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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2212.14258 | Sungyeon Kim | Sungyeon Kim, Boseung Jeong, Suha Kwak | HIER: Metric Learning Beyond Class Labels via Hierarchical
Regularization | Accepted to CVPR 2023 | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Supervision for metric learning has long been given in the form of
equivalence between human-labeled classes. Although this type of supervision
has been a basis of metric learning for decades, we argue that it hinders
further advances in the field. In this regard, we propose a new regularization
method, dubbed HIER, to discover the latent semantic hierarchy of training
data, and to deploy the hierarchy to provide richer and more fine-grained
supervision than inter-class separability induced by common metric learning
losses.HIER achieves this goal with no annotation for the semantic hierarchy
but by learning hierarchical proxies in hyperbolic spaces. The hierarchical
proxies are learnable parameters, and each of them is trained to serve as an
ancestor of a group of data or other proxies to approximate the semantic
hierarchy among them. HIER deals with the proxies along with data in hyperbolic
space since the geometric properties of the space are well-suited to represent
their hierarchical structure. The efficacy of HIER is evaluated on four
standard benchmarks, where it consistently improved the performance of
conventional methods when integrated with them, and consequently achieved the
best records, surpassing even the existing hyperbolic metric learning
technique, in almost all settings.
| [
{
"version": "v1",
"created": "Thu, 29 Dec 2022 11:05:47 GMT"
},
{
"version": "v2",
"created": "Sun, 22 Jan 2023 16:14:32 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Apr 2023 07:48:39 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Kim",
"Sungyeon",
""
],
[
"Jeong",
"Boseung",
""
],
[
"Kwak",
"Suha",
""
]
] | TITLE: HIER: Metric Learning Beyond Class Labels via Hierarchical
Regularization
ABSTRACT: Supervision for metric learning has long been given in the form of
equivalence between human-labeled classes. Although this type of supervision
has been a basis of metric learning for decades, we argue that it hinders
further advances in the field. In this regard, we propose a new regularization
method, dubbed HIER, to discover the latent semantic hierarchy of training
data, and to deploy the hierarchy to provide richer and more fine-grained
supervision than inter-class separability induced by common metric learning
losses.HIER achieves this goal with no annotation for the semantic hierarchy
but by learning hierarchical proxies in hyperbolic spaces. The hierarchical
proxies are learnable parameters, and each of them is trained to serve as an
ancestor of a group of data or other proxies to approximate the semantic
hierarchy among them. HIER deals with the proxies along with data in hyperbolic
space since the geometric properties of the space are well-suited to represent
their hierarchical structure. The efficacy of HIER is evaluated on four
standard benchmarks, where it consistently improved the performance of
conventional methods when integrated with them, and consequently achieved the
best records, surpassing even the existing hyperbolic metric learning
technique, in almost all settings.
|
2212.14613 | Yanbiao Ma | Yanbiao Ma, Licheng Jiao, Fang Liu, Yuxin Li, Shuyuan Yang, Xu Liu | Delving into Semantic Scale Imbalance | 47 pages, 26 figures, 12 tables, Published as a conference paper at
ICLR 2023 | null | null | null | cs.CV cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Model bias triggered by long-tailed data has been widely studied. However,
measure based on the number of samples cannot explicate three phenomena
simultaneously: (1) Given enough data, the classification performance gain is
marginal with additional samples. (2) Classification performance decays
precipitously as the number of training samples decreases when there is
insufficient data. (3) Model trained on sample-balanced datasets still has
different biases for different classes. In this work, we define and quantify
the semantic scale of classes, which is used to measure the feature diversity
of classes. It is exciting to find experimentally that there is a marginal
effect of semantic scale, which perfectly describes the first two phenomena.
Further, the quantitative measurement of semantic scale imbalance is proposed,
which can accurately reflect model bias on multiple datasets, even on
sample-balanced data, revealing a novel perspective for the study of class
imbalance. Due to the prevalence of semantic scale imbalance, we propose
semantic-scale-balanced learning, including a general loss improvement scheme
and a dynamic re-weighting training framework that overcomes the challenge of
calculating semantic scales in real-time during iterations. Comprehensive
experiments show that dynamic semantic-scale-balanced learning consistently
enables the model to perform superiorly on large-scale long-tailed and
non-long-tailed natural and medical datasets, which is a good starting point
for mitigating the prevalent but unnoticed model bias.
| [
{
"version": "v1",
"created": "Fri, 30 Dec 2022 09:40:09 GMT"
},
{
"version": "v2",
"created": "Mon, 2 Jan 2023 10:18:08 GMT"
},
{
"version": "v3",
"created": "Thu, 5 Jan 2023 06:04:01 GMT"
},
{
"version": "v4",
"created": "Mon, 9 Jan 2023 15:26:16 GMT"
},
{
"version": "v5",
"created": "Tue, 10 Jan 2023 09:20:44 GMT"
},
{
"version": "v6",
"created": "Mon, 16 Jan 2023 14:43:28 GMT"
},
{
"version": "v7",
"created": "Sun, 22 Jan 2023 13:12:48 GMT"
},
{
"version": "v8",
"created": "Sat, 8 Apr 2023 11:35:37 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Ma",
"Yanbiao",
""
],
[
"Jiao",
"Licheng",
""
],
[
"Liu",
"Fang",
""
],
[
"Li",
"Yuxin",
""
],
[
"Yang",
"Shuyuan",
""
],
[
"Liu",
"Xu",
""
]
] | TITLE: Delving into Semantic Scale Imbalance
ABSTRACT: Model bias triggered by long-tailed data has been widely studied. However,
measure based on the number of samples cannot explicate three phenomena
simultaneously: (1) Given enough data, the classification performance gain is
marginal with additional samples. (2) Classification performance decays
precipitously as the number of training samples decreases when there is
insufficient data. (3) Model trained on sample-balanced datasets still has
different biases for different classes. In this work, we define and quantify
the semantic scale of classes, which is used to measure the feature diversity
of classes. It is exciting to find experimentally that there is a marginal
effect of semantic scale, which perfectly describes the first two phenomena.
Further, the quantitative measurement of semantic scale imbalance is proposed,
which can accurately reflect model bias on multiple datasets, even on
sample-balanced data, revealing a novel perspective for the study of class
imbalance. Due to the prevalence of semantic scale imbalance, we propose
semantic-scale-balanced learning, including a general loss improvement scheme
and a dynamic re-weighting training framework that overcomes the challenge of
calculating semantic scales in real-time during iterations. Comprehensive
experiments show that dynamic semantic-scale-balanced learning consistently
enables the model to perform superiorly on large-scale long-tailed and
non-long-tailed natural and medical datasets, which is a good starting point
for mitigating the prevalent but unnoticed model bias.
|
2301.02560 | Vikram V. Ramaswamy | Vikram V. Ramaswamy, Sing Yu Lin, Dora Zhao, Aaron B. Adcock, Laurens
van der Maaten, Deepti Ghadiyaram, Olga Russakovsky | GeoDE: a Geographically Diverse Evaluation Dataset for Object
Recognition | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current dataset collection methods typically scrape large amounts of data
from the web. While this technique is extremely scalable, data collected in
this way tends to reinforce stereotypical biases, can contain personally
identifiable information, and typically originates from Europe and North
America. In this work, we rethink the dataset collection paradigm and introduce
GeoDE, a geographically diverse dataset with 61,940 images from 40 classes and
6 world regions, and no personally identifiable information, collected through
crowd-sourcing. We analyse GeoDE to understand differences in images collected
in this manner compared to web-scraping. Despite the smaller size of this
dataset, we demonstrate its use as both an evaluation and training dataset,
highlight shortcomings in current models, as well as show improved performances
when even small amounts of GeoDE (1000 - 2000 images per region) are added to a
training dataset. We release the full dataset and code at
https://geodiverse-data-collection.cs.princeton.edu/
| [
{
"version": "v1",
"created": "Thu, 5 Jan 2023 18:21:50 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Apr 2023 21:54:59 GMT"
},
{
"version": "v3",
"created": "Sat, 8 Apr 2023 00:10:46 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Ramaswamy",
"Vikram V.",
""
],
[
"Lin",
"Sing Yu",
""
],
[
"Zhao",
"Dora",
""
],
[
"Adcock",
"Aaron B.",
""
],
[
"van der Maaten",
"Laurens",
""
],
[
"Ghadiyaram",
"Deepti",
""
],
[
"Russakovsky",
"Olga",
""
]
] | TITLE: GeoDE: a Geographically Diverse Evaluation Dataset for Object
Recognition
ABSTRACT: Current dataset collection methods typically scrape large amounts of data
from the web. While this technique is extremely scalable, data collected in
this way tends to reinforce stereotypical biases, can contain personally
identifiable information, and typically originates from Europe and North
America. In this work, we rethink the dataset collection paradigm and introduce
GeoDE, a geographically diverse dataset with 61,940 images from 40 classes and
6 world regions, and no personally identifiable information, collected through
crowd-sourcing. We analyse GeoDE to understand differences in images collected
in this manner compared to web-scraping. Despite the smaller size of this
dataset, we demonstrate its use as both an evaluation and training dataset,
highlight shortcomings in current models, as well as show improved performances
when even small amounts of GeoDE (1000 - 2000 images per region) are added to a
training dataset. We release the full dataset and code at
https://geodiverse-data-collection.cs.princeton.edu/
|
2301.04224 | Xindi Wu | Xindi Wu, KwunFung Lau, Francesco Ferroni, Aljo\v{s}a O\v{s}ep, Deva
Ramanan | Pix2Map: Cross-modal Retrieval for Inferring Street Maps from Images | 12 pages, 8 figures | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-driving vehicles rely on urban street maps for autonomous navigation. In
this paper, we introduce Pix2Map, a method for inferring urban street map
topology directly from ego-view images, as needed to continually update and
expand existing maps. This is a challenging task, as we need to infer a complex
urban road topology directly from raw image data. The main insight of this
paper is that this problem can be posed as cross-modal retrieval by learning a
joint, cross-modal embedding space for images and existing maps, represented as
discrete graphs that encode the topological layout of the visual surroundings.
