanced from the
scaling of recent Large Language Models (LLMs). The key idea is to convert the
visual information into the language feature space so that the capacity of LLMs
can be fully exploited. Existing VideoQA methods typically take two paradigms:
(1) learning cross-modal alignment, and (2) using an off-the-shelf captioning
model to describe the visual data. However, the first design needs costly
training on many extra multi-modal data, whilst the second is further limited
by limited domain generalization. To address these limitations, a simple yet
effective Retrieving-to-Answer (R2A) framework is proposed.Given an input
video, R2A first retrieves a set of semantically similar texts from a generic
text corpus using a pre-trained multi-modal model (e.g., CLIP). With both the
question and the retrieved texts, a LLM (e.g., DeBERTa) can be directly used to
yield a desired answer. Without the need for cross-modal fine-tuning, R2A
allows for all the key components (e.g., LLM, retrieval model, and text corpus)
to plug-and-play. Extensive experiments on several VideoQA benchmarks show that
despite with 1.3B parameters and no fine-tuning, our R2A can outperform the 61
times larger Flamingo-80B model even additionally trained on nearly 2.1B
multi-modal data.
|
2306.09313 | Rohit Paturi | Rohit Paturi, Sundararajan Srinivasan, Xiang Li | Lexical Speaker Error Correction: Leveraging Language Models for Speaker
Diarization Error Correction | Accepted at INTERSPEECH 2023 | null | null | null | eess.AS cs.AI cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Speaker diarization (SD) is typically used with an automatic speech
recognition (ASR) system to ascribe speaker labels to recognized words. The
conventional approach reconciles outputs from independently optimized ASR and
SD systems, where the SD system typically uses only acoustic information to
identify the speakers in the audio stream. This approach can lead to speaker
errors especially around speaker turns and regions of speaker overlap. In this
paper, we propose a novel second-pass speaker error correction system using
lexical information, leveraging the power of modern language models (LMs). Our
experiments across multiple telephony datasets show that our approach is both
effective and robust. Training and tuning only on the Fisher dataset, this
error correction approach leads to relative word-level diarization error rate
(WDER) reductions of 15-30% on three telephony datasets: RT03-CTS, Callhome
American English and held-out portions of Fisher.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 17:47:41 GMT"
}
] | 2023-06-20T00:00:00 | [
[
"Paturi",
"Rohit",
""
],
[
"Srinivasan",
"Sundararajan",
""
],
[
"Li",
"Xiang",
""
]
] | TITLE: Lexical Speaker Error Correction: Leveraging Language Models for Speaker
Diarization Error Correction
ABSTRACT: Speaker diarization (SD) is typically used with an automatic speech
recognition (ASR) system to ascribe speaker labels to recognized words. The
conventional approach reconciles outputs from independently optimized ASR and
SD systems, where the SD system typically uses only acoustic information to
identify the speakers in the audio stream. This approach can lead to speaker
errors especially around speaker turns and regions of speaker overlap. In this
paper, we propose a novel second-pass speaker error correction system using
lexical information, leveraging the power of modern language models (LMs). Our
experiments across multiple telephony datasets show that our approach is both
effective and robust. Training and tuning only on the Fisher dataset, this
error correction approach leads to relative word-level diarization error rate
(WDER) reductions of 15-30% on three telephony datasets: RT03-CTS, Callhome
American English and held-out portions of Fisher.
|
2306.09628 | Arkaitz Bidaurrazaga Barrueta | Arkaitz Bidaurrazaga, Aritz P\'erez, Roberto Santana | Structural Restricted Boltzmann Machine for image denoising and
classification | null | null | null | null | cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Restricted Boltzmann Machines are generative models that consist of a layer
of hidden variables connected to another layer of visible units, and they are
used to model the distribution over visible variables. In order to gain a
higher representability power, many hidden units are commonly used, which, in
combination with a large number of visible units, leads to a high number of
trainable parameters. In this work we introduce the Structural Restricted
Boltzmann Machine model, which taking advantage of the structure of the data in
hand, constrains connections of hidden units to subsets of visible units in
order to reduce significantly the number of trainable parameters, without
compromising performance. As a possible area of application, we focus on image
modelling. Based on the nature of the images, the structure of the connections
is given in terms of spatial neighbourhoods over the pixels of the image that
constitute the visible variables of the model. We conduct extensive experiments
on various image domains. Image denoising is evaluated with corrupted images
from the MNIST dataset. The generative power of our models is compared to
vanilla RBMs, as well as their classification performance, which is assessed
with five different image domains. Results show that our proposed model has a
faster and more stable training, while also obtaining better results compared
to an RBM with no constrained connections between its visible and hidden units.
| [
{
"version": "v1",
"created": "Fri, 16 Jun 2023 05:18:26 GMT"
}
] | 2023-06-20T00:00:00 | [
[
"Bidaurrazaga",
"Arkaitz",
""
],
[
"Pérez",
"Aritz",
""
],
[
"Santana",
"Roberto",
""
]
] | TITLE: Structural Restricted Boltzmann Machine for image denoising and
classification
ABSTRACT: Restricted Boltzmann Machines are generative models that consist of a layer
of hidden variables connected to another layer of visible units, and they are
used to model the distribution over visible variables. In order to gain a
higher representability power, many hidden units are commonly used, which, in
combination with a large number of visible units, leads to a high number of
trainable parameters. In this work we introduce the Structural Restricted
Boltzmann Machine model, which taking advantage of the structure of the data in
hand, constrains connections of hidden units to subsets of visible units in
order to reduce significantly the number of trainable parameters, without
compromising performance. As a possible area of application, we focus on image
modelling. Based on the nature of the images, the structure of the connections
is given in terms of spatial neighbourhoods over the pixels of the image that
constitute the visible variables of the model. We conduct extensive experiments
on various image domains. Image denoising is evaluated with corrupted images
from the MNIST dataset. The generative power of our models is compared to
vanilla RBMs, as well as their classification performance, which is assessed
with five different image domains. Results show that our proposed model has a
faster and more stable training, while also obtaining better results compared
to an RBM with no constrained connections between its visible and hidden units.
|
1904.09757 | Tong Chen | Haojie Liu, Tong Chen, Peiyao Guo, Qiu Shen, Xun Cao, Yao Wang, Zhan
Ma | Non-local Attention Optimized Deep Image Compression | null | IEEE Transactions on Image Processing, vol. 30, pp. 3179-3191,
2021 | 10.1109/TIP.2021.3058615 | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a novel Non-Local Attention Optimized Deep Image
Compression (NLAIC) framework, which is built on top of the popular variational
auto-encoder (VAE) structure. Our NLAIC framework embeds non-local operations
in the encoders and decoders for both image and latent feature probability
information (known as hyperprior) to capture both local and global
correlations, and apply attention mechanism to generate masks that are used to
weigh the features for the image and hyperprior, which implicitly adapt bit
allocation for different features based on their importance. Furthermore, both
hyperpriors and spatial-channel neighbors of the latent features are used to
improve entropy coding. The proposed model outperforms the existing methods on
Kodak dataset, including learned (e.g., Balle2019, Balle2018) and conventional
(e.g., BPG, JPEG2000, JPEG) image compression methods, for both PSNR and
MS-SSIM distortion metrics.
| [
{
"version": "v1",
"created": "Mon, 22 Apr 2019 07:51:17 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Liu",
"Haojie",
""
],
[
"Chen",
"Tong",
""
],
[
"Guo",
"Peiyao",
""
],
[
"Shen",
"Qiu",
""
],
[
"Cao",
"Xun",
""
],
[
"Wang",
"Yao",
""
],
[
"Ma",
"Zhan",
""
]
] | TITLE: Non-local Attention Optimized Deep Image Compression
ABSTRACT: This paper proposes a novel Non-Local Attention Optimized Deep Image
Compression (NLAIC) framework, which is built on top of the popular variational
auto-encoder (VAE) structure. Our NLAIC framework embeds non-local operations
in the encoders and decoders for both image and latent feature probability
information (known as hyperprior) to capture both local and global
correlations, and apply attention mechanism to generate masks that are used to
weigh the features for the image and hyperprior, which implicitly adapt bit
allocation for different features based on their importance. Furthermore, both
hyperpriors and spatial-channel neighbors of the latent features are used to
improve entropy coding. The proposed model outperforms the existing methods on
Kodak dataset, including learned (e.g., Balle2019, Balle2018) and conventional
(e.g., BPG, JPEG2000, JPEG) image compression methods, for both PSNR and
MS-SSIM distortion metrics.
|
2011.13079 | Fnu Shilpika | Fnu Shilpika, Takanori Fujiwara, Naohisa Sakamoto, Jorji Nonaka,
Kwan-Liu Ma | A Visual Analytics Approach for Hardware System Monitoring with
Streaming Functional Data Analysis | 10 pages, 10 figures | null | 10.1109/TVCG.2022.3165348 | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many real-world applications involve analyzing time-dependent phenomena,
which are intrinsically functional, consisting of curves varying over a
continuum (e.g., time). When analyzing continuous data, functional data
analysis (FDA) provides substantial benefits, such as the ability to study the
derivatives and to restrict the ordering of data. However, continuous data
inherently has infinite dimensions, and for a long time series, FDA methods
often suffer from high computational costs. The analysis problem becomes even
more challenging when updating the FDA results for continuously arriving data.
In this paper, we present a visual analytics approach for monitoring and
reviewing time series data streamed from a hardware system with a focus on
identifying outliers by using FDA. To perform FDA while addressing the
computational problem, we introduce new incremental and progressive algorithms
that promptly generate the magnitude-shape (MS) plot, which conveys both the
functional magnitude and shape outlyingness of time series data. In addition,
by using an MS plot in conjunction with an FDA version of principal component
analysis, we enhance the analyst's ability to investigate the
visually-identified outliers. We illustrate the effectiveness of our approach
with two use scenarios using real-world datasets. The resulting tool is
evaluated by industry experts using real-world streaming datasets.
| [
{
"version": "v1",
"created": "Thu, 26 Nov 2020 00:48:06 GMT"
},
{
"version": "v2",
"created": "Sat, 23 Oct 2021 04:09:22 GMT"
},
{
"version": "v3",
"created": "Mon, 21 Feb 2022 21:00:51 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Shilpika",
"Fnu",
""
],
[
"Fujiwara",
"Takanori",
""
],
[
"Sakamoto",
"Naohisa",
""
],
[
"Nonaka",
"Jorji",
""
],
[
"Ma",
"Kwan-Liu",
""
]
] | TITLE: A Visual Analytics Approach for Hardware System Monitoring with
Streaming Functional Data Analysis
ABSTRACT: Many real-world applications involve analyzing time-dependent phenomena,
which are intrinsically functional, consisting of curves varying over a
continuum (e.g., time). When analyzing continuous data, functional data
analysis (FDA) provides substantial benefits, such as the ability to study the
derivatives and to restrict the ordering of data. However, continuous data
inherently has infinite dimensions, and for a long time series, FDA methods
often suffer from high computational costs. The analysis problem becomes even
more challenging when updating the FDA results for continuously arriving data.
In this paper, we present a visual analytics approach for monitoring and
reviewing time series data streamed from a hardware system with a focus on
identifying outliers by using FDA. To perform FDA while addressing the
computational problem, we introduce new incremental and progressive algorithms
that promptly generate the magnitude-shape (MS) plot, which conveys both the
functional magnitude and shape outlyingness of time series data. In addition,
by using an MS plot in conjunction with an FDA version of principal component
analysis, we enhance the analyst's ability to investigate the
visually-identified outliers. We illustrate the effectiveness of our approach
with two use scenarios using real-world datasets. The resulting tool is
evaluated by industry experts using real-world streaming datasets.
|
2111.04550 | Ece I\c{s}ik-Polat | Ece Isik-Polat, Gorkem Polat, Altan Kocyigit | ARFED: Attack-Resistant Federated averaging based on outlier elimination | null | Future Generation Computer Systems, Volume 141, April 2023, Pages
626-650 | 10.1016/j.future.2022.12.003 | null | cs.LG cs.CR | http://creativecommons.org/licenses/by/4.0/ | In federated learning, each participant trains its local model with its own
data and a global model is formed at a trusted server by aggregating model
updates coming from these participants. Since the server has no effect and
visibility on the training procedure of the participants to ensure privacy, the
global model becomes vulnerable to attacks such as data poisoning and model
poisoning. Although many defense algorithms have recently been proposed to
address these attacks, they often make strong assumptions that do not agree
with the nature of federated learning, such as assuming Non-IID datasets.
Moreover, they mostly lack comprehensive experimental analyses. In this work,
we propose a defense algorithm called ARFED that does not make any assumptions
about data distribution, update similarity of participants, or the ratio of the
malicious participants. ARFED mainly considers the outlier status of
participant updates for each layer of the model architecture based on the
distance to the global model. Hence, the participants that do not have any
outlier layer are involved in model aggregation. We have performed extensive
experiments on diverse scenarios and shown that the proposed approach provides
a robust defense against different attacks. To test the defense capability of
the ARFED in different conditions, we considered label flipping, Byzantine, and
partial knowledge attacks for both IID and Non-IID settings in our experimental
evaluations. Moreover, we proposed a new attack, called organized partial
knowledge attack, where malicious participants use their training statistics
collaboratively to define a common poisoned model. We have shown that organized
partial knowledge attacks are more effective than independent attacks.
| [
{
"version": "v1",
"created": "Mon, 8 Nov 2021 15:00:44 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 10:49:38 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Isik-Polat",
"Ece",
""
],
[
"Polat",
"Gorkem",
""
],
[
"Kocyigit",
"Altan",
""
]
] | TITLE: ARFED: Attack-Resistant Federated averaging based on outlier elimination
ABSTRACT: In federated learning, each participant trains its local model with its own
data and a global model is formed at a trusted server by aggregating model
updates coming from these participants. Since the server has no effect and
visibility on the training procedure of the participants to ensure privacy, the
global model becomes vulnerable to attacks such as data poisoning and model
poisoning. Although many defense algorithms have recently been proposed to
address these attacks, they often make strong assumptions that do not agree
with the nature of federated learning, such as assuming Non-IID datasets.
Moreover, they mostly lack comprehensive experimental analyses. In this work,
we propose a defense algorithm called ARFED that does not make any assumptions
about data distribution, update similarity of participants, or the ratio of the
malicious participants. ARFED mainly considers the outlier status of
participant updates for each layer of the model architecture based on the
distance to the global model. Hence, the participants that do not have any
outlier layer are involved in model aggregation. We have performed extensive
experiments on diverse scenarios and shown that the proposed approach provides
a robust defense against different attacks. To test the defense capability of
the ARFED in different conditions, we considered label flipping, Byzantine, and
partial knowledge attacks for both IID and Non-IID settings in our experimental
evaluations. Moreover, we proposed a new attack, called organized partial
knowledge attack, where malicious participants use their training statistics
collaboratively to define a common poisoned model. We have shown that organized
partial knowledge attacks are more effective than independent attacks.
|
2203.11400 | Kiet Nguyen | Kiet Van Nguyen, Son Quoc Tran, Luan Thanh Nguyen, Tin Van Huynh, Son
T. Luu, Ngan Luu-Thuy Nguyen | VLSP 2021 - ViMRC Challenge: Vietnamese Machine Reading Comprehension | The 8th International Workshop on Vietnamese Language and Speech
Processing (VLSP 2021) | null | 10.25073/2588-1086/vnucsce.340 | null | cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | One of the emerging research trends in natural language understanding is
machine reading comprehension (MRC) which is the task to find answers to human
questions based on textual data. Existing Vietnamese datasets for MRC research
concentrate solely on answerable questions. However, in reality, questions can
be unanswerable for which the correct answer is not stated in the given textual
data. To address the weakness, we provide the research community with a
benchmark dataset named UIT-ViQuAD 2.0 for evaluating the MRC task and question
answering systems for the Vietnamese language. We use UIT-ViQuAD 2.0 as a
benchmark dataset for the challenge on Vietnamese MRC at the Eighth Workshop on
Vietnamese Language and Speech Processing (VLSP 2021). This task attracted 77
participant teams from 34 universities and other organizations. In this
article, we present details of the organization of the challenge, an overview
of the methods employed by shared-task participants, and the results. The
highest performances are 77.24% in F1-score and 67.43% in Exact Match on the
private test set. The Vietnamese MRC systems proposed by the top 3 teams use
XLM-RoBERTa, a powerful pre-trained language model based on the transformer
architecture. The UIT-ViQuAD 2.0 dataset motivates researchers to further
explore the Vietnamese machine reading comprehension task and related tasks
such as question answering, question generation, and natural language
inference.
| [
{
"version": "v1",
"created": "Tue, 22 Mar 2022 00:44:41 GMT"
},
{
"version": "v2",
"created": "Thu, 31 Mar 2022 23:51:41 GMT"
},
{
"version": "v3",
"created": "Mon, 4 Apr 2022 11:58:38 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Van Nguyen",
"Kiet",
""
],
[
"Tran",
"Son Quoc",
""
],
[
"Nguyen",
"Luan Thanh",
""
],
[
"Van Huynh",
"Tin",
""
],
[
"Luu",
"Son T.",
""
],
[
"Nguyen",
"Ngan Luu-Thuy",
""
]
] | TITLE: VLSP 2021 - ViMRC Challenge: Vietnamese Machine Reading Comprehension
ABSTRACT: One of the emerging research trends in natural language understanding is
machine reading comprehension (MRC) which is the task to find answers to human
questions based on textual data. Existing Vietnamese datasets for MRC research
concentrate solely on answerable questions. However, in reality, questions can
be unanswerable for which the correct answer is not stated in the given textual
data. To address the weakness, we provide the research community with a
benchmark dataset named UIT-ViQuAD 2.0 for evaluating the MRC task and question
answering systems for the Vietnamese language. We use UIT-ViQuAD 2.0 as a
benchmark dataset for the challenge on Vietnamese MRC at the Eighth Workshop on
Vietnamese Language and Speech Processing (VLSP 2021). This task attracted 77
participant teams from 34 universities and other organizations. In this
article, we present details of the organization of the challenge, an overview
of the methods employed by shared-task participants, and the results. The
highest performances are 77.24% in F1-score and 67.43% in Exact Match on the
private test set. The Vietnamese MRC systems proposed by the top 3 teams use
XLM-RoBERTa, a powerful pre-trained language model based on the transformer
architecture. The UIT-ViQuAD 2.0 dataset motivates researchers to further
explore the Vietnamese machine reading comprehension task and related tasks
such as question answering, question generation, and natural language
inference.
|
2205.12696 | Qingyu Tan | Qingyu Tan, Lu Xu, Lidong Bing, Hwee Tou Ng, Sharifah Mahani Aljunied | Revisiting DocRED -- Addressing the False Negative Problem in Relation
Extraction | Accepted by EMNLP 2022 | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | The DocRED dataset is one of the most popular and widely used benchmarks for
document-level relation extraction (RE). It adopts a recommend-revise
annotation scheme so as to have a large-scale annotated dataset. However, we
find that the annotation of DocRED is incomplete, i.e., false negative samples
are prevalent. We analyze the causes and effects of the overwhelming false
negative problem in the DocRED dataset. To address the shortcoming, we
re-annotate 4,053 documents in the DocRED dataset by adding the missed relation
triples back to the original DocRED. We name our revised DocRED dataset
Re-DocRED. We conduct extensive experiments with state-of-the-art neural models
on both datasets, and the experimental results show that the models trained and
evaluated on our Re-DocRED achieve performance improvements of around 13 F1
points. Moreover, we conduct a comprehensive analysis to identify the potential
areas for further improvement. Our dataset is publicly available at
https://github.com/tonytan48/Re-DocRED.
| [
{
"version": "v1",
"created": "Wed, 25 May 2022 11:54:48 GMT"
},
{
"version": "v2",
"created": "Tue, 25 Oct 2022 09:36:12 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Jun 2023 05:44:07 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Tan",
"Qingyu",
""
],
[
"Xu",
"Lu",
""
],
[
"Bing",
"Lidong",
""
],
[
"Ng",
"Hwee Tou",
""
],
[
"Aljunied",
"Sharifah Mahani",
""
]
] | TITLE: Revisiting DocRED -- Addressing the False Negative Problem in Relation
Extraction
ABSTRACT: The DocRED dataset is one of the most popular and widely used benchmarks for
document-level relation extraction (RE). It adopts a recommend-revise
annotation scheme so as to have a large-scale annotated dataset. However, we
find that the annotation of DocRED is incomplete, i.e., false negative samples
are prevalent. We analyze the causes and effects of the overwhelming false
negative problem in the DocRED dataset. To address the shortcoming, we
re-annotate 4,053 documents in the DocRED dataset by adding the missed relation
triples back to the original DocRED. We name our revised DocRED dataset
Re-DocRED. We conduct extensive experiments with state-of-the-art neural models
on both datasets, and the experimental results show that the models trained and
evaluated on our Re-DocRED achieve performance improvements of around 13 F1
points. Moreover, we conduct a comprehensive analysis to identify the potential
areas for further improvement. Our dataset is publicly available at
https://github.com/tonytan48/Re-DocRED.
|
2206.00583 | Yifan Chen | Yifan Chen, Tianning Xu, Dilek Hakkani-Tur, Di Jin, Yun Yang, Ruoqing
Zhu | Calibrate and Debias Layer-wise Sampling for Graph Convolutional
Networks | Published at TMLR. Code is available at
https://github.com/ychen-stat-ml/GCN-layer-wise-sampling | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multiple sampling-based methods have been developed for approximating and
accelerating node embedding aggregation in graph convolutional networks (GCNs)
training. Among them, a layer-wise approach recursively performs importance
sampling to select neighbors jointly for existing nodes in each layer. This
paper revisits the approach from a matrix approximation perspective, and
identifies two issues in the existing layer-wise sampling methods: suboptimal
sampling probabilities and estimation biases induced by sampling without
replacement. To address these issues, we accordingly propose two remedies: a
new principle for constructing sampling probabilities and an efficient
debiasing algorithm. The improvements are demonstrated by extensive analyses of
estimation variance and experiments on common benchmarks. Code and algorithm
implementations are publicly available at
https://github.com/ychen-stat-ml/GCN-layer-wise-sampling .
| [
{
"version": "v1",
"created": "Wed, 1 Jun 2022 15:52:06 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 20:26:40 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Chen",
"Yifan",
""
],
[
"Xu",
"Tianning",
""
],
[
"Hakkani-Tur",
"Dilek",
""
],
[
"Jin",
"Di",
""
],
[
"Yang",
"Yun",
""
],
[
"Zhu",
"Ruoqing",
""
]
] | TITLE: Calibrate and Debias Layer-wise Sampling for Graph Convolutional
Networks
ABSTRACT: Multiple sampling-based methods have been developed for approximating and
accelerating node embedding aggregation in graph convolutional networks (GCNs)
training. Among them, a layer-wise approach recursively performs importance
sampling to select neighbors jointly for existing nodes in each layer. This
paper revisits the approach from a matrix approximation perspective, and
identifies two issues in the existing layer-wise sampling methods: suboptimal
sampling probabilities and estimation biases induced by sampling without
replacement. To address these issues, we accordingly propose two remedies: a
new principle for constructing sampling probabilities and an efficient
debiasing algorithm. The improvements are demonstrated by extensive analyses of
estimation variance and experiments on common benchmarks. Code and algorithm
implementations are publicly available at
https://github.com/ychen-stat-ml/GCN-layer-wise-sampling .
|
2206.02620 | Wanqi Xue | Wanqi Xue, Qingpeng Cai, Ruohan Zhan, Dong Zheng, Peng Jiang, Kun Gai,
Bo An | ResAct: Reinforcing Long-term Engagement in Sequential Recommendation
with Residual Actor | Accpetd by ICLR 2023 | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Long-term engagement is preferred over immediate engagement in sequential
recommendation as it directly affects product operational metrics such as daily
active users (DAUs) and dwell time. Meanwhile, reinforcement learning (RL) is
widely regarded as a promising framework for optimizing long-term engagement in
sequential recommendation. However, due to expensive online interactions, it is
very difficult for RL algorithms to perform state-action value estimation,
exploration and feature extraction when optimizing long-term engagement. In
this paper, we propose ResAct which seeks a policy that is close to, but better
than, the online-serving policy. In this way, we can collect sufficient data
near the learned policy so that state-action values can be properly estimated,
and there is no need to perform online exploration. ResAct optimizes the policy
by first reconstructing the online behaviors and then improving it via a
Residual Actor. To extract long-term information, ResAct utilizes two
information-theoretical regularizers to confirm the expressiveness and
conciseness of features. We conduct experiments on a benchmark dataset and a
large-scale industrial dataset which consists of tens of millions of
recommendation requests. Experimental results show that our method
significantly outperforms the state-of-the-art baselines in various long-term
engagement optimization tasks.
| [
{
"version": "v1",
"created": "Wed, 1 Jun 2022 02:45:35 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 08:38:52 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Xue",
"Wanqi",
""
],
[
"Cai",
"Qingpeng",
""
],
[
"Zhan",
"Ruohan",
""
],
[
"Zheng",
"Dong",
""
],
[
"Jiang",
"Peng",
""
],
[
"Gai",
"Kun",
""
],
[
"An",
"Bo",
""
]
] | TITLE: ResAct: Reinforcing Long-term Engagement in Sequential Recommendation
with Residual Actor
ABSTRACT: Long-term engagement is preferred over immediate engagement in sequential
recommendation as it directly affects product operational metrics such as daily
active users (DAUs) and dwell time. Meanwhile, reinforcement learning (RL) is
widely regarded as a promising framework for optimizing long-term engagement in
sequential recommendation. However, due to expensive online interactions, it is
very difficult for RL algorithms to perform state-action value estimation,
exploration and feature extraction when optimizing long-term engagement. In
this paper, we propose ResAct which seeks a policy that is close to, but better
than, the online-serving policy. In this way, we can collect sufficient data
near the learned policy so that state-action values can be properly estimated,
and there is no need to perform online exploration. ResAct optimizes the policy
by first reconstructing the online behaviors and then improving it via a
Residual Actor. To extract long-term information, ResAct utilizes two
information-theoretical regularizers to confirm the expressiveness and
conciseness of features. We conduct experiments on a benchmark dataset and a
large-scale industrial dataset which consists of tens of millions of
recommendation requests. Experimental results show that our method
significantly outperforms the state-of-the-art baselines in various long-term
engagement optimization tasks.
|
2206.04039 | Divyansh Kaushik | Divyansh Kaushik, Zachary C. Lipton, Alex John London | Resolving the Human Subjects Status of Machine Learning's Crowdworkers | null | null | null | null | cs.CY cs.AI cs.CL cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent years, machine learning (ML) has relied heavily on crowdworkers
both for building datasets and for addressing research questions requiring
human interaction or judgment. The diverse tasks performed and uses of the data
produced render it difficult to determine when crowdworkers are best thought of
as workers (versus human subjects). These difficulties are compounded by
conflicting policies, with some institutions and researchers regarding all ML
crowdworkers as human subjects and others holding that they rarely constitute
human subjects. Notably few ML papers involving crowdwork mention IRB
oversight, raising the prospect of non-compliance with ethical and regulatory
requirements. We investigate the appropriate designation of ML crowdsourcing
studies, focusing our inquiry on natural language processing to expose unique
challenges for research oversight. Crucially, under the U.S. Common Rule, these
judgments hinge on determinations of aboutness, concerning both whom (or what)
the collected data is about and whom (or what) the analysis is about. We
highlight two challenges posed by ML: the same set of workers can serve
multiple roles and provide many sorts of information; and ML research tends to
embrace a dynamic workflow, where research questions are seldom stated ex ante
and data sharing opens the door for future studies to aim questions at
different targets. Our analysis exposes a potential loophole in the Common
Rule, where researchers can elude research ethics oversight by splitting data
collection and analysis into distinct studies. Finally, we offer several policy
recommendations to address these concerns.
| [
{
"version": "v1",
"created": "Wed, 8 Jun 2022 17:55:01 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 20:10:08 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Kaushik",
"Divyansh",
""
],
[
"Lipton",
"Zachary C.",
""
],
[
"London",
"Alex John",
""
]
] | TITLE: Resolving the Human Subjects Status of Machine Learning's Crowdworkers
ABSTRACT: In recent years, machine learning (ML) has relied heavily on crowdworkers
both for building datasets and for addressing research questions requiring
human interaction or judgment. The diverse tasks performed and uses of the data
produced render it difficult to determine when crowdworkers are best thought of
as workers (versus human subjects). These difficulties are compounded by
conflicting policies, with some institutions and researchers regarding all ML
crowdworkers as human subjects and others holding that they rarely constitute
human subjects. Notably few ML papers involving crowdwork mention IRB
oversight, raising the prospect of non-compliance with ethical and regulatory
requirements. We investigate the appropriate designation of ML crowdsourcing
studies, focusing our inquiry on natural language processing to expose unique
challenges for research oversight. Crucially, under the U.S. Common Rule, these
judgments hinge on determinations of aboutness, concerning both whom (or what)
the collected data is about and whom (or what) the analysis is about. We
highlight two challenges posed by ML: the same set of workers can serve
multiple roles and provide many sorts of information; and ML research tends to
embrace a dynamic workflow, where research questions are seldom stated ex ante
and data sharing opens the door for future studies to aim questions at
different targets. Our analysis exposes a potential loophole in the Common
Rule, where researchers can elude research ethics oversight by splitting data
collection and analysis into distinct studies. Finally, we offer several policy
recommendations to address these concerns.