We conduct our experimental evaluation using the Argoverse dataset and show
that it is indeed possible to accurately retrieve street maps corresponding to
both seen and unseen roads solely from image data. Moreover, we show that our
retrieved maps can be used to update or expand existing maps and even show
proof-of-concept results for visual localization and image retrieval from
spatial graphs.
| [
{
"version": "v1",
"created": "Tue, 10 Jan 2023 22:05:35 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Apr 2023 21:30:05 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Wu",
"Xindi",
""
],
[
"Lau",
"KwunFung",
""
],
[
"Ferroni",
"Francesco",
""
],
[
"Ošep",
"Aljoša",
""
],
[
"Ramanan",
"Deva",
""
]
] | TITLE: Pix2Map: Cross-modal Retrieval for Inferring Street Maps from Images
ABSTRACT: Self-driving vehicles rely on urban street maps for autonomous navigation. In
this paper, we introduce Pix2Map, a method for inferring urban street map
topology directly from ego-view images, as needed to continually update and
expand existing maps. This is a challenging task, as we need to infer a complex
urban road topology directly from raw image data. The main insight of this
paper is that this problem can be posed as cross-modal retrieval by learning a
joint, cross-modal embedding space for images and existing maps, represented as
discrete graphs that encode the topological layout of the visual surroundings.
We conduct our experimental evaluation using the Argoverse dataset and show
that it is indeed possible to accurately retrieve street maps corresponding to
both seen and unseen roads solely from image data. Moreover, we show that our
retrieved maps can be used to update or expand existing maps and even show
proof-of-concept results for visual localization and image retrieval from
spatial graphs.
|
2301.05339 | Simbarashe Nyatsanga | Simbarashe Nyatsanga, Taras Kucherenko, Chaitanya Ahuja, Gustav Eje
Henter, Michael Neff | A Comprehensive Review of Data-Driven Co-Speech Gesture Generation | Accepted for EUROGRAPHICS 2023 | null | 10.1111/cgf.14776 | null | cs.GR cs.CV cs.HC cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Gestures that accompany speech are an essential part of natural and efficient
embodied human communication. The automatic generation of such co-speech
gestures is a long-standing problem in computer animation and is considered an
enabling technology in film, games, virtual social spaces, and for interaction
with social robots. The problem is made challenging by the idiosyncratic and
non-periodic nature of human co-speech gesture motion, and by the great
diversity of communicative functions that gestures encompass. Gesture
generation has seen surging interest recently, owing to the emergence of more
and larger datasets of human gesture motion, combined with strides in
deep-learning-based generative models, that benefit from the growing
availability of data. This review article summarizes co-speech gesture
generation research, with a particular focus on deep generative models. First,
we articulate the theory describing human gesticulation and how it complements
speech. Next, we briefly discuss rule-based and classical statistical gesture
synthesis, before delving into deep learning approaches. We employ the choice
of input modalities as an organizing principle, examining systems that generate
gestures from audio, text, and non-linguistic input. We also chronicle the
evolution of the related training data sets in terms of size, diversity, motion
quality, and collection method. Finally, we identify key research challenges in
gesture generation, including data availability and quality; producing
human-like motion; grounding the gesture in the co-occurring speech in
interaction with other speakers, and in the environment; performing gesture
evaluation; and integration of gesture synthesis into applications. We
highlight recent approaches to tackling the various key challenges, as well as
the limitations of these approaches, and point toward areas of future
development.
| [
{
"version": "v1",
"created": "Fri, 13 Jan 2023 00:20:05 GMT"
},
{
"version": "v2",
"created": "Tue, 21 Feb 2023 08:17:29 GMT"
},
{
"version": "v3",
"created": "Tue, 7 Mar 2023 19:35:43 GMT"
},
{
"version": "v4",
"created": "Mon, 10 Apr 2023 09:11:59 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Nyatsanga",
"Simbarashe",
""
],
[
"Kucherenko",
"Taras",
""
],
[
"Ahuja",
"Chaitanya",
""
],
[
"Henter",
"Gustav Eje",
""
],
[
"Neff",
"Michael",
""
]
] | TITLE: A Comprehensive Review of Data-Driven Co-Speech Gesture Generation
ABSTRACT: Gestures that accompany speech are an essential part of natural and efficient
embodied human communication. The automatic generation of such co-speech
gestures is a long-standing problem in computer animation and is considered an
enabling technology in film, games, virtual social spaces, and for interaction
with social robots. The problem is made challenging by the idiosyncratic and
non-periodic nature of human co-speech gesture motion, and by the great
diversity of communicative functions that gestures encompass. Gesture
generation has seen surging interest recently, owing to the emergence of more
and larger datasets of human gesture motion, combined with strides in
deep-learning-based generative models, that benefit from the growing
availability of data. This review article summarizes co-speech gesture
generation research, with a particular focus on deep generative models. First,
we articulate the theory describing human gesticulation and how it complements
speech. Next, we briefly discuss rule-based and classical statistical gesture
synthesis, before delving into deep learning approaches. We employ the choice
of input modalities as an organizing principle, examining systems that generate
gestures from audio, text, and non-linguistic input. We also chronicle the
evolution of the related training data sets in terms of size, diversity, motion
quality, and collection method. Finally, we identify key research challenges in
gesture generation, including data availability and quality; producing
human-like motion; grounding the gesture in the co-occurring speech in
interaction with other speakers, and in the environment; performing gesture
evaluation; and integration of gesture synthesis into applications. We
highlight recent approaches to tackling the various key challenges, as well as
the limitations of these approaches, and point toward areas of future
development.
|
2301.07850 | Shengyu Feng | Shengyu Feng, Hanghang Tong | Concept Discovery for Fast Adapatation | SDM23 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The advances in deep learning have enabled machine learning methods to
outperform human beings in various areas, but it remains a great challenge for
a well-trained model to quickly adapt to a new task. One promising solution to
realize this goal is through meta-learning, also known as learning to learn,
which has achieved promising results in few-shot learning. However, current
approaches are still enormously different from human beings' learning process,
especially in the ability to extract structural and transferable knowledge.
This drawback makes current meta-learning frameworks non-interpretable and hard
to extend to more complex tasks. We tackle this problem by introducing concept
discovery to the few-shot learning problem, where we achieve more effective
adaptation by meta-learning the structure among the data features, leading to a
composite representation of the data. Our proposed method Concept-Based
Model-Agnostic Meta-Learning (COMAML) has been shown to achieve consistent
improvements in the structured data for both synthesized datasets and
real-world datasets.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2023 02:33:58 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 00:32:17 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Feng",
"Shengyu",
""
],
[
"Tong",
"Hanghang",
""
]
] | TITLE: Concept Discovery for Fast Adapatation
ABSTRACT: The advances in deep learning have enabled machine learning methods to
outperform human beings in various areas, but it remains a great challenge for
a well-trained model to quickly adapt to a new task. One promising solution to
realize this goal is through meta-learning, also known as learning to learn,
which has achieved promising results in few-shot learning. However, current
approaches are still enormously different from human beings' learning process,
especially in the ability to extract structural and transferable knowledge.
This drawback makes current meta-learning frameworks non-interpretable and hard
to extend to more complex tasks. We tackle this problem by introducing concept
discovery to the few-shot learning problem, where we achieve more effective
adaptation by meta-learning the structure among the data features, leading to a
composite representation of the data. Our proposed method Concept-Based
Model-Agnostic Meta-Learning (COMAML) has been shown to achieve consistent
improvements in the structured data for both synthesized datasets and
real-world datasets.
|
2301.07944 | Jiazheng Xing | Jiazheng Xing, Mengmeng Wang, Yong Liu, Boyu Mu | Revisiting the Spatial and Temporal Modeling for Few-shot Action
Recognition | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Spatial and temporal modeling is one of the most core aspects of few-shot
action recognition. Most previous works mainly focus on long-term temporal
relation modeling based on high-level spatial representations, without
considering the crucial low-level spatial features and short-term temporal
relations. Actually, the former feature could bring rich local semantic
information, and the latter feature could represent motion characteristics of
adjacent frames, respectively. In this paper, we propose SloshNet, a new
framework that revisits the spatial and temporal modeling for few-shot action
recognition in a finer manner. First, to exploit the low-level spatial
features, we design a feature fusion architecture search module to
automatically search for the best combination of the low-level and high-level
spatial features. Next, inspired by the recent transformer, we introduce a
long-term temporal modeling module to model the global temporal relations based
on the extracted spatial appearance features. Meanwhile, we design another
short-term temporal modeling module to encode the motion characteristics
between adjacent frame representations. After that, the final predictions can
be obtained by feeding the embedded rich spatial-temporal features to a common
frame-level class prototype matcher. We extensively validate the proposed
SloshNet on four few-shot action recognition datasets, including
Something-Something V2, Kinetics, UCF101, and HMDB51. It achieves favorable
results against state-of-the-art methods in all datasets.