|
2208.05596 | Tristan Vanderbruggen | Patrick Flynn and Tristan Vanderbruggen and Chunhua Liao and Pei-Hung
Lin and Murali Emani and Xipeng Shen | Finding Reusable Machine Learning Components to Build Programming
Language Processing Pipelines | null | null | null | null | cs.LG cs.PL | http://creativecommons.org/licenses/by/4.0/ | Programming Language Processing (PLP) using machine learning has made vast
improvements in the past few years. Increasingly more people are interested in
exploring this promising field. However, it is challenging for new researchers
and developers to find the right components to construct their own machine
learning pipelines, given the diverse PLP tasks to be solved, the large number
of datasets and models being released, and the set of complex compilers or
tools involved. To improve the findability, accessibility, interoperability and
reusability (FAIRness) of machine learning components, we collect and analyze a
set of representative papers in the domain of machine learning-based PLP. We
then identify and characterize key concepts including PLP tasks, model
architectures and supportive tools. Finally, we show some example use cases of
leveraging the reusable components to construct machine learning pipelines to
solve a set of PLP tasks.
| [
{
"version": "v1",
"created": "Thu, 11 Aug 2022 00:24:10 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 18:25:33 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Flynn",
"Patrick",
""
],
[
"Vanderbruggen",
"Tristan",
""
],
[
"Liao",
"Chunhua",
""
],
[
"Lin",
"Pei-Hung",
""
],
[
"Emani",
"Murali",
""
],
[
"Shen",
"Xipeng",
""
]
] | TITLE: Finding Reusable Machine Learning Components to Build Programming
Language Processing Pipelines
ABSTRACT: Programming Language Processing (PLP) using machine learning has made vast
improvements in the past few years. Increasingly more people are interested in
exploring this promising field. However, it is challenging for new researchers
and developers to find the right components to construct their own machine
learning pipelines, given the diverse PLP tasks to be solved, the large number
of datasets and models being released, and the set of complex compilers or
tools involved. To improve the findability, accessibility, interoperability and
reusability (FAIRness) of machine learning components, we collect and analyze a
set of representative papers in the domain of machine learning-based PLP. We
then identify and characterize key concepts including PLP tasks, model
architectures and supportive tools. Finally, we show some example use cases of
leveraging the reusable components to construct machine learning pipelines to
solve a set of PLP tasks.
|
2210.04081 | Jaemin Yoo | Jaemin Yoo, Meng-Chieh Lee, Shubhranshu Shekhar, and Christos
Faloutsos | Less is More: SlimG for Accurate, Robust, and Interpretable Graph Mining | Accepted to KDD 2023 | null | null | null | cs.LG cs.SI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | How can we solve semi-supervised node classification in various graphs
possibly with noisy features and structures? Graph neural networks (GNNs) have
succeeded in many graph mining tasks, but their generalizability to various
graph scenarios is limited due to the difficulty of training, hyperparameter
tuning, and the selection of a model itself. Einstein said that we should "make
everything as simple as possible, but not simpler." We rephrase it into the
careful simplicity principle: a carefully-designed simple model can surpass
sophisticated ones in real-world graphs. Based on the principle, we propose
SlimG for semi-supervised node classification, which exhibits four desirable
properties: It is (a) accurate, winning or tying on 10 out of 13 real-world
datasets; (b) robust, being the only one that handles all scenarios of graph
data (homophily, heterophily, random structure, noisy features, etc.); (c) fast
and scalable, showing up to 18 times faster training in million-scale graphs;
and (d) interpretable, thanks to the linearity and sparsity. We explain the
success of SlimG through a systematic study of the designs of existing GNNs,
sanity checks, and comprehensive ablation studies.
| [
{
"version": "v1",
"created": "Sat, 8 Oct 2022 18:24:36 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Oct 2022 15:36:09 GMT"
},
{
"version": "v3",
"created": "Fri, 18 Nov 2022 03:10:55 GMT"
},
{
"version": "v4",
"created": "Fri, 16 Jun 2023 07:32:17 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Yoo",
"Jaemin",
""
],
[
"Lee",
"Meng-Chieh",
""
],
[
"Shekhar",
"Shubhranshu",
""
],
[
"Faloutsos",
"Christos",
""
]
] | TITLE: Less is More: SlimG for Accurate, Robust, and Interpretable Graph Mining
ABSTRACT: How can we solve semi-supervised node classification in various graphs
possibly with noisy features and structures? Graph neural networks (GNNs) have
succeeded in many graph mining tasks, but their generalizability to various
graph scenarios is limited due to the difficulty of training, hyperparameter
tuning, and the selection of a model itself. Einstein said that we should "make
everything as simple as possible, but not simpler." We rephrase it into the
careful simplicity principle: a carefully-designed simple model can surpass
sophisticated ones in real-world graphs. Based on the principle, we propose
SlimG for semi-supervised node classification, which exhibits four desirable
properties: It is (a) accurate, winning or tying on 10 out of 13 real-world
datasets; (b) robust, being the only one that handles all scenarios of graph
data (homophily, heterophily, random structure, noisy features, etc.); (c) fast
and scalable, showing up to 18 times faster training in million-scale graphs;
and (d) interpretable, thanks to the linearity and sparsity. We explain the
success of SlimG through a systematic study of the designs of existing GNNs,
sanity checks, and comprehensive ablation studies.
|
2211.06588 | Haodong Ouyang | Haodong Ouyang | DEYO: DETR with YOLO for Step-by-Step Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detection is an important topic in computer vision, with
post-processing, an essential part of the typical object detection pipeline,
posing a significant bottleneck affecting the performance of traditional object
detection models. The detection transformer (DETR), as the first end-to-end
target detection model, discards the requirement of manual components like the
anchor and non-maximum suppression (NMS), significantly simplifying the target
detection process. However, compared with most traditional object detection
models, DETR converges very slowly, and a query's meaning is obscure. Thus,
inspired by the Step-by-Step concept, this paper proposes a new two-stage
object detection model, named DETR with YOLO (DEYO), which relies on a
progressive inference to solve the above problems. DEYO is a two-stage
architecture comprising a classic target detection model and a DETR-like model
as the first and second stages, respectively. Specifically, the first stage
provides high-quality query and anchor feeding into the second stage, improving
the performance and efficiency of the second stage compared to the original
DETR model. Meanwhile, the second stage compensates for the performance
degradation caused by the first stage detector's limitations. Extensive
experiments demonstrate that DEYO attains 50.6 AP and 52.1 AP in 12 and 36
epochs, respectively, while utilizing ResNet-50 as the backbone and multi-scale
features on the COCO dataset. Compared with DINO, an optimal DETR-like model,
the developed DEYO model affords a significant performance improvement of 1.6
AP and 1.2 AP in two epoch settings.
| [
{
"version": "v1",
"created": "Sat, 12 Nov 2022 06:36:17 GMT"
},
{
"version": "v2",
"created": "Fri, 18 Nov 2022 22:07:57 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Jun 2023 03:49:48 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Ouyang",
"Haodong",
""
]
] | TITLE: DEYO: DETR with YOLO for Step-by-Step Object Detection
ABSTRACT: Object detection is an important topic in computer vision, with
post-processing, an essential part of the typical object detection pipeline,
posing a significant bottleneck affecting the performance of traditional object
detection models. The detection transformer (DETR), as the first end-to-end
target detection model, discards the requirement of manual components like the
anchor and non-maximum suppression (NMS), significantly simplifying the target
detection process. However, compared with most traditional object detection
models, DETR converges very slowly, and a query's meaning is obscure. Thus,
inspired by the Step-by-Step concept, this paper proposes a new two-stage
object detection model, named DETR with YOLO (DEYO), which relies on a
progressive inference to solve the above problems. DEYO is a two-stage
architecture comprising a classic target detection model and a DETR-like model
as the first and second stages, respectively. Specifically, the first stage
provides high-quality query and anchor feeding into the second stage, improving
the performance and efficiency of the second stage compared to the original
DETR model. Meanwhile, the second stage compensates for the performance
degradation caused by the first stage detector's limitations. Extensive
experiments demonstrate that DEYO attains 50.6 AP and 52.1 AP in 12 and 36
epochs, respectively, while utilizing ResNet-50 as the backbone and multi-scale
features on the COCO dataset. Compared with DINO, an optimal DETR-like model,
the developed DEYO model affords a significant performance improvement of 1.6
AP and 1.2 AP in two epoch settings.
|
2211.16963 | Saurav Sharma | Saurav Sharma, Chinedu Innocent Nwoye, Didier Mutter, Nicolas Padoy | Rendezvous in Time: An Attention-based Temporal Fusion approach for
Surgical Triplet Recognition | Accepted at IPCAI, 2023. Project Page:
https://github.com/CAMMA-public/rendezvous-in-time | null | 10.1007/s11548-023-02914-1 | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | One of the recent advances in surgical AI is the recognition of surgical
activities as triplets of (instrument, verb, target). Albeit providing detailed
information for computer-assisted intervention, current triplet recognition
approaches rely only on single frame features. Exploiting the temporal cues
from earlier frames would improve the recognition of surgical action triplets
from videos. In this paper, we propose Rendezvous in Time (RiT) - a deep
learning model that extends the state-of-the-art model, Rendezvous, with
temporal modeling. Focusing more on the verbs, our RiT explores the
connectedness of current and past frames to learn temporal attention-based
features for enhanced triplet recognition. We validate our proposal on the
challenging surgical triplet dataset, CholecT45, demonstrating an improved
recognition of the verb and triplet along with other interactions involving the
verb such as (instrument, verb). Qualitative results show that the RiT produces
smoother predictions for most triplet instances than the state-of-the-arts. We
present a novel attention-based approach that leverages the temporal fusion of
video frames to model the evolution of surgical actions and exploit their
benefits for surgical triplet recognition.
| [
{
"version": "v1",
"created": "Wed, 30 Nov 2022 13:18:07 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 09:55:05 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Sharma",
"Saurav",
""
],
[
"Nwoye",
"Chinedu Innocent",
""
],
[
"Mutter",
"Didier",
""
],
[
"Padoy",
"Nicolas",
""
]
] | TITLE: Rendezvous in Time: An Attention-based Temporal Fusion approach for
Surgical Triplet Recognition
ABSTRACT: One of the recent advances in surgical AI is the recognition of surgical
activities as triplets of (instrument, verb, target). Albeit providing detailed
information for computer-assisted intervention, current triplet recognition
approaches rely only on single frame features. Exploiting the temporal cues
from earlier frames would improve the recognition of surgical action triplets
from videos. In this paper, we propose Rendezvous in Time (RiT) - a deep
learning model that extends the state-of-the-art model, Rendezvous, with
temporal modeling. Focusing more on the verbs, our RiT explores the
connectedness of current and past frames to learn temporal attention-based
features for enhanced triplet recognition. We validate our proposal on the
challenging surgical triplet dataset, CholecT45, demonstrating an improved
recognition of the verb and triplet along with other interactions involving the
verb such as (instrument, verb). Qualitative results show that the RiT produces
smoother predictions for most triplet instances than the state-of-the-arts. We
present a novel attention-based approach that leverages the temporal fusion of
video frames to model the evolution of surgical actions and exploit their
benefits for surgical triplet recognition.
|
2212.04074 | Xiaohan Zhang | Xiaohan Zhang, Xingyu Li, Waqas Sultani, Yi Zhou, Safwan Wshah | Cross-view Geo-localization via Learning Disentangled Geometric Layout
Correspondence | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cross-view geo-localization aims to estimate the location of a query ground
image by matching it to a reference geo-tagged aerial images database. As an
extremely challenging task, its difficulties root in the drastic view changes
and different capturing time between two views. Despite these difficulties,
recent works achieve outstanding progress on cross-view geo-localization
benchmarks. However, existing methods still suffer from poor performance on the
cross-area benchmarks, in which the training and testing data are captured from
two different regions. We attribute this deficiency to the lack of ability to
extract the spatial configuration of visual feature layouts and models'
overfitting on low-level details from the training set. In this paper, we
propose GeoDTR which explicitly disentangles geometric information from raw
features and learns the spatial correlations among visual features from aerial
and ground pairs with a novel geometric layout extractor module. This module
generates a set of geometric layout descriptors, modulating the raw features
and producing high-quality latent representations. In addition, we elaborate on
two categories of data augmentations, (i) Layout simulation, which varies the
spatial configuration while keeping the low-level details intact. (ii) Semantic
augmentation, which alters the low-level details and encourages the model to
capture spatial configurations. These augmentations help to improve the
performance of the cross-view geo-localization models, especially on the
cross-area benchmarks. Moreover, we propose a counterfactual-based learning
process to benefit the geometric layout extractor in exploring spatial
information. Extensive experiments show that GeoDTR not only achieves
state-of-the-art results but also significantly boosts the performance on
same-area and cross-area benchmarks.
| [
{
"version": "v1",
"created": "Thu, 8 Dec 2022 04:54:01 GMT"
},
{
"version": "v2",
"created": "Tue, 20 Dec 2022 00:32:07 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Jun 2023 17:57:45 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Zhang",
"Xiaohan",
""
],
[
"Li",
"Xingyu",
""
],
[
"Sultani",
"Waqas",
""
],
[
"Zhou",
"Yi",
""
],
[
"Wshah",
"Safwan",
""
]
] | TITLE: Cross-view Geo-localization via Learning Disentangled Geometric Layout
Correspondence
ABSTRACT: Cross-view geo-localization aims to estimate the location of a query ground
image by matching it to a reference geo-tagged aerial images database. As an
extremely challenging task, its difficulties root in the drastic view changes
and different capturing time between two views. Despite these difficulties,
recent works achieve outstanding progress on cross-view geo-localization
benchmarks. However, existing methods still suffer from poor performance on the
cross-area benchmarks, in which the training and testing data are captured from
two different regions. We attribute this deficiency to the lack of ability to
extract the spatial configuration of visual feature layouts and models'
overfitting on low-level details from the training set. In this paper, we
propose GeoDTR which explicitly disentangles geometric information from raw
features and learns the spatial correlations among visual features from aerial
and ground pairs with a novel geometric layout extractor module. This module
generates a set of geometric layout descriptors, modulating the raw features
and producing high-quality latent representations. In addition, we elaborate on
two categories of data augmentations, (i) Layout simulation, which varies the
spatial configuration while keeping the low-level details intact. (ii) Semantic
augmentation, which alters the low-level details and encourages the model to
capture spatial configurations. These augmentations help to improve the
performance of the cross-view geo-localization models, especially on the
cross-area benchmarks. Moreover, we propose a counterfactual-based learning
process to benefit the geometric layout extractor in exploring spatial
information. Extensive experiments show that GeoDTR not only achieves
state-of-the-art results but also significantly boosts the performance on
same-area and cross-area benchmarks.
|
2212.09381 | Jianwu Fang | Jianwu Fang, Lei-Lei Li, Kuan Yang, Zhedong Zheng, Jianru Xue, and
Tat-Seng Chua | Cognitive Accident Prediction in Driving Scenes: A Multimodality
Benchmark | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traffic accident prediction in driving videos aims to provide an early
warning of the accident occurrence, and supports the decision making of safe
driving systems. Previous works usually concentrate on the spatial-temporal
correlation of object-level context, while they do not fit the inherent
long-tailed data distribution well and are vulnerable to severe environmental
change. In this work, we propose a Cognitive Accident Prediction (CAP) method
that explicitly leverages human-inspired cognition of text description on the
visual observation and the driver attention to facilitate model training. In
particular, the text description provides a dense semantic description guidance
for the primary context of the traffic scene, while the driver attention
provides a traction to focus on the critical region closely correlating with
safe driving. CAP is formulated by an attentive text-to-vision shift fusion
module, an attentive scene context transfer module, and the driver attention
guided accident prediction module. We leverage the attention mechanism in these
modules to explore the core semantic cues for accident prediction. In order to
train CAP, we extend an existing self-collected DADA-2000 dataset (with
annotated driver attention for each frame) with further factual text
descriptions for the visual observations before the accidents. Besides, we
construct a new large-scale benchmark consisting of 11,727 in-the-wild accident
videos with over 2.19 million frames (named as CAP-DATA) together with labeled
fact-effect-reason-introspection description and temporal accident frame label.
Based on extensive experiments, the superiority of CAP is validated compared
with state-of-the-art approaches. The code, CAP-DATA, and all results will be
released in \url{https://github.com/JWFanggit/LOTVS-CAP}.
| [
{
"version": "v1",
"created": "Mon, 19 Dec 2022 11:43:02 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 13:29:45 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Fang",
"Jianwu",
""
],
[
"Li",
"Lei-Lei",
""
],
[
"Yang",
"Kuan",
""
],
[
"Zheng",
"Zhedong",
""
],
[
"Xue",
"Jianru",
""
],
[
"Chua",
"Tat-Seng",
""
]
] | TITLE: Cognitive Accident Prediction in Driving Scenes: A Multimodality
Benchmark
ABSTRACT: Traffic accident prediction in driving videos aims to provide an early
warning of the accident occurrence, and supports the decision making of safe
driving systems. Previous works usually concentrate on the spatial-temporal
correlation of object-level context, while they do not fit the inherent
long-tailed data distribution well and are vulnerable to severe environmental
change. In this work, we propose a Cognitive Accident Prediction (CAP) method
that explicitly leverages human-inspired cognition of text description on the
visual observation and the driver attention to facilitate model training. In
particular, the text description provides a dense semantic description guidance
for the primary context of the traffic scene, while the driver attention
provides a traction to focus on the critical region closely correlating with
safe driving. CAP is formulated by an attentive text-to-vision shift fusion
module, an attentive scene context transfer module, and the driver attention
guided accident prediction module. We leverage the attention mechanism in these
modules to explore the core semantic cues for accident prediction. In order to
train CAP, we extend an existing self-collected DADA-2000 dataset (with
annotated driver attention for each frame) with further factual text
descriptions for the visual observations before the accidents. Besides, we
construct a new large-scale benchmark consisting of 11,727 in-the-wild accident
videos with over 2.19 million frames (named as CAP-DATA) together with labeled
fact-effect-reason-introspection description and temporal accident frame label.
Based on extensive experiments, the superiority of CAP is validated compared
with state-of-the-art approaches. The code, CAP-DATA, and all results will be
released in \url{https://github.com/JWFanggit/LOTVS-CAP}.
|
2212.10525 | Suwon Shon | Suwon Shon, Siddhant Arora, Chyi-Jiunn Lin, Ankita Pasad, Felix Wu,
Roshan Sharma, Wei-Lun Wu, Hung-Yi Lee, Karen Livescu, Shinji Watanabe | SLUE Phase-2: A Benchmark Suite of Diverse Spoken Language Understanding
Tasks | accepted in ACL 2023 (long paper) | null | null | null | cs.CL eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spoken language understanding (SLU) tasks have been studied for many decades
in the speech research community, but have not received as much attention as
lower-level tasks like speech and speaker recognition. In particular, there are
not nearly as many SLU task benchmarks, and many of the existing ones use data
that is not freely available to all researchers. Recent work has begun to
introduce such benchmark datasets for several tasks. In this work, we introduce
several new annotated SLU benchmark tasks based on freely available speech
data, which complement existing benchmarks and address gaps in the SLU
evaluation landscape. We contribute four tasks: question answering and
summarization involve inference over longer speech sequences; named entity
localization addresses the speech-specific task of locating the targeted
content in the signal; dialog act classification identifies the function of a
given speech utterance. We follow the blueprint of the Spoken Language
Understanding Evaluation (SLUE) benchmark suite. In order to facilitate the
development of SLU models that leverage the success of pre-trained speech
representations, we will be publishing for each task (i) annotations for a
relatively small fine-tuning set, (ii) annotated development and test sets, and
(iii) baseline models for easy reproducibility and comparisons. In this work,
we present the details of data collection and annotation and the performance of
the baseline models. We also perform sensitivity analysis of pipeline models'
performance (speech recognizer + text model) to the speech recognition
accuracy, using more than 20 state-of-the-art speech recognition models.
| [
{
"version": "v1",
"created": "Tue, 20 Dec 2022 18:39:59 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 22:51:09 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Shon",
"Suwon",
""
],
[
"Arora",
"Siddhant",
""
],
[
"Lin",
"Chyi-Jiunn",
""
],
[
"Pasad",
"Ankita",
""
],
[
"Wu",
"Felix",
""
],
[
"Sharma",
"Roshan",
""
],
[
"Wu",
"Wei-Lun",
""
],
[
"Lee",
"Hung-Yi",
""
],
[
"Livescu",
"Karen",
""
],
[
"Watanabe",
"Shinji",
""
]
] | TITLE: SLUE Phase-2: A Benchmark Suite of Diverse Spoken Language Understanding
Tasks
ABSTRACT: Spoken language understanding (SLU) tasks have been studied for many decades
in the speech research community, but have not received as much attention as
lower-level tasks like speech and speaker recognition. In particular, there are
not nearly as many SLU task benchmarks, and many of the existing ones use data
that is not freely available to all researchers. Recent work has begun to
introduce such benchmark datasets for several tasks. In this work, we introduce
several new annotated SLU benchmark tasks based on freely available speech
data, which complement existing benchmarks and address gaps in the SLU
evaluation landscape. We contribute four tasks: question answering and
summarization involve inference over longer speech sequences; named entity
localization addresses the speech-specific task of locating the targeted
content in the signal; dialog act classification identifies the function of a
given speech utterance. We follow the blueprint of the Spoken Language
Understanding Evaluation (SLUE) benchmark suite. In order to facilitate the
development of SLU models that leverage the success of pre-trained speech
representations, we will be publishing for each task (i) annotations for a
relatively small fine-tuning set, (ii) annotated development and test sets, and
(iii) baseline models for easy reproducibility and comparisons. In this work,
we present the details of data collection and annotation and the performance of
the baseline models. We also perform sensitivity analysis of pipeline models'
performance (speech recognizer + text model) to the speech recognition
accuracy, using more than 20 state-of-the-art speech recognition models.
|
2302.01736 | Corentin Puffay | Corentin Puffay, Bernd Accou, Lies Bollens, Mohammad Jalilpour Monesi,
Jonas Vanthornhout, Hugo Van hamme, Tom Francart | Relating EEG to continuous speech using deep neural networks: a review | null | null | null | null | eess.AS cs.SD | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Objective. When a person listens to continuous speech, a corresponding
response is elicited in the brain and can be recorded using
electroencephalography (EEG). Linear models are presently used to relate the
EEG recording to the corresponding speech signal. The ability of linear models
to find a mapping between these two signals is used as a measure of neural
tracking of speech. Such models are limited as they assume linearity in the
EEG-speech relationship, which omits the nonlinear dynamics of the brain. As an
alternative, deep learning models have recently been used to relate EEG to
continuous speech. Approach. This paper reviews and comments on
deep-learning-based studies that relate EEG to continuous speech in single- or
multiple-speakers paradigms. We point out recurrent methodological pitfalls and
the need for a standard benchmark of model analysis. Main results. We gathered
29 studies. The main methodological issues we found are biased
cross-validations, data leakage leading to over-fitted models, or
disproportionate data size compared to the model's complexity. In addition, we
address requirements for a standard benchmark model analysis, such as public
datasets, common evaluation metrics, and good practices for the match-mismatch
task. Significance. We present a review paper summarizing the main
deep-learning-based studies that relate EEG to speech while addressing
methodological pitfalls and important considerations for this newly expanding
field. Our study is particularly relevant given the growing application of deep
learning in EEG-speech decoding.
| [
{
"version": "v1",
"created": "Fri, 3 Feb 2023 13:51:01 GMT"
},
{
"version": "v2",
"created": "Mon, 6 Feb 2023 09:05:07 GMT"
},
{
"version": "v3",
"created": "Wed, 26 Apr 2023 12:26:37 GMT"
},
{
"version": "v4",
"created": "Fri, 16 Jun 2023 15:50:26 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Puffay",
"Corentin",
""
],
[
"Accou",
"Bernd",
""
],
[
"Bollens",
"Lies",
""
],
[
"Monesi",
"Mohammad Jalilpour",
""
],
[
"Vanthornhout",
"Jonas",
""
],
[
"Van hamme",
"Hugo",
""
],
[
"Francart",
"Tom",
""
]
] | TITLE: Relating EEG to continuous speech using deep neural networks: a review
ABSTRACT: Objective. When a person listens to continuous speech, a corresponding
response is elicited in the brain and can be recorded using
electroencephalography (EEG). Linear models are presently used to relate the
EEG recording to the corresponding speech signal. The ability of linear models
to find a mapping between these two signals is used as a measure of neural
tracking of speech. Such models are limited as they assume linearity in the
EEG-speech relationship, which omits the nonlinear dynamics of the brain. As an
alternative, deep learning models have recently been used to relate EEG to
continuous speech. Approach. This paper reviews and comments on
deep-learning-based studies that relate EEG to continuous speech in single- or
multiple-speakers paradigms. We point out recurrent methodological pitfalls and
the need for a standard benchmark of model analysis. Main results. We gathered
29 studies. The main methodological issues we found are biased
cross-validations, data leakage leading to over-fitted models, or
disproportionate data size compared to the model's complexity. In addition, we
address requirements for a standard benchmark model analysis, such as public
datasets, common evaluation metrics, and good practices for the match-mismatch
task. Significance. We present a review paper summarizing the main
deep-learning-based studies that relate EEG to speech while addressing
methodological pitfalls and important considerations for this newly expanding
field. Our study is particularly relevant given the growing application of deep
learning in EEG-speech decoding.
|
2302.02031 | Nguyen Phong Hoang | Jacob Brown, Xi Jiang, Van Tran, Arjun Nitin Bhagoji, Nguyen Phong
Hoang, Nick Feamster, Prateek Mittal, Vinod Yegneswaran | Augmenting Rule-based DNS Censorship Detection at Scale with Machine
Learning | To appear in Proceedings of the 29th ACM SIGKDD Conference on
Knowledge Discovery and Data Mining (KDD '23) | null | null | null | cs.LG cs.AI cs.CY cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The proliferation of global censorship has led to the development of a
plethora of measurement platforms to monitor and expose it. Censorship of the
domain name system (DNS) is a key mechanism used across different countries. It
is currently detected by applying heuristics to samples of DNS queries and
responses (probes) for specific destinations. These heuristics, however, are
both platform-specific and have been found to be brittle when censors change
their blocking behavior, necessitating a more reliable automated process for
detecting censorship.
In this paper, we explore how machine learning (ML) models can (1) help
streamline the detection process, (2) improve the potential of using
large-scale datasets for censorship detection, and (3) discover new censorship
instances and blocking signatures missed by existing heuristic methods. Our
study shows that supervised models, trained using expert-derived labels on
instances of known anomalies and possible censorship, can learn the detection
heuristics employed by different measurement platforms. More crucially, we find
that unsupervised models, trained solely on uncensored instances, can identify
new instances and variations of censorship missed by existing heuristics.
Moreover, both methods demonstrate the capability to uncover a substantial
number of new DNS blocking signatures, i.e., injected fake IP addresses
overlooked by existing heuristics. These results are underpinned by an
important methodological finding: comparing the outputs of models trained using
the same probes but with labels arising from independent processes allows us to
more reliably detect cases of censorship in the absence of ground-truth labels
of censorship.
| [
{
"version": "v1",
"created": "Fri, 3 Feb 2023 23:36:30 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 20:52:14 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Brown",
"Jacob",
""
],
[
"Jiang",
"Xi",
""
],
[
"Tran",
"Van",
""
],
[
"Bhagoji",
"Arjun Nitin",
""
],
[
"Hoang",
"Nguyen Phong",
""
],
[
"Feamster",
"Nick",
""
],
[
"Mittal",
"Prateek",
""
],
[
"Yegneswaran",
"Vinod",
""
]
] | TITLE: Augmenting Rule-based DNS Censorship Detection at Scale with Machine
Learning
ABSTRACT: The proliferation of global censorship has led to the development of a
plethora of measurement platforms to monitor and expose it. Censorship of the
domain name system (DNS) is a key mechanism used across different countries. It
is currently detected by applying heuristics to samples of DNS queries and
responses (probes) for specific destinations. These heuristics, however, are
both platform-specific and have been found to be brittle when censors change
their blocking behavior, necessitating a more reliable automated process for
detecting censorship.
In this paper, we explore how machine learning (ML) models can (1) help
streamline the detection process, (2) improve the potential of using
large-scale datasets for censorship detection, and (3) discover new censorship
instances and blocking signatures missed by existing heuristic methods. Our
study shows that supervised models, trained using expert-derived labels on
instances of known anomalies and possible censorship, can learn the detection
heuristics employed by different measurement platforms. More crucially, we find
that unsupervised models, trained solely on uncensored instances, can identify
new instances and variations of censorship missed by existing heuristics.