| [
{
"version": "v1",
"created": "Thu, 19 Jan 2023 08:34:04 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 03:29:05 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Xing",
"Jiazheng",
""
],
[
"Wang",
"Mengmeng",
""
],
[
"Liu",
"Yong",
""
],
[
"Mu",
"Boyu",
""
]
] | TITLE: Revisiting the Spatial and Temporal Modeling for Few-shot Action
Recognition
ABSTRACT: Spatial and temporal modeling is one of the most core aspects of few-shot
action recognition. Most previous works mainly focus on long-term temporal
relation modeling based on high-level spatial representations, without
considering the crucial low-level spatial features and short-term temporal
relations. Actually, the former feature could bring rich local semantic
information, and the latter feature could represent motion characteristics of
adjacent frames, respectively. In this paper, we propose SloshNet, a new
framework that revisits the spatial and temporal modeling for few-shot action
recognition in a finer manner. First, to exploit the low-level spatial
features, we design a feature fusion architecture search module to
automatically search for the best combination of the low-level and high-level
spatial features. Next, inspired by the recent transformer, we introduce a
long-term temporal modeling module to model the global temporal relations based
on the extracted spatial appearance features. Meanwhile, we design another
short-term temporal modeling module to encode the motion characteristics
between adjacent frame representations. After that, the final predictions can
be obtained by feeding the embedded rich spatial-temporal features to a common
frame-level class prototype matcher. We extensively validate the proposed
SloshNet on four few-shot action recognition datasets, including
Something-Something V2, Kinetics, UCF101, and HMDB51. It achieves favorable
results against state-of-the-art methods in all datasets.
|
2301.08968 | Huancheng Chen | Huancheng Chen, Johnny (Chaining) Wang, Haris Vikalo | The Best of Both Worlds: Accurate Global and Personalized Models through
Federated Learning with Data-Free Hyper-Knowledge Distillation | Accepted by ICLR2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Heterogeneity of data distributed across clients limits the performance of
global models trained through federated learning, especially in the settings
with highly imbalanced class distributions of local datasets. In recent years,
personalized federated learning (pFL) has emerged as a potential solution to
the challenges presented by heterogeneous data. However, existing pFL methods
typically enhance performance of local models at the expense of the global
model's accuracy. We propose FedHKD (Federated Hyper-Knowledge Distillation), a
novel FL algorithm in which clients rely on knowledge distillation (KD) to
train local models. In particular, each client extracts and sends to the server
the means of local data representations and the corresponding soft predictions
-- information that we refer to as ``hyper-knowledge". The server aggregates
this information and broadcasts it to the clients in support of local training.
Notably, unlike other KD-based pFL methods, FedHKD does not rely on a public
dataset nor it deploys a generative model at the server. We analyze convergence
of FedHKD and conduct extensive experiments on visual datasets in a variety of
scenarios, demonstrating that FedHKD provides significant improvement in both
personalized as well as global model performance compared to state-of-the-art
FL methods designed for heterogeneous data settings.
| [
{
"version": "v1",
"created": "Sat, 21 Jan 2023 16:20:57 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Apr 2023 15:03:20 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Chen",
"Huancheng",
"",
"Chaining"
],
[
"Johnny",
"",
"",
"Chaining"
],
[
"Wang",
"",
""
],
[
"Vikalo",
"Haris",
""
]
] | TITLE: The Best of Both Worlds: Accurate Global and Personalized Models through
Federated Learning with Data-Free Hyper-Knowledge Distillation
ABSTRACT: Heterogeneity of data distributed across clients limits the performance of
global models trained through federated learning, especially in the settings
with highly imbalanced class distributions of local datasets. In recent years,
personalized federated learning (pFL) has emerged as a potential solution to
the challenges presented by heterogeneous data. However, existing pFL methods
typically enhance performance of local models at the expense of the global
model's accuracy. We propose FedHKD (Federated Hyper-Knowledge Distillation), a
novel FL algorithm in which clients rely on knowledge distillation (KD) to
train local models. In particular, each client extracts and sends to the server
the means of local data representations and the corresponding soft predictions
-- information that we refer to as ``hyper-knowledge". The server aggregates
this information and broadcasts it to the clients in support of local training.
Notably, unlike other KD-based pFL methods, FedHKD does not rely on a public
dataset nor it deploys a generative model at the server. We analyze convergence
of FedHKD and conduct extensive experiments on visual datasets in a variety of
scenarios, demonstrating that FedHKD provides significant improvement in both
personalized as well as global model performance compared to state-of-the-art
FL methods designed for heterogeneous data settings.
|
2301.10051 | Zanjia Tong | Zanjia Tong, Yuhang Chen, Zewei Xu, Rong Yu | Wise-IoU: Bounding Box Regression Loss with Dynamic Focusing Mechanism | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The loss function for bounding box regression (BBR) is essential to object
detection. Its good definition will bring significant performance improvement
to the model. Most existing works assume that the examples in the training data
are high-quality and focus on strengthening the fitting ability of BBR loss. If
we blindly strengthen BBR on low-quality examples, it will jeopardize
localization performance. Focal-EIoU v1 was proposed to solve this problem, but
due to its static focusing mechanism (FM), the potential of non-monotonic FM
was not fully exploited. Based on this idea, we propose an IoU-based loss with
a dynamic non-monotonic FM named Wise-IoU (WIoU). The dynamic non-monotonic FM
uses the outlier degree instead of IoU to evaluate the quality of anchor boxes
and provides a wise gradient gain allocation strategy. This strategy reduces
the competitiveness of high-quality anchor boxes while also reducing the
harmful gradient generated by low-quality examples. This allows WIoU to focus
on ordinary-quality anchor boxes and improve the detector's overall
performance. When WIoU is applied to the state-of-the-art real-time detector
YOLOv7, the AP-75 on the MS-COCO dataset is improved from 53.03% to 54.50%.
Code is available at https://github.com/Instinct323/wiou.
| [
{
"version": "v1",
"created": "Tue, 24 Jan 2023 14:50:40 GMT"
},
{
"version": "v2",
"created": "Fri, 3 Feb 2023 14:05:02 GMT"
},
{
"version": "v3",
"created": "Sat, 8 Apr 2023 13:58:40 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Tong",
"Zanjia",
""
],
[
"Chen",
"Yuhang",
""
],
[
"Xu",
"Zewei",
""
],
[
"Yu",
"Rong",
""
]
] | TITLE: Wise-IoU: Bounding Box Regression Loss with Dynamic Focusing Mechanism
ABSTRACT: The loss function for bounding box regression (BBR) is essential to object
detection. Its good definition will bring significant performance improvement
to the model. Most existing works assume that the examples in the training data
are high-quality and focus on strengthening the fitting ability of BBR loss. If
we blindly strengthen BBR on low-quality examples, it will jeopardize
localization performance. Focal-EIoU v1 was proposed to solve this problem, but
due to its static focusing mechanism (FM), the potential of non-monotonic FM
was not fully exploited. Based on this idea, we propose an IoU-based loss with
a dynamic non-monotonic FM named Wise-IoU (WIoU). The dynamic non-monotonic FM
uses the outlier degree instead of IoU to evaluate the quality of anchor boxes
and provides a wise gradient gain allocation strategy. This strategy reduces
the competitiveness of high-quality anchor boxes while also reducing the
harmful gradient generated by low-quality examples. This allows WIoU to focus
on ordinary-quality anchor boxes and improve the detector's overall
performance. When WIoU is applied to the state-of-the-art real-time detector
YOLOv7, the AP-75 on the MS-COCO dataset is improved from 53.03% to 54.50%.
Code is available at https://github.com/Instinct323/wiou.
|
2302.03003 | Wenhui Zhu | Wenhui Zhu, Peijie Qiu, Oana M. Dumitrascu, Jacob M. Sobczak, Mohammad
Farazi, Zhangsihao Yang, Keshav Nandakumar, Yalin Wang | OTRE: Where Optimal Transport Guided Unpaired Image-to-Image Translation
Meets Regularization by Enhancing | Accepted as a conference paper to The 28th biennial international
conference on Information Processing in Medical Imaging (IPMI 2023) | null | null | null | eess.IV cs.CV stat.ML | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Non-mydriatic retinal color fundus photography (CFP) is widely available due
to the advantage of not requiring pupillary dilation, however, is prone to poor
quality due to operators, systemic imperfections, or patient-related causes.
Optimal retinal image quality is mandated for accurate medical diagnoses and
automated analyses. Herein, we leveraged the Optimal Transport (OT) theory to
propose an unpaired image-to-image translation scheme for mapping low-quality
retinal CFPs to high-quality counterparts. Furthermore, to improve the
flexibility, robustness, and applicability of our image enhancement pipeline in
the clinical practice, we generalized a state-of-the-art model-based image
reconstruction method, regularization by denoising, by plugging in priors
learned by our OT-guided image-to-image translation network. We named it as
regularization by enhancing (RE). We validated the integrated framework, OTRE,
on three publicly available retinal image datasets by assessing the quality
after enhancement and their performance on various downstream tasks, including
diabetic retinopathy grading, vessel segmentation, and diabetic lesion
segmentation. The experimental results demonstrated the superiority of our
proposed framework over some state-of-the-art unsupervised competitors and a
state-of-the-art supervised method.
| [
{
"version": "v1",
"created": "Mon, 6 Feb 2023 18:39:40 GMT"
},
{
"version": "v2",
"created": "Tue, 7 Feb 2023 16:31:32 GMT"
},
{
"version": "v3",
"created": "Thu, 9 Feb 2023 02:15:34 GMT"
},
{
"version": "v4",
"created": "Sun, 9 Apr 2023 00:56:15 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Zhu",
"Wenhui",
""
],
[
"Qiu",
"Peijie",
""
],
[
"Dumitrascu",
"Oana M.",
""
],
[
"Sobczak",
"Jacob M.",
""
],
[
"Farazi",
"Mohammad",
""
],
[
"Yang",
"Zhangsihao",
""
],
[
"Nandakumar",
"Keshav",
""
],
[
"Wang",
"Yalin",
""
]
] | TITLE: OTRE: Where Optimal Transport Guided Unpaired Image-to-Image Translation
Meets Regularization by Enhancing
ABSTRACT: Non-mydriatic retinal color fundus photography (CFP) is widely available due
to the advantage of not requiring pupillary dilation, however, is prone to poor
quality due to operators, systemic imperfections, or patient-related causes.