Moreover, both methods demonstrate the capability to uncover a substantial
number of new DNS blocking signatures, i.e., injected fake IP addresses
overlooked by existing heuristics. These results are underpinned by an
important methodological finding: comparing the outputs of models trained using
the same probes but with labels arising from independent processes allows us to
more reliably detect cases of censorship in the absence of ground-truth labels
of censorship.
|
2303.00332 | Hui Wang | Hui Wang, Siqi Zheng, Yafeng Chen, Luyao Cheng, Qian Chen | CAM++: A Fast and Efficient Network for Speaker Verification Using
Context-Aware Masking | null | null | null | null | cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time delay neural network (TDNN) has been proven to be efficient for speaker
verification. One of its successful variants, ECAPA-TDNN, achieved
state-of-the-art performance at the cost of much higher computational
complexity and slower inference speed. This makes it inadequate for scenarios
with demanding inference rate and limited computational resources. We are thus
interested in finding an architecture that can achieve the performance of
ECAPA-TDNN and the efficiency of vanilla TDNN. In this paper, we propose an
efficient network based on context-aware masking, namely CAM++, which uses
densely connected time delay neural network (D-TDNN) as backbone and adopts a
novel multi-granularity pooling to capture contextual information at different
levels. Extensive experiments on two public benchmarks, VoxCeleb and CN-Celeb,
demonstrate that the proposed architecture outperforms other mainstream speaker
verification systems with lower computational cost and faster inference speed.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 08:50:31 GMT"
},
{
"version": "v2",
"created": "Thu, 2 Mar 2023 03:17:41 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Jun 2023 09:07:11 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Wang",
"Hui",
""
],
[
"Zheng",
"Siqi",
""
],
[
"Chen",
"Yafeng",
""
],
[
"Cheng",
"Luyao",
""
],
[
"Chen",
"Qian",
""
]
] | TITLE: CAM++: A Fast and Efficient Network for Speaker Verification Using
Context-Aware Masking
ABSTRACT: Time delay neural network (TDNN) has been proven to be efficient for speaker
verification. One of its successful variants, ECAPA-TDNN, achieved
state-of-the-art performance at the cost of much higher computational
complexity and slower inference speed. This makes it inadequate for scenarios
with demanding inference rate and limited computational resources. We are thus
interested in finding an architecture that can achieve the performance of
ECAPA-TDNN and the efficiency of vanilla TDNN. In this paper, we propose an
efficient network based on context-aware masking, namely CAM++, which uses
densely connected time delay neural network (D-TDNN) as backbone and adopts a
novel multi-granularity pooling to capture contextual information at different
levels. Extensive experiments on two public benchmarks, VoxCeleb and CN-Celeb,
demonstrate that the proposed architecture outperforms other mainstream speaker
verification systems with lower computational cost and faster inference speed.
|
2303.03363 | Philipp Seidl | Philipp Seidl, Andreu Vall, Sepp Hochreiter, G\"unter Klambauer | Enhancing Activity Prediction Models in Drug Discovery with the Ability
to Understand Human Language | ICML version, 15 pages + 18 pages appendix | null | null | null | q-bio.BM cs.CL cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Activity and property prediction models are the central workhorses in drug
discovery and materials sciences, but currently they have to be trained or
fine-tuned for new tasks. Without training or fine-tuning, scientific language
models could be used for such low-data tasks through their announced zero- and
few-shot capabilities. However, their predictive quality at activity prediction
is lacking. In this work, we envision a novel type of activity prediction model
that is able to adapt to new prediction tasks at inference time, via
understanding textual information describing the task. To this end, we propose
a new architecture with separate modules for chemical and natural language
inputs, and a contrastive pre-training objective on data from large biochemical
databases. In extensive experiments, we show that our method CLAMP yields
improved predictive performance on few-shot learning benchmarks and zero-shot
problems in drug discovery. We attribute the advances of our method to the
modularized architecture and to our pre-training objective.
| [
{
"version": "v1",
"created": "Mon, 6 Mar 2023 18:49:09 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 09:59:34 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Seidl",
"Philipp",
""
],
[
"Vall",
"Andreu",
""
],
[
"Hochreiter",
"Sepp",
""
],
[
"Klambauer",
"Günter",
""
]
] | TITLE: Enhancing Activity Prediction Models in Drug Discovery with the Ability
to Understand Human Language
ABSTRACT: Activity and property prediction models are the central workhorses in drug
discovery and materials sciences, but currently they have to be trained or
fine-tuned for new tasks. Without training or fine-tuning, scientific language
models could be used for such low-data tasks through their announced zero- and
few-shot capabilities. However, their predictive quality at activity prediction
is lacking. In this work, we envision a novel type of activity prediction model
that is able to adapt to new prediction tasks at inference time, via
understanding textual information describing the task. To this end, we propose
a new architecture with separate modules for chemical and natural language
inputs, and a contrastive pre-training objective on data from large biochemical
databases. In extensive experiments, we show that our method CLAMP yields
improved predictive performance on few-shot learning benchmarks and zero-shot
problems in drug discovery. We attribute the advances of our method to the
modularized architecture and to our pre-training objective.
|
2303.18080 | Yasser Benigmim | Yasser Benigmim, Subhankar Roy, Slim Essid, Vicky Kalogeiton,
St\'ephane Lathuili\`ere | One-shot Unsupervised Domain Adaptation with Personalized Diffusion
Models | Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern
Recognition- Workshop on Generative Models for Computer Vision (CVPR-W 2023) | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Adapting a segmentation model from a labeled source domain to a target
domain, where a single unlabeled datum is available, is one the most
challenging problems in domain adaptation and is otherwise known as one-shot
unsupervised domain adaptation (OSUDA). Most of the prior works have addressed
the problem by relying on style transfer techniques, where the source images
are stylized to have the appearance of the target domain. Departing from the
common notion of transferring only the target ``texture'' information, we
leverage text-to-image diffusion models (e.g., Stable Diffusion) to generate a
synthetic target dataset with photo-realistic images that not only faithfully
depict the style of the target domain, but are also characterized by novel
scenes in diverse contexts. The text interface in our method Data AugmenTation
with diffUsion Models (DATUM) endows us with the possibility of guiding the
generation of images towards desired semantic concepts while respecting the
original spatial context of a single training image, which is not possible in
existing OSUDA methods. Extensive experiments on standard benchmarks show that
our DATUM surpasses the state-of-the-art OSUDA methods by up to +7.1%. The
implementation is available at https://github.com/yasserben/DATUM
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 14:16:38 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 16:27:52 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Benigmim",
"Yasser",
""
],
[
"Roy",
"Subhankar",
""
],
[
"Essid",
"Slim",
""
],
[
"Kalogeiton",
"Vicky",
""
],
[
"Lathuilière",
"Stéphane",
""
]
] | TITLE: One-shot Unsupervised Domain Adaptation with Personalized Diffusion
Models
ABSTRACT: Adapting a segmentation model from a labeled source domain to a target
domain, where a single unlabeled datum is available, is one the most
challenging problems in domain adaptation and is otherwise known as one-shot
unsupervised domain adaptation (OSUDA). Most of the prior works have addressed
the problem by relying on style transfer techniques, where the source images
are stylized to have the appearance of the target domain. Departing from the
common notion of transferring only the target ``texture'' information, we
leverage text-to-image diffusion models (e.g., Stable Diffusion) to generate a
synthetic target dataset with photo-realistic images that not only faithfully
depict the style of the target domain, but are also characterized by novel
scenes in diverse contexts. The text interface in our method Data AugmenTation
with diffUsion Models (DATUM) endows us with the possibility of guiding the
generation of images towards desired semantic concepts while respecting the
original spatial context of a single training image, which is not possible in
existing OSUDA methods. Extensive experiments on standard benchmarks show that
our DATUM surpasses the state-of-the-art OSUDA methods by up to +7.1%. The
implementation is available at https://github.com/yasserben/DATUM
|
2304.03378 | Giorgos Kordopatis-Zilos | Giorgos Kordopatis-Zilos and Giorgos Tolias and Christos Tzelepis and
Ioannis Kompatsiaris and Ioannis Patras and Symeon Papadopoulos | Self-Supervised Video Similarity Learning | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce S$^2$VS, a video similarity learning approach with
self-supervision. Self-Supervised Learning (SSL) is typically used to train
deep models on a proxy task so as to have strong transferability on target
tasks after fine-tuning. Here, in contrast to prior work, SSL is used to
perform video similarity learning and address multiple retrieval and detection
tasks at once with no use of labeled data. This is achieved by learning via
instance-discrimination with task-tailored augmentations and the widely used
InfoNCE loss together with an additional loss operating jointly on
self-similarity and hard-negative similarity. We benchmark our method on tasks
where video relevance is defined with varying granularity, ranging from video
copies to videos depicting the same incident or event. We learn a single
universal model that achieves state-of-the-art performance on all tasks,
surpassing previously proposed methods that use labeled data. The code and
pretrained models are publicly available at: https://github.com/gkordo/s2vs
| [
{
"version": "v1",
"created": "Thu, 6 Apr 2023 21:15:27 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 14:11:58 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Kordopatis-Zilos",
"Giorgos",
""
],
[
"Tolias",
"Giorgos",
""
],
[
"Tzelepis",
"Christos",
""
],
[
"Kompatsiaris",
"Ioannis",
""
],
[
"Patras",
"Ioannis",
""
],
[
"Papadopoulos",
"Symeon",
""
]
] | TITLE: Self-Supervised Video Similarity Learning
ABSTRACT: We introduce S$^2$VS, a video similarity learning approach with
self-supervision. Self-Supervised Learning (SSL) is typically used to train
deep models on a proxy task so as to have strong transferability on target
tasks after fine-tuning. Here, in contrast to prior work, SSL is used to
perform video similarity learning and address multiple retrieval and detection
tasks at once with no use of labeled data. This is achieved by learning via
instance-discrimination with task-tailored augmentations and the widely used
InfoNCE loss together with an additional loss operating jointly on
self-similarity and hard-negative similarity. We benchmark our method on tasks
where video relevance is defined with varying granularity, ranging from video
copies to videos depicting the same incident or event. We learn a single
universal model that achieves state-of-the-art performance on all tasks,
surpassing previously proposed methods that use labeled data. The code and
pretrained models are publicly available at: https://github.com/gkordo/s2vs
|
2304.07142 | William Ravenscroft | William Ravenscroft and Stefan Goetze and Thomas Hain | On Data Sampling Strategies for Training Neural Network Speech
Separation Models | Accepted for EUSIPCO 2023 | null | null | null | cs.SD cs.AI cs.LG cs.NE eess.AS | http://creativecommons.org/licenses/by/4.0/ | Speech separation remains an important area of multi-speaker signal
processing. Deep neural network (DNN) models have attained the best performance
on many speech separation benchmarks. Some of these models can take significant
time to train and have high memory requirements. Previous work has proposed
shortening training examples to address these issues but the impact of this on
model performance is not yet well understood. In this work, the impact of
applying these training signal length (TSL) limits is analysed for two speech
separation models: SepFormer, a transformer model, and Conv-TasNet, a
convolutional model. The WJS0-2Mix, WHAMR and Libri2Mix datasets are analysed
in terms of signal length distribution and its impact on training efficiency.
It is demonstrated that, for specific distributions, applying specific TSL
limits results in better performance. This is shown to be mainly due to
randomly sampling the start index of the waveforms resulting in more unique
examples for training. A SepFormer model trained using a TSL limit of 4.42s and
dynamic mixing (DM) is shown to match the best-performing SepFormer model
trained with DM and unlimited signal lengths. Furthermore, the 4.42s TSL limit
results in a 44% reduction in training time with WHAMR.
| [
{
"version": "v1",
"created": "Fri, 14 Apr 2023 14:05:52 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 13:42:10 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Ravenscroft",
"William",
""
],
[
"Goetze",
"Stefan",
""
],
[
"Hain",
"Thomas",
""
]
] | TITLE: On Data Sampling Strategies for Training Neural Network Speech
Separation Models
ABSTRACT: Speech separation remains an important area of multi-speaker signal
processing. Deep neural network (DNN) models have attained the best performance
on many speech separation benchmarks. Some of these models can take significant
time to train and have high memory requirements. Previous work has proposed
shortening training examples to address these issues but the impact of this on
model performance is not yet well understood. In this work, the impact of
applying these training signal length (TSL) limits is analysed for two speech
separation models: SepFormer, a transformer model, and Conv-TasNet, a
convolutional model. The WJS0-2Mix, WHAMR and Libri2Mix datasets are analysed
in terms of signal length distribution and its impact on training efficiency.
It is demonstrated that, for specific distributions, applying specific TSL
limits results in better performance. This is shown to be mainly due to
randomly sampling the start index of the waveforms resulting in more unique
examples for training. A SepFormer model trained using a TSL limit of 4.42s and
dynamic mixing (DM) is shown to match the best-performing SepFormer model
trained with DM and unlimited signal lengths. Furthermore, the 4.42s TSL limit
results in a 44% reduction in training time with WHAMR.
|
2305.06940 | Ning Ding | Ning Ding, Ce Zhang, Azim Eskandarian | SalienDet: A Saliency-based Feature Enhancement Algorithm for Object
Detection for Autonomous Driving | This paper is accepted and being published at IEEE Transactions on
Intelligent Vehicles | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detection (OD) is crucial to autonomous driving. On the other hand,
unknown objects, which have not been seen in training sample set, are one of
the reasons that hinder autonomous vehicles from driving beyond the operational
domain. To addresss this issue, we propose a saliency-based OD algorithm
(SalienDet) to detect unknown objects. Our SalienDet utilizes a saliency-based
algorithm to enhance image features for object proposal generation. Moreover,
we design a dataset relabeling approach to differentiate the unknown objects
from all objects in training sample set to achieve Open-World Detection. To
validate the performance of SalienDet, we evaluate SalienDet on KITTI,
nuScenes, and BDD datasets, and the result indicates that it outperforms
existing algorithms for unknown object detection. Notably, SalienDet can be
easily adapted for incremental learning in open-world detection tasks. The
project page is
\url{https://github.com/dingmike001/SalienDet-Open-Detection.git}.
| [
{
"version": "v1",
"created": "Thu, 11 May 2023 16:19:44 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 05:28:33 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Ding",
"Ning",
""
],
[
"Zhang",
"Ce",
""
],
[
"Eskandarian",
"Azim",
""
]
] | TITLE: SalienDet: A Saliency-based Feature Enhancement Algorithm for Object
Detection for Autonomous Driving
ABSTRACT: Object detection (OD) is crucial to autonomous driving. On the other hand,
unknown objects, which have not been seen in training sample set, are one of
the reasons that hinder autonomous vehicles from driving beyond the operational
domain. To addresss this issue, we propose a saliency-based OD algorithm
(SalienDet) to detect unknown objects. Our SalienDet utilizes a saliency-based
algorithm to enhance image features for object proposal generation. Moreover,
we design a dataset relabeling approach to differentiate the unknown objects
from all objects in training sample set to achieve Open-World Detection. To
validate the performance of SalienDet, we evaluate SalienDet on KITTI,
nuScenes, and BDD datasets, and the result indicates that it outperforms
existing algorithms for unknown object detection. Notably, SalienDet can be
easily adapted for incremental learning in open-world detection tasks. The
project page is
\url{https://github.com/dingmike001/SalienDet-Open-Detection.git}.
|
2305.14065 | Peng Xu | Peng Xu, Lin Zhang, Xuanzhou Liu, Jiaqi Sun, Yue Zhao, Haiqin Yang,
Bei Yu | Do Not Train It: A Linear Neural Architecture Search of Graph Neural
Networks | null | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural architecture search (NAS) for Graph neural networks (GNNs), called
NAS-GNNs, has achieved significant performance over manually designed GNN
architectures. However, these methods inherit issues from the conventional NAS
methods, such as high computational cost and optimization difficulty. More
importantly, previous NAS methods have ignored the uniqueness of GNNs, where
GNNs possess expressive power without training. With the randomly-initialized
weights, we can then seek the optimal architecture parameters via the sparse
coding objective and derive a novel NAS-GNNs method, namely neural architecture
coding (NAC). Consequently, our NAC holds a no-update scheme on GNNs and can
efficiently compute in linear time. Empirical evaluations on multiple GNN
benchmark datasets demonstrate that our approach leads to state-of-the-art
performance, which is up to $200\times$ faster and $18.8\%$ more accurate than
the strong baselines.
| [
{
"version": "v1",
"created": "Tue, 23 May 2023 13:44:04 GMT"
},
{
"version": "v2",
"created": "Thu, 15 Jun 2023 07:23:46 GMT"
},
{
"version": "v3",
"created": "Fri, 16 Jun 2023 10:33:21 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Xu",
"Peng",
""
],
[
"Zhang",
"Lin",
""
],
[
"Liu",
"Xuanzhou",
""
],
[
"Sun",
"Jiaqi",
""
],
[
"Zhao",
"Yue",
""
],
[
"Yang",
"Haiqin",
""
],
[
"Yu",
"Bei",
""
]
] | TITLE: Do Not Train It: A Linear Neural Architecture Search of Graph Neural
Networks
ABSTRACT: Neural architecture search (NAS) for Graph neural networks (GNNs), called
NAS-GNNs, has achieved significant performance over manually designed GNN
architectures. However, these methods inherit issues from the conventional NAS
methods, such as high computational cost and optimization difficulty. More
importantly, previous NAS methods have ignored the uniqueness of GNNs, where
GNNs possess expressive power without training. With the randomly-initialized
weights, we can then seek the optimal architecture parameters via the sparse
coding objective and derive a novel NAS-GNNs method, namely neural architecture
coding (NAC). Consequently, our NAC holds a no-update scheme on GNNs and can
efficiently compute in linear time. Empirical evaluations on multiple GNN
benchmark datasets demonstrate that our approach leads to state-of-the-art
performance, which is up to $200\times$ faster and $18.8\%$ more accurate than
the strong baselines.
|
2305.16391 | Xiaohui Chen | Xiaohui Chen, Jiankai Sun, Taiqing Wang, Ruocheng Guo, Li-Ping Liu,
Aonan Zhang | Graph-Based Model-Agnostic Data Subsampling for Recommendation Systems | KDD 2023, fix typo | null | 10.1145/3580305.3599834 | null | cs.IR cs.LG cs.SI | http://creativecommons.org/licenses/by/4.0/ | Data subsampling is widely used to speed up the training of large-scale
recommendation systems. Most subsampling methods are model-based and often
require a pre-trained pilot model to measure data importance via e.g. sample
hardness. However, when the pilot model is misspecified, model-based
subsampling methods deteriorate. Since model misspecification is persistent in
real recommendation systems, we instead propose model-agnostic data subsampling
methods by only exploring input data structure represented by graphs.
Specifically, we study the topology of the user-item graph to estimate the
importance of each user-item interaction (an edge in the user-item graph) via
graph conductance, followed by a propagation step on the network to smooth out
the estimated importance value.
Since our proposed method is model-agnostic, we can marry the merits of both
model-agnostic and model-based subsampling methods. Empirically, we show that
combing the two consistently improves over any single method on the used
datasets.
Experimental results on KuaiRec and MIND datasets demonstrate that our
proposed methods achieve superior results compared to baseline approaches.
| [
{
"version": "v1",
"created": "Thu, 25 May 2023 18:00:15 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 05:38:39 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Chen",
"Xiaohui",
""
],
[
"Sun",
"Jiankai",
""
],
[
"Wang",
"Taiqing",
""
],
[
"Guo",
"Ruocheng",
""
],
[
"Liu",
"Li-Ping",
""
],
[
"Zhang",
"Aonan",
""
]
] | TITLE: Graph-Based Model-Agnostic Data Subsampling for Recommendation Systems
ABSTRACT: Data subsampling is widely used to speed up the training of large-scale
recommendation systems. Most subsampling methods are model-based and often
require a pre-trained pilot model to measure data importance via e.g. sample
hardness. However, when the pilot model is misspecified, model-based
subsampling methods deteriorate. Since model misspecification is persistent in
real recommendation systems, we instead propose model-agnostic data subsampling
methods by only exploring input data structure represented by graphs.
Specifically, we study the topology of the user-item graph to estimate the
importance of each user-item interaction (an edge in the user-item graph) via
graph conductance, followed by a propagation step on the network to smooth out
the estimated importance value.
Since our proposed method is model-agnostic, we can marry the merits of both
model-agnostic and model-based subsampling methods. Empirically, we show that
combing the two consistently improves over any single method on the used
datasets.
Experimental results on KuaiRec and MIND datasets demonstrate that our
proposed methods achieve superior results compared to baseline approaches.
|
2306.04370 | Yuting Zhang | Yuting Zhang, Yiqing Wu, Ran Le, Yongchun Zhu, Fuzhen Zhuang, Ruidong
Han, Xiang Li, Wei Lin, Zhulin An, Yongjun Xu | Modeling Dual Period-Varying Preferences for Takeaway Recommendation | accepted by KDD (The 29th ACM SIGKDD Conference on Knowledge
Discovery and Data Mining) 2023 Applied Data Science (ADS) track | null | 10.1145/3580305.3599866 | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Takeaway recommender systems, which aim to accurately provide stores that
offer foods meeting users' interests, have served billions of users in our
daily life. Different from traditional recommendation, takeaway recommendation
faces two main challenges: (1) Dual Interaction-Aware Preference Modeling.
Traditional recommendation commonly focuses on users' single preferences for
items while takeaway recommendation needs to comprehensively consider users'
dual preferences for stores and foods. (2) Period-Varying Preference Modeling.
Conventional recommendation generally models continuous changes in users'
preferences from a session-level or day-level perspective. However, in
practical takeaway systems, users' preferences vary significantly during the
morning, noon, night, and late night periods of the day. To address these
challenges, we propose a Dual Period-Varying Preference modeling (DPVP) for
takeaway recommendation. Specifically, we design a dual interaction-aware
module, aiming to capture users' dual preferences based on their interactions
with stores and foods. Moreover, to model various preferences in different time
periods of the day, we propose a time-based decomposition module as well as a
time-aware gating mechanism. Extensive offline and online experiments
demonstrate that our model outperforms state-of-the-art methods on real-world
datasets and it is capable of modeling the dual period-varying preferences.
Moreover, our model has been deployed online on Meituan Takeaway platform,
leading to an average improvement in GMV (Gross Merchandise Value) of 0.70%.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 12:06:10 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 07:29:28 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Zhang",
"Yuting",
""
],
[
"Wu",
"Yiqing",
""
],
[
"Le",
"Ran",
""
],
[
"Zhu",
"Yongchun",
""
],
[
"Zhuang",
"Fuzhen",
""
],
[
"Han",
"Ruidong",
""
],
[
"Li",
"Xiang",
""
],
[
"Lin",
"Wei",
""
],
[
"An",
"Zhulin",
""
],
[
"Xu",
"Yongjun",
""
]
] | TITLE: Modeling Dual Period-Varying Preferences for Takeaway Recommendation
ABSTRACT: Takeaway recommender systems, which aim to accurately provide stores that
offer foods meeting users' interests, have served billions of users in our
daily life. Different from traditional recommendation, takeaway recommendation
faces two main challenges: (1) Dual Interaction-Aware Preference Modeling.
Traditional recommendation commonly focuses on users' single preferences for
items while takeaway recommendation needs to comprehensively consider users'
dual preferences for stores and foods. (2) Period-Varying Preference Modeling.
Conventional recommendation generally models continuous changes in users'
preferences from a session-level or day-level perspective. However, in
practical takeaway systems, users' preferences vary significantly during the
morning, noon, night, and late night periods of the day. To address these
challenges, we propose a Dual Period-Varying Preference modeling (DPVP) for
takeaway recommendation. Specifically, we design a dual interaction-aware
module, aiming to capture users' dual preferences based on their interactions
with stores and foods. Moreover, to model various preferences in different time
periods of the day, we propose a time-based decomposition module as well as a
time-aware gating mechanism. Extensive offline and online experiments
demonstrate that our model outperforms state-of-the-art methods on real-world
datasets and it is capable of modeling the dual period-varying preferences.
Moreover, our model has been deployed online on Meituan Takeaway platform,
leading to an average improvement in GMV (Gross Merchandise Value) of 0.70%.
|
2306.04668 | Lisa Y.W. Tang | Lisa Y.W. Tang | SMRVIS: Point cloud extraction from 3-D ultrasound for non-destructive
testing | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | We propose to formulate point cloud extraction from ultrasound volumes as an
image segmentation problem. Through this convenient formulation, a quick
prototype exploring various variants of the Residual Network, U-Net, and the
Squeeze and Excitation Network was developed and evaluated. This report
documents the experimental results compiled using a training dataset of five
labeled ultrasound volumes and 84 unlabeled volumes that got completed in a
two-week period as part of a submission to the open challenge "3D Surface Mesh
Estimation for CVPR workshop on Deep Learning in Ultrasound Image Analysis".
Based on external evaluation performed by the challenge's organizers, the
framework came first place on the challenge's
\href{https://www.cvpr2023-dl-ultrasound.com/}{Leaderboard}. Source code is
shared with the research community at a
\href{https://github.com/lisatwyw/smrvis}{public repository}.
| [
{
"version": "v1",
"created": "Wed, 7 Jun 2023 14:53:02 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 00:41:44 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Tang",
"Lisa Y. W.",
""
]
] | TITLE: SMRVIS: Point cloud extraction from 3-D ultrasound for non-destructive
testing
ABSTRACT: We propose to formulate point cloud extraction from ultrasound volumes as an
image segmentation problem. Through this convenient formulation, a quick
prototype exploring various variants of the Residual Network, U-Net, and the
Squeeze and Excitation Network was developed and evaluated. This report
documents the experimental results compiled using a training dataset of five
labeled ultrasound volumes and 84 unlabeled volumes that got completed in a
two-week period as part of a submission to the open challenge "3D Surface Mesh
Estimation for CVPR workshop on Deep Learning in Ultrasound Image Analysis".
Based on external evaluation performed by the challenge's organizers, the
framework came first place on the challenge's
\href{https://www.cvpr2023-dl-ultrasound.com/}{Leaderboard}. Source code is
shared with the research community at a
\href{https://github.com/lisatwyw/smrvis}{public repository}.
|
2306.07279 | Tiange Luo | Tiange Luo, Chris Rockwell, Honglak Lee, Justin Johnson | Scalable 3D Captioning with Pretrained Models | Dataset link: https://huggingface.co/datasets/tiange/Cap3D | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce Cap3D, an automatic approach for generating descriptive text for
3D objects. This approach utilizes pretrained models from image captioning,
image-text alignment, and LLM to consolidate captions from multiple views of a
3D asset, completely side-stepping the time-consuming and costly process of
manual annotation. We apply Cap3D to the recently introduced large-scale 3D
dataset, Objaverse, resulting in 660k 3D-text pairs. Our evaluation, conducted
using 41k human annotations from the same dataset, demonstrates that Cap3D
surpasses human-authored descriptions in terms of quality, cost, and speed.
Through effective prompt engineering, Cap3D rivals human performance in
generating geometric descriptions on 17k collected annotations from the ABO
dataset. Finally, we finetune Text-to-3D models on Cap3D and human captions,
and show Cap3D outperforms; and benchmark the SOTA including Point-E, Shape-E,
and DreamFusion.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 17:59:03 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 03:58:15 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Luo",
"Tiange",
""
],
[
"Rockwell",
"Chris",
""
],
[
"Lee",
"Honglak",
""
],
[
"Johnson",
"Justin",
""
]
] | TITLE: Scalable 3D Captioning with Pretrained Models
ABSTRACT: We introduce Cap3D, an automatic approach for generating descriptive text for
3D objects. This approach utilizes pretrained models from image captioning,
image-text alignment, and LLM to consolidate captions from multiple views of a
3D asset, completely side-stepping the time-consuming and costly process of
manual annotation. We apply Cap3D to the recently introduced large-scale 3D
dataset, Objaverse, resulting in 660k 3D-text pairs. Our evaluation, conducted
using 41k human annotations from the same dataset, demonstrates that Cap3D
surpasses human-authored descriptions in terms of quality, cost, and speed.