Optimal retinal image quality is mandated for accurate medical diagnoses and
automated analyses. Herein, we leveraged the Optimal Transport (OT) theory to
propose an unpaired image-to-image translation scheme for mapping low-quality
retinal CFPs to high-quality counterparts. Furthermore, to improve the
flexibility, robustness, and applicability of our image enhancement pipeline in
the clinical practice, we generalized a state-of-the-art model-based image
reconstruction method, regularization by denoising, by plugging in priors
learned by our OT-guided image-to-image translation network. We named it as
regularization by enhancing (RE). We validated the integrated framework, OTRE,
on three publicly available retinal image datasets by assessing the quality
after enhancement and their performance on various downstream tasks, including
diabetic retinopathy grading, vessel segmentation, and diabetic lesion
segmentation. The experimental results demonstrated the superiority of our
proposed framework over some state-of-the-art unsupervised competitors and a
state-of-the-art supervised method.
|
2302.10126 | Radu Tudor Ionescu | Eduard Poesina, Radu Tudor Ionescu, Josiane Mothe | iQPP: A Benchmark for Image Query Performance Prediction | Accepted at SIGIR 2023 | null | null | null | cs.CV cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To date, query performance prediction (QPP) in the context of content-based
image retrieval remains a largely unexplored task, especially in the
query-by-example scenario, where the query is an image. To boost the
exploration of the QPP task in image retrieval, we propose the first benchmark
for image query performance prediction (iQPP). First, we establish a set of
four data sets (PASCAL VOC 2012, Caltech-101, ROxford5k and RParis6k) and
estimate the ground-truth difficulty of each query as the average precision or
the precision@k, using two state-of-the-art image retrieval models. Next, we
propose and evaluate novel pre-retrieval and post-retrieval query performance
predictors, comparing them with existing or adapted (from text to image)
predictors. The empirical results show that most predictors do not generalize
across evaluation scenarios. Our comprehensive experiments indicate that iQPP
is a challenging benchmark, revealing an important research gap that needs to
be addressed in future work. We release our code and data as open source at
https://github.com/Eduard6421/iQPP, to foster future research.
| [
{
"version": "v1",
"created": "Mon, 20 Feb 2023 17:56:57 GMT"
},
{
"version": "v2",
"created": "Tue, 21 Feb 2023 09:13:06 GMT"
},
{
"version": "v3",
"created": "Mon, 10 Apr 2023 06:41:46 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Poesina",
"Eduard",
""
],
[
"Ionescu",
"Radu Tudor",
""
],
[
"Mothe",
"Josiane",
""
]
] | TITLE: iQPP: A Benchmark for Image Query Performance Prediction
ABSTRACT: To date, query performance prediction (QPP) in the context of content-based
image retrieval remains a largely unexplored task, especially in the
query-by-example scenario, where the query is an image. To boost the
exploration of the QPP task in image retrieval, we propose the first benchmark
for image query performance prediction (iQPP). First, we establish a set of
four data sets (PASCAL VOC 2012, Caltech-101, ROxford5k and RParis6k) and
estimate the ground-truth difficulty of each query as the average precision or
the precision@k, using two state-of-the-art image retrieval models. Next, we
propose and evaluate novel pre-retrieval and post-retrieval query performance
predictors, comparing them with existing or adapted (from text to image)
predictors. The empirical results show that most predictors do not generalize
across evaluation scenarios. Our comprehensive experiments indicate that iQPP
is a challenging benchmark, revealing an important research gap that needs to
be addressed in future work. We release our code and data as open source at
https://github.com/Eduard6421/iQPP, to foster future research.
|
2302.11559 | MD Shamimul Islam | Md Shamimul Islam, A.J.M. Akhtarujjaman Joha, Md Nur Hossain, Sohaib
Abdullah, Ibrahim Elwarfalli, Md Mahedi Hasan | Word level Bangla Sign Language Dataset for Continuous BSL Recognition | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | An robust sign language recognition system can greatly alleviate
communication barriers, particularly for people who struggle with verbal
communication. This is crucial for human growth and progress as it enables the
expression of thoughts, feelings, and ideas. However, sign recognition is a
complex task that faces numerous challenges such as same gesture patterns for
multiple signs, lighting, clothing, carrying conditions, and the presence of
large poses, as well as illumination discrepancies across different views.
Additionally, the absence of an extensive Bangla sign language video dataset
makes it even more challenging to operate recognition systems, particularly
when utilizing deep learning techniques. In order to address this issue,
firstly, we created a large-scale dataset called the MVBSL-W50, which comprises
50 isolated words across 13 categories. Secondly, we developed an
attention-based Bi-GRU model that captures the temporal dynamics of pose
information for individuals communicating through sign language. The proposed
model utilizes human pose information, which has shown to be successful in
analyzing sign language patterns. By focusing solely on movement information
and disregarding body appearance and environmental factors, the model is
simplified and can achieve a speedier performance. The accuracy of the model is
reported to be 85.64%.
| [
{
"version": "v1",
"created": "Wed, 22 Feb 2023 18:55:54 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Apr 2023 18:48:21 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Islam",
"Md Shamimul",
""
],
[
"Joha",
"A. J. M. Akhtarujjaman",
""
],
[
"Hossain",
"Md Nur",
""
],
[
"Abdullah",
"Sohaib",
""
],
[
"Elwarfalli",
"Ibrahim",
""
],
[
"Hasan",
"Md Mahedi",
""
]
] | TITLE: Word level Bangla Sign Language Dataset for Continuous BSL Recognition
ABSTRACT: An robust sign language recognition system can greatly alleviate
communication barriers, particularly for people who struggle with verbal
communication. This is crucial for human growth and progress as it enables the
expression of thoughts, feelings, and ideas. However, sign recognition is a
complex task that faces numerous challenges such as same gesture patterns for
multiple signs, lighting, clothing, carrying conditions, and the presence of
large poses, as well as illumination discrepancies across different views.
Additionally, the absence of an extensive Bangla sign language video dataset
makes it even more challenging to operate recognition systems, particularly
when utilizing deep learning techniques. In order to address this issue,
firstly, we created a large-scale dataset called the MVBSL-W50, which comprises
50 isolated words across 13 categories. Secondly, we developed an
attention-based Bi-GRU model that captures the temporal dynamics of pose
information for individuals communicating through sign language. The proposed
model utilizes human pose information, which has shown to be successful in
analyzing sign language patterns. By focusing solely on movement information
and disregarding body appearance and environmental factors, the model is
simplified and can achieve a speedier performance. The accuracy of the model is
reported to be 85.64%.
|
2302.13721 | Thushan Sivalingam | Maheshi Lokumarambage, Vishnu Gowrisetty, Hossein Rezaei, Thushan
Sivalingam, Nandana Rajatheva, Anil Fernando | Wireless End-to-End Image Transmission System using Semantic
Communications | Accepted for IEEE Access | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semantic communication is considered the future of mobile communication,
which aims to transmit data beyond Shannon's theorem of communications by
transmitting the semantic meaning of the data rather than the bit-by-bit
reconstruction of the data at the receiver's end. The semantic communication
paradigm aims to bridge the gap of limited bandwidth problems in modern
high-volume multimedia application content transmission. Integrating AI
technologies with the 6G communications networks paved the way to develop
semantic communication-based end-to-end communication systems. In this study,
we have implemented a semantic communication-based end-to-end image
transmission system, and we discuss potential design considerations in
developing semantic communication systems in conjunction with physical channel
characteristics. A Pre-trained GAN network is used at the receiver as the
transmission task to reconstruct the realistic image based on the Semantic
segmented image at the receiver input. The semantic segmentation task at the
transmitter (encoder) and the GAN network at the receiver (decoder) is trained
on a common knowledge base, the COCO-Stuff dataset. The research shows that the
resource gain in the form of bandwidth saving is immense when transmitting the
semantic segmentation map through the physical channel instead of the ground
truth image in contrast to conventional communication systems. Furthermore, the
research studies the effect of physical channel distortions and quantization
noise on semantic communication-based multimedia content transmission.
| [
{
"version": "v1",
"created": "Mon, 27 Feb 2023 12:33:53 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 12:41:42 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Lokumarambage",
"Maheshi",
""
],
[
"Gowrisetty",
"Vishnu",
""
],
[
"Rezaei",
"Hossein",
""
],
[
"Sivalingam",
"Thushan",
""
],
[
"Rajatheva",
"Nandana",
""
],
[
"Fernando",
"Anil",
""
]
] | TITLE: Wireless End-to-End Image Transmission System using Semantic
Communications
ABSTRACT: Semantic communication is considered the future of mobile communication,
which aims to transmit data beyond Shannon's theorem of communications by
transmitting the semantic meaning of the data rather than the bit-by-bit
reconstruction of the data at the receiver's end. The semantic communication
paradigm aims to bridge the gap of limited bandwidth problems in modern
high-volume multimedia application content transmission. Integrating AI
technologies with the 6G communications networks paved the way to develop
semantic communication-based end-to-end communication systems. In this study,
we have implemented a semantic communication-based end-to-end image
transmission system, and we discuss potential design considerations in
developing semantic communication systems in conjunction with physical channel
characteristics. A Pre-trained GAN network is used at the receiver as the
transmission task to reconstruct the realistic image based on the Semantic
segmented image at the receiver input. The semantic segmentation task at the
transmitter (encoder) and the GAN network at the receiver (decoder) is trained
on a common knowledge base, the COCO-Stuff dataset. The research shows that the
resource gain in the form of bandwidth saving is immense when transmitting the
semantic segmentation map through the physical channel instead of the ground
truth image in contrast to conventional communication systems. Furthermore, the
research studies the effect of physical channel distortions and quantization
noise on semantic communication-based multimedia content transmission.
|
2303.01559 | Haozhe Liu | Haozhe Liu, Wentian Zhang, Bing Li, Haoqian Wu, Nanjun He, Yawen
Huang, Yuexiang Li, Bernard Ghanem, Yefeng Zheng | Improving GAN Training via Feature Space Shrinkage | Accepted by CVPR'2023. Code and Demo are available at
https://github.com/WentianZhang-ML/AdaptiveMix | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the outstanding capability for data generation, Generative Adversarial
Networks (GANs) have attracted considerable attention in unsupervised learning.
However, training GANs is difficult, since the training distribution is dynamic
for the discriminator, leading to unstable image representation. In this paper,
we address the problem of training GANs from a novel perspective, \emph{i.e.,}
robust image classification. Motivated by studies on robust image
representation, we propose a simple yet effective module, namely AdaptiveMix,
for GANs, which shrinks the regions of training data in the image
representation space of the discriminator. Considering it is intractable to
directly bound feature space, we propose to construct hard samples and narrow
down the feature distance between hard and easy samples. The hard samples are
constructed by mixing a pair of training images. We evaluate the effectiveness
of our AdaptiveMix with widely-used and state-of-the-art GAN architectures. The
evaluation results demonstrate that our AdaptiveMix can facilitate the training
of GANs and effectively improve the image quality of generated samples. We also
show that our AdaptiveMix can be further applied to image classification and
Out-Of-Distribution (OOD) detection tasks, by equipping it with
state-of-the-art methods. Extensive experiments on seven publicly available
datasets show that our method effectively boosts the performance of baselines.
The code is publicly available at
https://github.com/WentianZhang-ML/AdaptiveMix.
| [
{
"version": "v1",
"created": "Thu, 2 Mar 2023 20:22:24 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 11:36:38 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Liu",
"Haozhe",
""
],
[
"Zhang",
"Wentian",
""
],
[
"Li",
"Bing",
""
],
[
"Wu",
"Haoqian",
""
],
[
"He",
"Nanjun",
""
],
[
"Huang",
"Yawen",
""
],
[
"Li",
"Yuexiang",
""
],
[
"Ghanem",
"Bernard",
""
],
[
"Zheng",
"Yefeng",
""
]
] | TITLE: Improving GAN Training via Feature Space Shrinkage
ABSTRACT: Due to the outstanding capability for data generation, Generative Adversarial
Networks (GANs) have attracted considerable attention in unsupervised learning.
However, training GANs is difficult, since the training distribution is dynamic
for the discriminator, leading to unstable image representation. In this paper,
we address the problem of training GANs from a novel perspective, \emph{i.e.,}
robust image classification. Motivated by studies on robust image
representation, we propose a simple yet effective module, namely AdaptiveMix,
for GANs, which shrinks the regions of training data in the image
representation space of the discriminator. Considering it is intractable to
directly bound feature space, we propose to construct hard samples and narrow
down the feature distance between hard and easy samples. The hard samples are
constructed by mixing a pair of training images. We evaluate the effectiveness
of our AdaptiveMix with widely-used and state-of-the-art GAN architectures. The
evaluation results demonstrate that our AdaptiveMix can facilitate the training
of GANs and effectively improve the image quality of generated samples. We also
show that our AdaptiveMix can be further applied to image classification and
Out-Of-Distribution (OOD) detection tasks, by equipping it with
state-of-the-art methods. Extensive experiments on seven publicly available
datasets show that our method effectively boosts the performance of baselines.
The code is publicly available at
https://github.com/WentianZhang-ML/AdaptiveMix.
|
2303.06420 | Amit Puri | Amit Puri, John Jose, Tamarapalli Venkatesh | Design and Evaluation of a Rack-Scale Disaggregated Memory Architecture
For Data Centers | null | null | 10.1109/HPCC-DSS-SmartCity-DependSys57074.2022.00060 | null | cs.DC cs.AR | http://creativecommons.org/licenses/by/4.0/ | Memory disaggregation is being considered as a strong alternative to
traditional architecture to deal with the memory under-utilization in data
centers. Disaggregated memory can adapt to dynamically changing memory
requirements for the data center applications like data analytics, big data,
etc., that require in-memory processing. However, such systems can face high
remote memory access latency due to the interconnect speeds. In this paper, we
explore a rack-scale disaggregated memory architecture and discuss the various
design aspects. We design a trace-driven simulator that combines an event-based
interconnect and a cycle-accurate memory simulator to evaluate the performance
of disaggregated memory system at the rack scale. Our study shows that not only
the interconnect but the contention in the remote memory queues also adds
significantly to remote memory access latency. We introduces a memory
allocation policy to reduce the latency compared to the conventional policies.
We conduct experiments using various benchmarks with diverse memory access
patterns. Our study shows encouraging results towards the rack-scale memory
disaggregation and acceptable average memory access latency.
| [
{
"version": "v1",
"created": "Sat, 11 Mar 2023 15:04:58 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 07:35:23 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Puri",
"Amit",
""
],
[
"Jose",
"John",
""
],
[
"Venkatesh",
"Tamarapalli",
""
]
] | TITLE: Design and Evaluation of a Rack-Scale Disaggregated Memory Architecture
For Data Centers
ABSTRACT: Memory disaggregation is being considered as a strong alternative to
traditional architecture to deal with the memory under-utilization in data
centers. Disaggregated memory can adapt to dynamically changing memory
requirements for the data center applications like data analytics, big data,
etc., that require in-memory processing. However, such systems can face high
remote memory access latency due to the interconnect speeds. In this paper, we
explore a rack-scale disaggregated memory architecture and discuss the various
design aspects. We design a trace-driven simulator that combines an event-based
interconnect and a cycle-accurate memory simulator to evaluate the performance
of disaggregated memory system at the rack scale. Our study shows that not only
the interconnect but the contention in the remote memory queues also adds
significantly to remote memory access latency. We introduces a memory
allocation policy to reduce the latency compared to the conventional policies.
We conduct experiments using various benchmarks with diverse memory access
patterns. Our study shows encouraging results towards the rack-scale memory
disaggregation and acceptable average memory access latency.
|
2303.08435 | Guojin Chen | Guojin Chen, Zehua Pei, Haoyu Yang, Yuzhe Ma, Bei Yu, Martin D. F.
Wong | Physics-Informed Optical Kernel Regression Using Complex-valued Neural
Fields | Accepted by DAC23 | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Lithography is fundamental to integrated circuit fabrication, necessitating
large computation overhead. The advancement of machine learning (ML)-based
lithography models alleviates the trade-offs between manufacturing process
expense and capability. However, all previous methods regard the lithography
system as an image-to-image black box mapping, utilizing network parameters to
learn by rote mappings from massive mask-to-aerial or mask-to-resist image
pairs, resulting in poor generalization capability. In this paper, we propose a
new ML-based paradigm disassembling the rigorous lithographic model into
non-parametric mask operations and learned optical kernels containing
determinant source, pupil, and lithography information. By optimizing
complex-valued neural fields to perform optical kernel regression from
coordinates, our method can accurately restore lithography system using a
small-scale training dataset with fewer parameters, demonstrating superior
generalization capability as well. Experiments show that our framework can use
31% of parameters while achieving 69$\times$ smaller mean squared error with
1.3$\times$ higher throughput than the state-of-the-art.
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 08:17:07 GMT"
},
{
"version": "v2",
"created": "Mon, 27 Mar 2023 07:11:17 GMT"
},
{
"version": "v3",
"created": "Sun, 2 Apr 2023 08:20:29 GMT"
},
{
"version": "v4",
"created": "Sun, 9 Apr 2023 07:57:46 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Chen",
"Guojin",
""
],
[
"Pei",
"Zehua",
""
],
[
"Yang",
"Haoyu",
""
],
[
"Ma",
"Yuzhe",
""
],
[
"Yu",
"Bei",
""
],
[
"Wong",
"Martin D. F.",
""
]
] | TITLE: Physics-Informed Optical Kernel Regression Using Complex-valued Neural
Fields
ABSTRACT: Lithography is fundamental to integrated circuit fabrication, necessitating
large computation overhead. The advancement of machine learning (ML)-based
lithography models alleviates the trade-offs between manufacturing process
expense and capability. However, all previous methods regard the lithography
system as an image-to-image black box mapping, utilizing network parameters to
learn by rote mappings from massive mask-to-aerial or mask-to-resist image
pairs, resulting in poor generalization capability. In this paper, we propose a
new ML-based paradigm disassembling the rigorous lithographic model into
non-parametric mask operations and learned optical kernels containing
determinant source, pupil, and lithography information. By optimizing
complex-valued neural fields to perform optical kernel regression from
coordinates, our method can accurately restore lithography system using a
small-scale training dataset with fewer parameters, demonstrating superior
generalization capability as well. Experiments show that our framework can use
31% of parameters while achieving 69$\times$ smaller mean squared error with
1.3$\times$ higher throughput than the state-of-the-art.