Through effective prompt engineering, Cap3D rivals human performance in
generating geometric descriptions on 17k collected annotations from the ABO
dataset. Finally, we finetune Text-to-3D models on Cap3D and human captions,
and show Cap3D outperforms; and benchmark the SOTA including Point-E, Shape-E,
and DreamFusion.
|
2306.08183 | Kelly Marshall | Kelly O. Marshall, Minh Pham, Ameya Joshi, Anushrut Jignasu, Aditya
Balu, Adarsh Krishnamurthy, Chinmay Hegde | ZeroForge: Feedforward Text-to-Shape Without 3D Supervision | 19 pages, High resolution figures needed to demonstrate 3D results | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Current state-of-the-art methods for text-to-shape generation either require
supervised training using a labeled dataset of pre-defined 3D shapes, or
perform expensive inference-time optimization of implicit neural
representations. In this work, we present ZeroForge, an approach for zero-shot
text-to-shape generation that avoids both pitfalls. To achieve open-vocabulary
shape generation, we require careful architectural adaptation of existing
feed-forward approaches, as well as a combination of data-free CLIP-loss and
contrastive losses to avoid mode collapse. Using these techniques, we are able
to considerably expand the generative ability of existing feed-forward
text-to-shape models such as CLIP-Forge. We support our method via extensive
qualitative and quantitative evaluations
| [
{
"version": "v1",
"created": "Wed, 14 Jun 2023 00:38:14 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 00:48:13 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Marshall",
"Kelly O.",
""
],
[
"Pham",
"Minh",
""
],
[
"Joshi",
"Ameya",
""
],
[
"Jignasu",
"Anushrut",
""
],
[
"Balu",
"Aditya",
""
],
[
"Krishnamurthy",
"Adarsh",
""
],
[
"Hegde",
"Chinmay",
""
]
] | TITLE: ZeroForge: Feedforward Text-to-Shape Without 3D Supervision
ABSTRACT: Current state-of-the-art methods for text-to-shape generation either require
supervised training using a labeled dataset of pre-defined 3D shapes, or
perform expensive inference-time optimization of implicit neural
representations. In this work, we present ZeroForge, an approach for zero-shot
text-to-shape generation that avoids both pitfalls. To achieve open-vocabulary
shape generation, we require careful architectural adaptation of existing
feed-forward approaches, as well as a combination of data-free CLIP-loss and
contrastive losses to avoid mode collapse. Using these techniques, we are able
to considerably expand the generative ability of existing feed-forward
text-to-shape models such as CLIP-Forge. We support our method via extensive
qualitative and quantitative evaluations
|
2306.08830 | Xiao Jin | Xiao Jin, Xin-Yue Mu, Jing Xu | Searching for the Fakes: Efficient Neural Architecture Search for
General Face Forgery Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | As the saying goes, "seeing is believing". However, with the development of
digital face editing tools, we can no longer trust what we can see. Although
face forgery detection has made promising progress, most current methods are
designed manually by human experts, which is labor-consuming. In this paper, we
develop an end-to-end framework based on neural architecture search (NAS) for
deepfake detection, which can automatically design network architectures
without human intervention. First, a forgery-oriented search space is created
to choose appropriate operations for this task. Second, we propose a novel
performance estimation metric, which guides the search process to select more
general models. The cross-dataset search is also considered to develop more
general architectures. Eventually, we connect the cells in a cascaded pyramid
way for final forgery classification. Compared with state-of-the-art networks
artificially designed, our method achieves competitive performance in both
in-dataset and cross-dataset scenarios.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 03:01:13 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 05:27:51 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Jin",
"Xiao",
""
],
[
"Mu",
"Xin-Yue",
""
],
[
"Xu",
"Jing",
""
]
] | TITLE: Searching for the Fakes: Efficient Neural Architecture Search for
General Face Forgery Detection
ABSTRACT: As the saying goes, "seeing is believing". However, with the development of
digital face editing tools, we can no longer trust what we can see. Although
face forgery detection has made promising progress, most current methods are
designed manually by human experts, which is labor-consuming. In this paper, we
develop an end-to-end framework based on neural architecture search (NAS) for
deepfake detection, which can automatically design network architectures
without human intervention. First, a forgery-oriented search space is created
to choose appropriate operations for this task. Second, we propose a novel
performance estimation metric, which guides the search process to select more
general models. The cross-dataset search is also considered to develop more
general architectures. Eventually, we connect the cells in a cascaded pyramid
way for final forgery classification. Compared with state-of-the-art networks
artificially designed, our method achieves competitive performance in both
in-dataset and cross-dataset scenarios.
|
2306.09303 | Zuxin Liu | Zuxin Liu, Zijian Guo, Haohong Lin, Yihang Yao, Jiacheng Zhu, Zhepeng
Cen, Hanjiang Hu, Wenhao Yu, Tingnan Zhang, Jie Tan, Ding Zhao | Datasets and Benchmarks for Offline Safe Reinforcement Learning | 22 pages.13 figures, 7 tables | null | null | null | cs.LG cs.AI cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a comprehensive benchmarking suite tailored to offline
safe reinforcement learning (RL) challenges, aiming to foster progress in the
development and evaluation of safe learning algorithms in both the training and
deployment phases. Our benchmark suite contains three packages: 1) expertly
crafted safe policies, 2) D4RL-styled datasets along with environment wrappers,
and 3) high-quality offline safe RL baseline implementations. We feature a
methodical data collection pipeline powered by advanced safe RL algorithms,
which facilitates the generation of diverse datasets across 38 popular safe RL
tasks, from robot control to autonomous driving. We further introduce an array
of data post-processing filters, capable of modifying each dataset's diversity,
thereby simulating various data collection conditions. Additionally, we provide
elegant and extensible implementations of prevalent offline safe RL algorithms
to accelerate research in this area. Through extensive experiments with over
50000 CPU and 800 GPU hours of computations, we evaluate and compare the
performance of these baseline algorithms on the collected datasets, offering
insights into their strengths, limitations, and potential areas of improvement.
Our benchmarking framework serves as a valuable resource for researchers and
practitioners, facilitating the development of more robust and reliable offline
safe RL solutions in safety-critical applications. The benchmark website is
available at \url{www.offline-saferl.org}.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 17:31:26 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 17:54:06 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Liu",
"Zuxin",
""
],
[
"Guo",
"Zijian",
""
],
[
"Lin",
"Haohong",
""
],
[
"Yao",
"Yihang",
""
],
[
"Zhu",
"Jiacheng",
""
],
[
"Cen",
"Zhepeng",
""
],
[
"Hu",
"Hanjiang",
""
],
[
"Yu",
"Wenhao",
""
],
[
"Zhang",
"Tingnan",
""
],
[
"Tan",
"Jie",
""
],
[
"Zhao",
"Ding",
""
]
] | TITLE: Datasets and Benchmarks for Offline Safe Reinforcement Learning
ABSTRACT: This paper presents a comprehensive benchmarking suite tailored to offline
safe reinforcement learning (RL) challenges, aiming to foster progress in the
development and evaluation of safe learning algorithms in both the training and
deployment phases. Our benchmark suite contains three packages: 1) expertly
crafted safe policies, 2) D4RL-styled datasets along with environment wrappers,
and 3) high-quality offline safe RL baseline implementations. We feature a
methodical data collection pipeline powered by advanced safe RL algorithms,
which facilitates the generation of diverse datasets across 38 popular safe RL
tasks, from robot control to autonomous driving. We further introduce an array
of data post-processing filters, capable of modifying each dataset's diversity,
thereby simulating various data collection conditions. Additionally, we provide
elegant and extensible implementations of prevalent offline safe RL algorithms
to accelerate research in this area. Through extensive experiments with over
50000 CPU and 800 GPU hours of computations, we evaluate and compare the
performance of these baseline algorithms on the collected datasets, offering
insights into their strengths, limitations, and potential areas of improvement.
Our benchmarking framework serves as a valuable resource for researchers and
practitioners, facilitating the development of more robust and reliable offline
safe RL solutions in safety-critical applications. The benchmark website is
available at \url{www.offline-saferl.org}.
|
2306.09368 | Jiawen Zhang | Jiawen Zhang, Shun Zheng, Wei Cao, Jiang Bian, Jia Li | Warpformer: A Multi-scale Modeling Approach for Irregular Clinical Time
Series | KDD23 Research Track | null | 10.1145/3580305.3599543 | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Irregularly sampled multivariate time series are ubiquitous in various
fields, particularly in healthcare, and exhibit two key characteristics:
intra-series irregularity and inter-series discrepancy. Intra-series
irregularity refers to the fact that time-series signals are often recorded at
irregular intervals, while inter-series discrepancy refers to the significant
variability in sampling rates among diverse series. However, recent advances in
irregular time series have primarily focused on addressing intra-series
irregularity, overlooking the issue of inter-series discrepancy. To bridge this
gap, we present Warpformer, a novel approach that fully considers these two
characteristics. In a nutshell, Warpformer has several crucial designs,
including a specific input representation that explicitly characterizes both
intra-series irregularity and inter-series discrepancy, a warping module that
adaptively unifies irregular time series in a given scale, and a customized
attention module for representation learning. Additionally, we stack multiple
warping and attention modules to learn at different scales, producing
multi-scale representations that balance coarse-grained and fine-grained
signals for downstream tasks. We conduct extensive experiments on widely used
datasets and a new large-scale benchmark built from clinical databases. The
results demonstrate the superiority of Warpformer over existing
state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Wed, 14 Jun 2023 13:23:14 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Zhang",
"Jiawen",
""
],
[
"Zheng",
"Shun",
""
],
[
"Cao",
"Wei",
""
],
[
"Bian",
"Jiang",
""
],
[
"Li",
"Jia",
""
]
] | TITLE: Warpformer: A Multi-scale Modeling Approach for Irregular Clinical Time
Series
ABSTRACT: Irregularly sampled multivariate time series are ubiquitous in various
fields, particularly in healthcare, and exhibit two key characteristics:
intra-series irregularity and inter-series discrepancy. Intra-series
irregularity refers to the fact that time-series signals are often recorded at
irregular intervals, while inter-series discrepancy refers to the significant
variability in sampling rates among diverse series. However, recent advances in
irregular time series have primarily focused on addressing intra-series
irregularity, overlooking the issue of inter-series discrepancy. To bridge this
gap, we present Warpformer, a novel approach that fully considers these two
characteristics. In a nutshell, Warpformer has several crucial designs,
including a specific input representation that explicitly characterizes both
intra-series irregularity and inter-series discrepancy, a warping module that
adaptively unifies irregular time series in a given scale, and a customized
attention module for representation learning. Additionally, we stack multiple
warping and attention modules to learn at different scales, producing
multi-scale representations that balance coarse-grained and fine-grained
signals for downstream tasks. We conduct extensive experiments on widely used
datasets and a new large-scale benchmark built from clinical databases. The
results demonstrate the superiority of Warpformer over existing
state-of-the-art approaches.
|
2306.09372 | Mijanur Palash | Mijanur Palash, Bharat Bhargava | SAFER: Situation Aware Facial Emotion Recognition | Facial Expression Recognition, Emotion Recognition, Deep Learning,
Covid-19, Contextual information, Open-world AI | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this paper, we present SAFER, a novel system for emotion recognition from
facial expressions. It employs state-of-the-art deep learning techniques to
extract various features from facial images and incorporates contextual
information, such as background and location type, to enhance its performance.
The system has been designed to operate in an open-world setting, meaning it
can adapt to unseen and varied facial expressions, making it suitable for
real-world applications. An extensive evaluation of SAFER against existing
works in the field demonstrates improved performance, achieving an accuracy of
91.4% on the CAER-S dataset. Additionally, the study investigates the effect of
novelty such as face masks during the Covid-19 pandemic on facial emotion
recognition and critically examines the limitations of mainstream facial
expressions datasets. To address these limitations, a novel dataset for facial
emotion recognition is proposed. The proposed dataset and the system are
expected to be useful for various applications such as human-computer
interaction, security, and surveillance.
| [
{
"version": "v1",
"created": "Wed, 14 Jun 2023 20:42:26 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Palash",
"Mijanur",
""
],
[
"Bhargava",
"Bharat",
""
]
] | TITLE: SAFER: Situation Aware Facial Emotion Recognition
ABSTRACT: In this paper, we present SAFER, a novel system for emotion recognition from
facial expressions. It employs state-of-the-art deep learning techniques to
extract various features from facial images and incorporates contextual
information, such as background and location type, to enhance its performance.
The system has been designed to operate in an open-world setting, meaning it
can adapt to unseen and varied facial expressions, making it suitable for
real-world applications. An extensive evaluation of SAFER against existing
works in the field demonstrates improved performance, achieving an accuracy of
91.4% on the CAER-S dataset. Additionally, the study investigates the effect of
novelty such as face masks during the Covid-19 pandemic on facial emotion
recognition and critically examines the limitations of mainstream facial
expressions datasets. To address these limitations, a novel dataset for facial
emotion recognition is proposed. The proposed dataset and the system are
expected to be useful for various applications such as human-computer
interaction, security, and surveillance.
|
2306.09375 | Shengchao Liu | Shengchao Liu, Weitao Du, Yanjing Li, Zhuoxinran Li, Zhiling Zheng,
Chenru Duan, Zhiming Ma, Omar Yaghi, Anima Anandkumar, Christian Borgs,
Jennifer Chayes, Hongyu Guo, Jian Tang | Symmetry-Informed Geometric Representation for Molecules, Proteins, and
Crystalline Materials | null | null | null | null | cs.LG physics.chem-ph q-bio.QM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Artificial intelligence for scientific discovery has recently generated
significant interest within the machine learning and scientific communities,
particularly in the domains of chemistry, biology, and material discovery. For
these scientific problems, molecules serve as the fundamental building blocks,
and machine learning has emerged as a highly effective and powerful tool for
modeling their geometric structures. Nevertheless, due to the rapidly evolving
process of the field and the knowledge gap between science (e.g., physics,
chemistry, & biology) and machine learning communities, a benchmarking study on
geometrical representation for such data has not been conducted. To address
such an issue, in this paper, we first provide a unified view of the current
symmetry-informed geometric methods, classifying them into three main
categories: invariance, equivariance with spherical frame basis, and
equivariance with vector frame basis. Then we propose a platform, coined
Geom3D, which enables benchmarking the effectiveness of geometric strategies.
Geom3D contains 16 advanced symmetry-informed geometric representation models
and 14 geometric pretraining methods over 46 diverse datasets, including small
molecules, proteins, and crystalline materials. We hope that Geom3D can, on the
one hand, eliminate barriers for machine learning researchers interested in
exploring scientific problems; and, on the other hand, provide valuable
guidance for researchers in computational chemistry, structural biology, and
materials science, aiding in the informed selection of representation
techniques for specific applications.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 05:37:25 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Liu",
"Shengchao",
""
],
[
"Du",
"Weitao",
""
],
[
"Li",
"Yanjing",
""
],
[
"Li",
"Zhuoxinran",
""
],
[
"Zheng",
"Zhiling",
""
],
[
"Duan",
"Chenru",
""
],
[
"Ma",
"Zhiming",
""
],
[
"Yaghi",
"Omar",
""
],
[
"Anandkumar",
"Anima",
""
],
[
"Borgs",
"Christian",
""
],
[
"Chayes",
"Jennifer",
""
],
[
"Guo",
"Hongyu",
""
],
[
"Tang",
"Jian",
""
]
] | TITLE: Symmetry-Informed Geometric Representation for Molecules, Proteins, and
Crystalline Materials
ABSTRACT: Artificial intelligence for scientific discovery has recently generated
significant interest within the machine learning and scientific communities,
particularly in the domains of chemistry, biology, and material discovery. For
these scientific problems, molecules serve as the fundamental building blocks,
and machine learning has emerged as a highly effective and powerful tool for
modeling their geometric structures. Nevertheless, due to the rapidly evolving
process of the field and the knowledge gap between science (e.g., physics,
chemistry, & biology) and machine learning communities, a benchmarking study on
geometrical representation for such data has not been conducted. To address
such an issue, in this paper, we first provide a unified view of the current
symmetry-informed geometric methods, classifying them into three main
categories: invariance, equivariance with spherical frame basis, and
equivariance with vector frame basis. Then we propose a platform, coined
Geom3D, which enables benchmarking the effectiveness of geometric strategies.