|
2303.10800 | Nathan Inkawhich | Nathan Inkawhich | A Global Model Approach to Robust Few-Shot SAR Automatic Target
Recognition | null | null | 10.1109/LGRS.2023.3264535 | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In real-world scenarios, it may not always be possible to collect hundreds of
labeled samples per class for training deep learning-based SAR Automatic Target
Recognition (ATR) models. This work specifically tackles the few-shot SAR ATR
problem, where only a handful of labeled samples may be available to support
the task of interest. Our approach is composed of two stages. In the first, a
global representation model is trained via self-supervised learning on a large
pool of diverse and unlabeled SAR data. In the second stage, the global model
is used as a fixed feature extractor and a classifier is trained to partition
the feature space given the few-shot support samples, while simultaneously
being calibrated to detect anomalous inputs. Unlike competing approaches which
require a pristine labeled dataset for pretraining via meta-learning, our
approach learns highly transferable features from unlabeled data that have
little-to-no relation to the downstream task. We evaluate our method in
standard and extended MSTAR operating conditions and find it to achieve high
accuracy and robust out-of-distribution detection in many different few-shot
settings. Our results are particularly significant because they show the merit
of a global model approach to SAR ATR, which makes minimal assumptions, and
provides many axes for extendability.
| [
{
"version": "v1",
"created": "Mon, 20 Mar 2023 00:24:05 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Inkawhich",
"Nathan",
""
]
] | TITLE: A Global Model Approach to Robust Few-Shot SAR Automatic Target
Recognition
ABSTRACT: In real-world scenarios, it may not always be possible to collect hundreds of
labeled samples per class for training deep learning-based SAR Automatic Target
Recognition (ATR) models. This work specifically tackles the few-shot SAR ATR
problem, where only a handful of labeled samples may be available to support
the task of interest. Our approach is composed of two stages. In the first, a
global representation model is trained via self-supervised learning on a large
pool of diverse and unlabeled SAR data. In the second stage, the global model
is used as a fixed feature extractor and a classifier is trained to partition
the feature space given the few-shot support samples, while simultaneously
being calibrated to detect anomalous inputs. Unlike competing approaches which
require a pristine labeled dataset for pretraining via meta-learning, our
approach learns highly transferable features from unlabeled data that have
little-to-no relation to the downstream task. We evaluate our method in
standard and extended MSTAR operating conditions and find it to achieve high
accuracy and robust out-of-distribution detection in many different few-shot
settings. Our results are particularly significant because they show the merit
of a global model approach to SAR ATR, which makes minimal assumptions, and
provides many axes for extendability.
|
2303.11470 | Ruixiang Tang | Ruixiang Tang, Qizhang Feng, Ninghao Liu, Fan Yang, Xia Hu | Did You Train on My Dataset? Towards Public Dataset Protection with
Clean-Label Backdoor Watermarking | null | null | null | null | cs.CR cs.AI cs.LG cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The huge supporting training data on the Internet has been a key factor in
the success of deep learning models. However, this abundance of
public-available data also raises concerns about the unauthorized exploitation
of datasets for commercial purposes, which is forbidden by dataset licenses. In
this paper, we propose a backdoor-based watermarking approach that serves as a
general framework for safeguarding public-available data. By inserting a small
number of watermarking samples into the dataset, our approach enables the
learning model to implicitly learn a secret function set by defenders. This
hidden function can then be used as a watermark to track down third-party
models that use the dataset illegally. Unfortunately, existing backdoor
insertion methods often entail adding arbitrary and mislabeled data to the
training set, leading to a significant drop in performance and easy detection
by anomaly detection algorithms. To overcome this challenge, we introduce a
clean-label backdoor watermarking framework that uses imperceptible
perturbations to replace mislabeled samples. As a result, the watermarking
samples remain consistent with the original labels, making them difficult to
detect. Our experiments on text, image, and audio datasets demonstrate that the
proposed framework effectively safeguards datasets with minimal impact on
original task performance. We also show that adding just 1% of watermarking
samples can inject a traceable watermarking function and that our watermarking
samples are stealthy and look benign upon visual inspection.
| [
{
"version": "v1",
"created": "Mon, 20 Mar 2023 21:54:30 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 06:21:19 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Tang",
"Ruixiang",
""
],
[
"Feng",
"Qizhang",
""
],
[
"Liu",
"Ninghao",
""
],
[
"Yang",
"Fan",
""
],
[
"Hu",
"Xia",
""
]
] | TITLE: Did You Train on My Dataset? Towards Public Dataset Protection with
Clean-Label Backdoor Watermarking
ABSTRACT: The huge supporting training data on the Internet has been a key factor in
the success of deep learning models. However, this abundance of
public-available data also raises concerns about the unauthorized exploitation
of datasets for commercial purposes, which is forbidden by dataset licenses. In
this paper, we propose a backdoor-based watermarking approach that serves as a
general framework for safeguarding public-available data. By inserting a small
number of watermarking samples into the dataset, our approach enables the
learning model to implicitly learn a secret function set by defenders. This
hidden function can then be used as a watermark to track down third-party
models that use the dataset illegally. Unfortunately, existing backdoor
insertion methods often entail adding arbitrary and mislabeled data to the
training set, leading to a significant drop in performance and easy detection
by anomaly detection algorithms. To overcome this challenge, we introduce a
clean-label backdoor watermarking framework that uses imperceptible
perturbations to replace mislabeled samples. As a result, the watermarking
samples remain consistent with the original labels, making them difficult to
detect. Our experiments on text, image, and audio datasets demonstrate that the
proposed framework effectively safeguards datasets with minimal impact on
original task performance. We also show that adding just 1% of watermarking
samples can inject a traceable watermarking function and that our watermarking
samples are stealthy and look benign upon visual inspection.
|
2303.14042 | Zilin Luo | Zilin Luo, Yaoyao Liu, Bernt Schiele, Qianru Sun | Class-Incremental Exemplar Compression for Class-Incremental Learning | Accepted to CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Exemplar-based class-incremental learning (CIL) finetunes the model with all
samples of new classes but few-shot exemplars of old classes in each
incremental phase, where the "few-shot" abides by the limited memory budget. In
this paper, we break this "few-shot" limit based on a simple yet surprisingly
effective idea: compressing exemplars by downsampling non-discriminative pixels
and saving "many-shot" compressed exemplars in the memory. Without needing any
manual annotation, we achieve this compression by generating 0-1 masks on
discriminative pixels from class activation maps (CAM). We propose an adaptive
mask generation model called class-incremental masking (CIM) to explicitly
resolve two difficulties of using CAM: 1) transforming the heatmaps of CAM to
0-1 masks with an arbitrary threshold leads to a trade-off between the coverage
on discriminative pixels and the quantity of exemplars, as the total memory is
fixed; and 2) optimal thresholds vary for different object classes, which is
particularly obvious in the dynamic environment of CIL. We optimize the CIM
model alternatively with the conventional CIL model through a bilevel
optimization problem. We conduct extensive experiments on high-resolution CIL
benchmarks including Food-101, ImageNet-100, and ImageNet-1000, and show that
using the compressed exemplars by CIM can achieve a new state-of-the-art CIL
accuracy, e.g., 4.8 percentage points higher than FOSTER on 10-Phase
ImageNet-1000. Our code is available at https://github.com/xfflzl/CIM-CIL.
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 14:51:20 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 03:25:41 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Luo",
"Zilin",
""
],
[
"Liu",
"Yaoyao",
""
],
[
"Schiele",
"Bernt",
""
],
[
"Sun",
"Qianru",
""
]
] | TITLE: Class-Incremental Exemplar Compression for Class-Incremental Learning
ABSTRACT: Exemplar-based class-incremental learning (CIL) finetunes the model with all
samples of new classes but few-shot exemplars of old classes in each
incremental phase, where the "few-shot" abides by the limited memory budget. In
this paper, we break this "few-shot" limit based on a simple yet surprisingly
effective idea: compressing exemplars by downsampling non-discriminative pixels
and saving "many-shot" compressed exemplars in the memory. Without needing any
manual annotation, we achieve this compression by generating 0-1 masks on
discriminative pixels from class activation maps (CAM). We propose an adaptive
mask generation model called class-incremental masking (CIM) to explicitly
resolve two difficulties of using CAM: 1) transforming the heatmaps of CAM to
0-1 masks with an arbitrary threshold leads to a trade-off between the coverage
on discriminative pixels and the quantity of exemplars, as the total memory is
fixed; and 2) optimal thresholds vary for different object classes, which is
particularly obvious in the dynamic environment of CIL. We optimize the CIM
model alternatively with the conventional CIL model through a bilevel
optimization problem. We conduct extensive experiments on high-resolution CIL
benchmarks including Food-101, ImageNet-100, and ImageNet-1000, and show that
using the compressed exemplars by CIM can achieve a new state-of-the-art CIL
accuracy, e.g., 4.8 percentage points higher than FOSTER on 10-Phase
ImageNet-1000. Our code is available at https://github.com/xfflzl/CIM-CIL.
|
2303.14584 | Phani Krishna Uppala | Phani Krishna Uppala, Abhishek Bamotra, Shriti Priya, Vaidehi Joshi | Learning video embedding space with Natural Language Supervision | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The recent success of the CLIP model has shown its potential to be applied to
a wide range of vision and language tasks. However this only establishes
embedding space relationship of language to images, not to the video domain. In
this paper, we propose a novel approach to map video embedding space to natural
langugage. We propose a two-stage approach that first extracts visual features
from each frame of a video using a pre-trained CNN, and then uses the CLIP
model to encode the visual features for the video domain, along with the
corresponding text descriptions. We evaluate our method on two benchmark
datasets, UCF101 and HMDB51, and achieve state-of-the-art performance on both
tasks.
| [
{
"version": "v1",
"created": "Sat, 25 Mar 2023 23:24:57 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 02:44:20 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Uppala",
"Phani Krishna",
""
],
[
"Bamotra",
"Abhishek",
""
],
[
"Priya",
"Shriti",
""
],
[
"Joshi",
"Vaidehi",
""
]
] | TITLE: Learning video embedding space with Natural Language Supervision
ABSTRACT: The recent success of the CLIP model has shown its potential to be applied to
a wide range of vision and language tasks. However this only establishes
embedding space relationship of language to images, not to the video domain. In
this paper, we propose a novel approach to map video embedding space to natural
langugage. We propose a two-stage approach that first extracts visual features
from each frame of a video using a pre-trained CNN, and then uses the CLIP
model to encode the visual features for the video domain, along with the
corresponding text descriptions. We evaluate our method on two benchmark
datasets, UCF101 and HMDB51, and achieve state-of-the-art performance on both
tasks.
|
2303.16071 | Li-Hsiang Shen | Chih-Yu Chen, Li-Hsiang Shen, Kai-Ten Feng, Lie-Liang Yang, and
Jen-Ming Wu | Edge Selection and Clustering for Federated Learning in Optical
Inter-LEO Satellite Constellation | null | null | null | null | cs.LG cs.NI eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Low-Earth orbit (LEO) satellites have been prosperously deployed for various
Earth observation missions due to its capability of collecting a large amount
of image or sensor data. However, traditionally, the data training process is
performed in the terrestrial cloud server, which leads to a high transmission
overhead. With the recent development of LEO, it is more imperative to provide
ultra-dense LEO constellation with enhanced on-board computation capability.
Benefited from it, we have proposed a collaborative federated learning for low
Earth orbit (FELLO). We allocate the entire process on LEOs with low payload
inter-satellite transmissions, whilst the low-delay terrestrial gateway server
(GS) only takes care for initial signal controlling. The GS initially selects
an LEO server, whereas its LEO clients are all determined by clustering
mechanism and communication capability through the optical inter-satellite
links (ISLs). The re-clustering of changing LEO server will be executed once
with low communication quality of FELLO. In the simulations, we have
numerically analyzed the proposed FELLO under practical Walker-based LEO
constellation configurations along with MNIST training dataset for
classification mission. The proposed FELLO outperforms the conventional
centralized and distributed architectures with higher classification accuracy
as well as comparably lower latency of joint communication and computing.
| [
{
"version": "v1",
"created": "Sat, 25 Mar 2023 04:24:16 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 06:13:34 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Chen",
"Chih-Yu",
""
],
[
"Shen",
"Li-Hsiang",
""
],
[
"Feng",
"Kai-Ten",
""
],
[
"Yang",
"Lie-Liang",
""
],
[
"Wu",
"Jen-Ming",
""
]
] | TITLE: Edge Selection and Clustering for Federated Learning in Optical
Inter-LEO Satellite Constellation
ABSTRACT: Low-Earth orbit (LEO) satellites have been prosperously deployed for various
Earth observation missions due to its capability of collecting a large amount
of image or sensor data. However, traditionally, the data training process is
performed in the terrestrial cloud server, which leads to a high transmission
overhead. With the recent development of LEO, it is more imperative to provide
ultra-dense LEO constellation with enhanced on-board computation capability.
Benefited from it, we have proposed a collaborative federated learning for low
Earth orbit (FELLO). We allocate the entire process on LEOs with low payload
inter-satellite transmissions, whilst the low-delay terrestrial gateway server
(GS) only takes care for initial signal controlling. The GS initially selects
an LEO server, whereas its LEO clients are all determined by clustering
mechanism and communication capability through the optical inter-satellite
links (ISLs). The re-clustering of changing LEO server will be executed once
with low communication quality of FELLO. In the simulations, we have
numerically analyzed the proposed FELLO under practical Walker-based LEO
constellation configurations along with MNIST training dataset for
classification mission. The proposed FELLO outperforms the conventional
centralized and distributed architectures with higher classification accuracy
as well as comparably lower latency of joint communication and computing.
|
2304.00160 | Tuo Zhang | Duygu Nur Yaldiz, Tuo Zhang and Salman Avestimehr | Secure Federated Learning against Model Poisoning Attacks via Client
Filtering | null | ICLR 2023 Workshop on Backdoor Attacks and Defenses in Machine
Learning | null | null | cs.CR cs.DC | http://creativecommons.org/licenses/by/4.0/ | Given the distributed nature, detecting and defending against the backdoor
attack under federated learning (FL) systems is challenging. In this paper, we
observe that the cosine similarity of the last layer's weight between the
global model and each local update could be used effectively as an indicator of
malicious model updates. Therefore, we propose CosDefense, a
cosine-similarity-based attacker detection algorithm. Specifically, under
CosDefense, the server calculates the cosine similarity score of the last
layer's weight between the global model and each client update, labels
malicious clients whose score is much higher than the average, and filters them
out of the model aggregation in each round. Compared to existing defense
schemes, CosDefense does not require any extra information besides the received
model updates to operate and is compatible with client sampling. Experiment
results on three real-world datasets demonstrate that CosDefense could provide
robust performance under the state-of-the-art FL poisoning attack.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 22:49:01 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Apr 2023 01:19:30 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Yaldiz",
"Duygu Nur",
""
],
[
"Zhang",
"Tuo",
""
],
[
"Avestimehr",
"Salman",
""
]
] | TITLE: Secure Federated Learning against Model Poisoning Attacks via Client
Filtering
ABSTRACT: Given the distributed nature, detecting and defending against the backdoor
attack under federated learning (FL) systems is challenging. In this paper, we
observe that the cosine similarity of the last layer's weight between the
global model and each local update could be used effectively as an indicator of
malicious model updates. Therefore, we propose CosDefense, a
cosine-similarity-based attacker detection algorithm. Specifically, under
CosDefense, the server calculates the cosine similarity score of the last
layer's weight between the global model and each client update, labels
malicious clients whose score is much higher than the average, and filters them
out of the model aggregation in each round. Compared to existing defense
schemes, CosDefense does not require any extra information besides the received
model updates to operate and is compatible with client sampling. Experiment
results on three real-world datasets demonstrate that CosDefense could provide
robust performance under the state-of-the-art FL poisoning attack.
|
2304.00601 | Ligong Han | Ligong Han, Seungwook Han, Shivchander Sudalairaj, Charlotte Loh,
Rumen Dangovski, Fei Deng, Pulkit Agrawal, Dimitris Metaxas, Leonid
Karlinsky, Tsui-Wei Weng, Akash Srivastava | Constructive Assimilation: Boosting Contrastive Learning Performance
through View Generation Strategies | Accepted at Generative Models for Computer Vision Workshop 2023 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformations based on domain expertise (expert transformations), such as
random-resized-crop and color-jitter, have proven critical to the success of
contrastive learning techniques such as SimCLR. Recently, several attempts have
been made to replace such domain-specific, human-designed transformations with
generated views that are learned. However for imagery data, so far none of
these view-generation methods has been able to outperform expert
transformations. In this work, we tackle a different question: instead of
replacing expert transformations with generated views, can we constructively
assimilate generated views with expert transformations? We answer this question
in the affirmative and propose a view generation method and a simple, effective
assimilation method that together improve the state-of-the-art by up to ~3.6%
on three different datasets. Importantly, we conduct a detailed empirical study
that systematically analyzes a range of view generation and assimilation
methods and provides a holistic picture of the efficacy of learned views in
contrastive representation learning.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 19:09:01 GMT"
},
{
"version": "v2",
"created": "Sat, 8 Apr 2023 22:26:50 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Han",
"Ligong",
""
],
[
"Han",
"Seungwook",
""
],
[
"Sudalairaj",
"Shivchander",
""
],
[
"Loh",
"Charlotte",
""
],
[
"Dangovski",
"Rumen",
""
],
[
"Deng",
"Fei",
""
],
[
"Agrawal",
"Pulkit",
""
],
[
"Metaxas",
"Dimitris",
""
],
[
"Karlinsky",
"Leonid",
""
],
[
"Weng",
"Tsui-Wei",
""
],
[
"Srivastava",
"Akash",
""
]
] | TITLE: Constructive Assimilation: Boosting Contrastive Learning Performance
through View Generation Strategies
ABSTRACT: Transformations based on domain expertise (expert transformations), such as
random-resized-crop and color-jitter, have proven critical to the success of
contrastive learning techniques such as SimCLR. Recently, several attempts have
been made to replace such domain-specific, human-designed transformations with
generated views that are learned. However for imagery data, so far none of
these view-generation methods has been able to outperform expert
transformations. In this work, we tackle a different question: instead of
replacing expert transformations with generated views, can we constructively
assimilate generated views with expert transformations? We answer this question
in the affirmative and propose a view generation method and a simple, effective
assimilation method that together improve the state-of-the-art by up to ~3.6%
on three different datasets. Importantly, we conduct a detailed empirical study
that systematically analyzes a range of view generation and assimilation
methods and provides a holistic picture of the efficacy of learned views in
contrastive representation learning.
|
2304.00961 | PengWan Yang | Pengwan Yang, Cees G. M. Snoek, Yuki M. Asano | Self-Ordering Point Clouds | null | null | null | null | cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | In this paper we address the task of finding representative subsets of points
in a 3D point cloud by means of a point-wise ordering. Only a few works have
tried to address this challenging vision problem, all with the help of hard to
obtain point and cloud labels. Different from these works, we introduce the
task of point-wise ordering in 3D point clouds through self-supervision, which
we call self-ordering. We further contribute the first end-to-end trainable
network that learns a point-wise ordering in a self-supervised fashion. It
utilizes a novel differentiable point scoring-sorting strategy and it
constructs an hierarchical contrastive scheme to obtain self-supervision
signals. We extensively ablate the method and show its scalability and superior
performance even compared to supervised ordering methods on multiple datasets
and tasks including zero-shot ordering of point clouds from unseen categories.