Geom3D contains 16 advanced symmetry-informed geometric representation models
and 14 geometric pretraining methods over 46 diverse datasets, including small
molecules, proteins, and crystalline materials. We hope that Geom3D can, on the
one hand, eliminate barriers for machine learning researchers interested in
exploring scientific problems; and, on the other hand, provide valuable
guidance for researchers in computational chemistry, structural biology, and
materials science, aiding in the informed selection of representation
techniques for specific applications.
|
2306.09379 | Shengqi Xu | Shengqi Xu, Shuning Cao, Haoyue Liu, Xueyao Xiao, Yi Chang, Luxin Yan | 1st Solution Places for CVPR 2023 UG$^2$+ Challenge Track 2.2-Coded
Target Restoration through Atmospheric Turbulence | arXiv admin note: text overlap with arXiv:2306.08963 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-sa/4.0/ | In this technical report, we briefly introduce the solution of our team
VIELab-HUST for coded target restoration through atmospheric turbulence in CVPR
2023 UG$^2$+ Track 2.2. In this task, we propose an efficient multi-stage
framework to restore a high quality image from distorted frames. Specifically,
each distorted frame is initially aligned using image registration to suppress
geometric distortion. We subsequently select the sharpest set of registered
frames by employing a frame selection approach based on image sharpness, and
average them to produce an image that is largely free of geometric distortion,
albeit with blurriness. A learning-based deblurring method is then applied to
remove the residual blur in the averaged image. Finally, post-processing
techniques are utilized to further enhance the quality of the output image. Our
framework is capable of handling different kinds of coded target dataset
provided in the final testing phase, and ranked 1st on the final leaderboard.
Our code will be available at https://github.com/xsqhust/Turbulence_Removal.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 09:06:48 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Xu",
"Shengqi",
""
],
[
"Cao",
"Shuning",
""
],
[
"Liu",
"Haoyue",
""
],
[
"Xiao",
"Xueyao",
""
],
[
"Chang",
"Yi",
""
],
[
"Yan",
"Luxin",
""
]
] | TITLE: 1st Solution Places for CVPR 2023 UG$^2$+ Challenge Track 2.2-Coded
Target Restoration through Atmospheric Turbulence
ABSTRACT: In this technical report, we briefly introduce the solution of our team
VIELab-HUST for coded target restoration through atmospheric turbulence in CVPR
2023 UG$^2$+ Track 2.2. In this task, we propose an efficient multi-stage
framework to restore a high quality image from distorted frames. Specifically,
each distorted frame is initially aligned using image registration to suppress
geometric distortion. We subsequently select the sharpest set of registered
frames by employing a frame selection approach based on image sharpness, and
average them to produce an image that is largely free of geometric distortion,
albeit with blurriness. A learning-based deblurring method is then applied to
remove the residual blur in the averaged image. Finally, post-processing
techniques are utilized to further enhance the quality of the output image. Our
framework is capable of handling different kinds of coded target dataset
provided in the final testing phase, and ranked 1st on the final leaderboard.
Our code will be available at https://github.com/xsqhust/Turbulence_Removal.
|
2306.09386 | Ziyue Li Dr | Yirong Chen, Ziyue Li, Wanli Ouyang, Michael Lepech | Adaptive Hierarchical SpatioTemporal Network for Traffic Forecasting | Accepted to IEEE CASE 2023 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurate traffic forecasting is vital to intelligent transportation systems,
which are widely adopted to solve urban traffic issues. Existing traffic
forecasting studies focus on modeling spatial-temporal dynamics in traffic
data, among which the graph convolution network (GCN) is at the center for
exploiting the spatial dependency embedded in the road network graphs. However,
these GCN-based methods operate intrinsically on the node level (e.g., road and
intersection) only whereas overlooking the spatial hierarchy of the whole city.
Nodes such as intersections and road segments can form clusters (e.g.,
regions), which could also have interactions with each other and share
similarities at a higher level. In this work, we propose an Adaptive
Hierarchical SpatioTemporal Network (AHSTN) to promote traffic forecasting by
exploiting the spatial hierarchy and modeling multi-scale spatial correlations.
Apart from the node-level spatiotemporal blocks, AHSTN introduces the adaptive
spatiotemporal downsampling module to infer the spatial hierarchy for
spatiotemporal modeling at the cluster level. Then, an adaptive spatiotemporal
upsampling module is proposed to upsample the cluster-level representations to
the node-level and obtain the multi-scale representations for generating
predictions. Experiments on two real-world datasets show that AHSTN achieves
better performance over several strong baselines.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 14:50:27 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Chen",
"Yirong",
""
],
[
"Li",
"Ziyue",
""
],
[
"Ouyang",
"Wanli",
""
],
[
"Lepech",
"Michael",
""
]
] | TITLE: Adaptive Hierarchical SpatioTemporal Network for Traffic Forecasting
ABSTRACT: Accurate traffic forecasting is vital to intelligent transportation systems,
which are widely adopted to solve urban traffic issues. Existing traffic
forecasting studies focus on modeling spatial-temporal dynamics in traffic
data, among which the graph convolution network (GCN) is at the center for
exploiting the spatial dependency embedded in the road network graphs. However,
these GCN-based methods operate intrinsically on the node level (e.g., road and
intersection) only whereas overlooking the spatial hierarchy of the whole city.
Nodes such as intersections and road segments can form clusters (e.g.,
regions), which could also have interactions with each other and share
similarities at a higher level. In this work, we propose an Adaptive
Hierarchical SpatioTemporal Network (AHSTN) to promote traffic forecasting by
exploiting the spatial hierarchy and modeling multi-scale spatial correlations.
Apart from the node-level spatiotemporal blocks, AHSTN introduces the adaptive
spatiotemporal downsampling module to infer the spatial hierarchy for
spatiotemporal modeling at the cluster level. Then, an adaptive spatiotemporal
upsampling module is proposed to upsample the cluster-level representations to
the node-level and obtain the multi-scale representations for generating
predictions. Experiments on two real-world datasets show that AHSTN achieves
better performance over several strong baselines.
|
2306.09390 | Hamideh Ghanadian | Hamideh Ghanadian, Isar Nejadgholi, Hussein Al Osman | ChatGPT for Suicide Risk Assessment on Social Media: Quantitative
Evaluation of Model Performance, Potentials and Limitations | null | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper presents a novel framework for quantitatively evaluating the
interactive ChatGPT model in the context of suicidality assessment from social
media posts, utilizing the University of Maryland Reddit suicidality dataset.
We conduct a technical evaluation of ChatGPT's performance on this task using
Zero-Shot and Few-Shot experiments and compare its results with those of two
fine-tuned transformer-based models. Additionally, we investigate the impact of
different temperature parameters on ChatGPT's response generation and discuss
the optimal temperature based on the inconclusiveness rate of ChatGPT. Our
results indicate that while ChatGPT attains considerable accuracy in this task,
transformer-based models fine-tuned on human-annotated datasets exhibit
superior performance. Moreover, our analysis sheds light on how adjusting the
ChatGPT's hyperparameters can improve its ability to assist mental health
professionals in this critical task.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 16:01:30 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Ghanadian",
"Hamideh",
""
],
[
"Nejadgholi",
"Isar",
""
],
[
"Osman",
"Hussein Al",
""
]
] | TITLE: ChatGPT for Suicide Risk Assessment on Social Media: Quantitative
Evaluation of Model Performance, Potentials and Limitations
ABSTRACT: This paper presents a novel framework for quantitatively evaluating the
interactive ChatGPT model in the context of suicidality assessment from social
media posts, utilizing the University of Maryland Reddit suicidality dataset.
We conduct a technical evaluation of ChatGPT's performance on this task using
Zero-Shot and Few-Shot experiments and compare its results with those of two
fine-tuned transformer-based models. Additionally, we investigate the impact of
different temperature parameters on ChatGPT's response generation and discuss
the optimal temperature based on the inconclusiveness rate of ChatGPT. Our
results indicate that while ChatGPT attains considerable accuracy in this task,
transformer-based models fine-tuned on human-annotated datasets exhibit
superior performance. Moreover, our analysis sheds light on how adjusting the
ChatGPT's hyperparameters can improve its ability to assist mental health
professionals in this critical task.
|
2306.09426 | Valdivino Alexandre De Santiago J\'unior | Valdivino Alexandre de Santiago J\'unior | Deep learning techniques for blind image super-resolution: A high-scale
multi-domain perspective evaluation | 21 pages | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Despite several solutions and experiments have been conducted recently
addressing image super-resolution (SR), boosted by deep learning (DL)
techniques, they do not usually design evaluations with high scaling factors,
capping it at 2x or 4x. Moreover, the datasets are generally benchmarks which
do not truly encompass significant diversity of domains to proper evaluate the
techniques. It is also interesting to remark that blind SR is attractive for
real-world scenarios since it is based on the idea that the degradation process
is unknown, and hence techniques in this context rely basically on
low-resolution (LR) images. In this article, we present a high-scale (8x)
controlled experiment which evaluates five recent DL techniques tailored for
blind image SR: Adaptive Pseudo Augmentation (APA), Blind Image SR with
Spatially Variant Degradations (BlindSR), Deep Alternating Network (DAN),
FastGAN, and Mixture of Experts Super-Resolution (MoESR). We consider 14 small
datasets from five different broader domains which are: aerial, fauna, flora,
medical, and satellite. Another distinctive characteristic of our evaluation is
that some of the DL approaches were designed for single-image SR but others
not. Two no-reference metrics were selected, being the classical natural image
quality evaluator (NIQE) and the recent transformer-based multi-dimension
attention network for no-reference image quality assessment (MANIQA) score, to
assess the techniques. Overall, MoESR can be regarded as the best solution
although the perceptual quality of the created HR images of all the techniques
still needs to improve. Supporting code: https://github.com/vsantjr/DL_BlindSR.
Datasets:
https://www.kaggle.com/datasets/valdivinosantiago/dl-blindsr-datasets.
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 18:18:02 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Júnior",
"Valdivino Alexandre de Santiago",
""
]
] | TITLE: Deep learning techniques for blind image super-resolution: A high-scale
multi-domain perspective evaluation
ABSTRACT: Despite several solutions and experiments have been conducted recently
addressing image super-resolution (SR), boosted by deep learning (DL)
techniques, they do not usually design evaluations with high scaling factors,
capping it at 2x or 4x. Moreover, the datasets are generally benchmarks which
do not truly encompass significant diversity of domains to proper evaluate the
techniques. It is also interesting to remark that blind SR is attractive for
real-world scenarios since it is based on the idea that the degradation process
is unknown, and hence techniques in this context rely basically on
low-resolution (LR) images. In this article, we present a high-scale (8x)
controlled experiment which evaluates five recent DL techniques tailored for
blind image SR: Adaptive Pseudo Augmentation (APA), Blind Image SR with
Spatially Variant Degradations (BlindSR), Deep Alternating Network (DAN),
FastGAN, and Mixture of Experts Super-Resolution (MoESR). We consider 14 small
datasets from five different broader domains which are: aerial, fauna, flora,
medical, and satellite. Another distinctive characteristic of our evaluation is
that some of the DL approaches were designed for single-image SR but others
not. Two no-reference metrics were selected, being the classical natural image
quality evaluator (NIQE) and the recent transformer-based multi-dimension
attention network for no-reference image quality assessment (MANIQA) score, to
assess the techniques. Overall, MoESR can be regarded as the best solution
although the perceptual quality of the created HR images of all the techniques
still needs to improve. Supporting code: https://github.com/vsantjr/DL_BlindSR.
Datasets:
https://www.kaggle.com/datasets/valdivinosantiago/dl-blindsr-datasets.
|
2306.09431 | Wenxuan Hou | Wenxuan Hou, Guangyao Li, Yapeng Tian, Di Hu | Towards Long Form Audio-visual Video Understanding | null | null | null | null | cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We live in a world filled with never-ending streams of multimodal
information. As a more natural recording of the real scenario, long form
audio-visual videos are expected as an important bridge for better exploring
and understanding the world. In this paper, we propose the multisensory
temporal event localization task in long form videos and strive to tackle the
associated challenges. To facilitate this study, we first collect a large-scale
Long Form Audio-visual Video (LFAV) dataset with 5,175 videos and an average
video length of 210 seconds. Each of the collected videos is elaborately
annotated with diversified modality-aware events, in a long-range temporal
sequence. We then propose an event-centric framework for localizing
multisensory events as well as understanding their relations in long form
videos. It includes three phases in different levels: snippet prediction phase
to learn snippet features, event extraction phase to extract event-level
features, and event interaction phase to study event relations. Experiments
demonstrate that the proposed method, utilizing the new LFAV dataset, exhibits
considerable effectiveness in localizing multiple modality-aware events within
long form videos. Project website: http://gewu-lab.github.io/LFAV/
| [
{
"version": "v1",
"created": "Thu, 15 Jun 2023 18:22:32 GMT"
}
] | 2023-06-19T00:00:00 | [
[
"Hou",
"Wenxuan",
""
],
[
"Li",
"Guangyao",
""
],
[
"Tian",
"Yapeng",
""
],
[
"Hu",
"Di",
""
]
] | TITLE: Towards Long Form Audio-visual Video Understanding
ABSTRACT: We live in a world filled with never-ending streams of multimodal
information. As a more natural recording of the real scenario, long form
audio-visual videos are expected as an important bridge for better exploring
and understanding the world. In this paper, we propose the multisensory
temporal event localization task in long form videos and strive to tackle the
associated challenges. To facilitate this study, we first collect a large-scale
Long Form Audio-visual Video (LFAV) dataset with 5,175 videos and an average
video length of 210 seconds. Each of the collected videos is elaborately
annotated with diversified modality-aware events, in a long-range temporal
sequence. We then propose an event-centric framework for localizing
multisensory events as well as understanding their relations in long form
videos. It includes three phases in different levels: snippet prediction phase
to learn snippet features, event extraction phase to extract event-level
features, and event interaction phase to study event relations. Experiments
demonstrate that the proposed method, utilizing the new LFAV dataset, exhibits
considerable effectiveness in localizing multiple modality-aware events within
long form videos. Project website: http://gewu-lab.github.io/LFAV/
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.