| [
{
"version": "v1",
"created": "Mon, 3 Apr 2023 13:26:52 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 10:31:07 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Yang",
"Pengwan",
""
],
[
"Snoek",
"Cees G. M.",
""
],
[
"Asano",
"Yuki M.",
""
]
] | TITLE: Self-Ordering Point Clouds
ABSTRACT: In this paper we address the task of finding representative subsets of points
in a 3D point cloud by means of a point-wise ordering. Only a few works have
tried to address this challenging vision problem, all with the help of hard to
obtain point and cloud labels. Different from these works, we introduce the
task of point-wise ordering in 3D point clouds through self-supervision, which
we call self-ordering. We further contribute the first end-to-end trainable
network that learns a point-wise ordering in a self-supervised fashion. It
utilizes a novel differentiable point scoring-sorting strategy and it
constructs an hierarchical contrastive scheme to obtain self-supervision
signals. We extensively ablate the method and show its scalability and superior
performance even compared to supervised ordering methods on multiple datasets
and tasks including zero-shot ordering of point clouds from unseen categories.
|
2304.01108 | Jordan Suchow | Jordan W. Suchow and Necdet G\"urkan | Coincidental Generation | null | null | null | null | cs.CV cs.CR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative A.I. models have emerged as versatile tools across diverse
industries, with applications in privacy-preserving data sharing, computational
art, personalization of products and services, and immersive entertainment.
Here, we introduce a new privacy concern in the adoption and use of generative
A.I. models: that of coincidental generation, where a generative model's output
is similar enough to an existing entity, beyond those represented in the
dataset used to train the model, to be mistaken for it. Consider, for example,
synthetic portrait generators, which are today deployed in commercial
applications such as virtual modeling agencies and synthetic stock photography.
Due to the low intrinsic dimensionality of human face perception, every
synthetically generated face will coincidentally resemble an actual person.
Such examples of coincidental generation all but guarantee the misappropriation
of likeness and expose organizations that use generative A.I. to legal and
regulatory risk.
| [
{
"version": "v1",
"created": "Mon, 3 Apr 2023 16:08:22 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 15:16:04 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Suchow",
"Jordan W.",
""
],
[
"Gürkan",
"Necdet",
""
]
] | TITLE: Coincidental Generation
ABSTRACT: Generative A.I. models have emerged as versatile tools across diverse
industries, with applications in privacy-preserving data sharing, computational
art, personalization of products and services, and immersive entertainment.
Here, we introduce a new privacy concern in the adoption and use of generative
A.I. models: that of coincidental generation, where a generative model's output
is similar enough to an existing entity, beyond those represented in the
dataset used to train the model, to be mistaken for it. Consider, for example,
synthetic portrait generators, which are today deployed in commercial
applications such as virtual modeling agencies and synthetic stock photography.
Due to the low intrinsic dimensionality of human face perception, every
synthetically generated face will coincidentally resemble an actual person.
Such examples of coincidental generation all but guarantee the misappropriation
of likeness and expose organizations that use generative A.I. to legal and
regulatory risk.
|
2304.03588 | Feiyang Xiao | Jian Guan, Feiyang Xiao, Youde Liu, Qiaoxi Zhu, Wenwu Wang | Anomalous Sound Detection using Audio Representation with Machine ID
based Contrastive Learning Pretraining | To appear in IEEE International Conference on Acoustics, Speech, and
Signal Processing (ICASSP 2023) | null | null | null | cs.SD cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing contrastive learning methods for anomalous sound detection refine
the audio representation of each audio sample by using the contrast between the
samples' augmentations (e.g., with time or frequency masking). However, they
might be biased by the augmented data, due to the lack of physical properties
of machine sound, thereby limiting the detection performance. This paper uses
contrastive learning to refine audio representations for each machine ID,
rather than for each audio sample. The proposed two-stage method uses
contrastive learning to pretrain the audio representation model by
incorporating machine ID and a self-supervised ID classifier to fine-tune the
learnt model, while enhancing the relation between audio features from the same
ID. Experiments show that our method outperforms the state-of-the-art methods
using contrastive learning or self-supervised classification in overall anomaly
detection performance and stability on DCASE 2020 Challenge Task2 dataset.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 11:08:31 GMT"
},
{
"version": "v2",
"created": "Mon, 10 Apr 2023 04:04:38 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Guan",
"Jian",
""
],
[
"Xiao",
"Feiyang",
""
],
[
"Liu",
"Youde",
""
],
[
"Zhu",
"Qiaoxi",
""
],
[
"Wang",
"Wenwu",
""
]
] | TITLE: Anomalous Sound Detection using Audio Representation with Machine ID
based Contrastive Learning Pretraining
ABSTRACT: Existing contrastive learning methods for anomalous sound detection refine
the audio representation of each audio sample by using the contrast between the
samples' augmentations (e.g., with time or frequency masking). However, they
might be biased by the augmented data, due to the lack of physical properties
of machine sound, thereby limiting the detection performance. This paper uses
contrastive learning to refine audio representations for each machine ID,
rather than for each audio sample. The proposed two-stage method uses
contrastive learning to pretrain the audio representation model by
incorporating machine ID and a self-supervised ID classifier to fine-tune the
learnt model, while enhancing the relation between audio features from the same
ID. Experiments show that our method outperforms the state-of-the-art methods
using contrastive learning or self-supervised classification in overall anomaly
detection performance and stability on DCASE 2020 Challenge Task2 dataset.
|
2304.03785 | Ayan Das | Ayan Das, Yongxin Yang, Timothy Hospedales, Tao Xiang, Yi-Zhe Song | ChiroDiff: Modelling chirographic data with Diffusion Models | Accepted at ICLR '23 | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Generative modelling over continuous-time geometric constructs, a.k.a such as
handwriting, sketches, drawings etc., have been accomplished through
autoregressive distributions. Such strictly-ordered discrete factorization
however falls short of capturing key properties of chirographic data -- it
fails to build holistic understanding of the temporal concept due to one-way
visibility (causality). Consequently, temporal data has been modelled as
discrete token sequences of fixed sampling rate instead of capturing the true
underlying concept. In this paper, we introduce a powerful model-class namely
"Denoising Diffusion Probabilistic Models" or DDPMs for chirographic data that
specifically addresses these flaws. Our model named "ChiroDiff", being
non-autoregressive, learns to capture holistic concepts and therefore remains
resilient to higher temporal sampling rate up to a good extent. Moreover, we
show that many important downstream utilities (e.g. conditional sampling,
creative mixing) can be flexibly implemented using ChiroDiff. We further show
some unique use-cases like stochastic vectorization, de-noising/healing,
abstraction are also possible with this model-class. We perform quantitative
and qualitative evaluation of our framework on relevant datasets and found it
to be better or on par with competing approaches.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 15:17:48 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Das",
"Ayan",
""
],
[
"Yang",
"Yongxin",
""
],
[
"Hospedales",
"Timothy",
""
],
[
"Xiang",
"Tao",
""
],
[
"Song",
"Yi-Zhe",
""
]
] | TITLE: ChiroDiff: Modelling chirographic data with Diffusion Models
ABSTRACT: Generative modelling over continuous-time geometric constructs, a.k.a such as
handwriting, sketches, drawings etc., have been accomplished through
autoregressive distributions. Such strictly-ordered discrete factorization
however falls short of capturing key properties of chirographic data -- it
fails to build holistic understanding of the temporal concept due to one-way
visibility (causality). Consequently, temporal data has been modelled as
discrete token sequences of fixed sampling rate instead of capturing the true
underlying concept. In this paper, we introduce a powerful model-class namely
"Denoising Diffusion Probabilistic Models" or DDPMs for chirographic data that
specifically addresses these flaws. Our model named "ChiroDiff", being
non-autoregressive, learns to capture holistic concepts and therefore remains
resilient to higher temporal sampling rate up to a good extent. Moreover, we
show that many important downstream utilities (e.g. conditional sampling,
creative mixing) can be flexibly implemented using ChiroDiff. We further show
some unique use-cases like stochastic vectorization, de-noising/healing,
abstraction are also possible with this model-class. We perform quantitative
and qualitative evaluation of our framework on relevant datasets and found it
to be better or on par with competing approaches.
|
2304.03801 | Mukund Telukunta | Mukund Telukunta and Venkata Sriram Siddhardh Nadendla | Towards Inclusive Fairness Evaluation via Eliciting Disagreement
Feedback from Non-Expert Stakeholders | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by/4.0/ | Traditional algorithmic fairness notions rely on label feedback, which can
only be elicited from expert critics. However, in most practical applications,
several non-expert stakeholders also play a major role in the system and can
have distinctive opinions about the decision making philosophy. For example, in
kidney placement programs, transplant surgeons are very wary about accepting
kidney offers for black patients due to genetic reasons. However, non-expert
stakeholders in kidney placement programs (e.g. patients, donors and their
family members) may misinterpret such decisions from the perspective of social
discrimination. This paper evaluates group fairness notions from the viewpoint
of non-expert stakeholders, who can only provide binary
\emph{agreement/disagreement feedback} regarding the decision in context.
Specifically, two types of group fairness notions have been identified: (i)
\emph{definite notions} (e.g. calibration), which can be evaluated exactly
using disagreement feedback, and (ii) \emph{indefinite notions} (e.g. equal
opportunity) which suffer from uncertainty due to lack of label feedback. In
the case of indefinite notions, bounds are presented based on disagreement
rates, and an estimate is constructed based on established bounds. The efficacy
of all our findings are validated empirically on real human feedback dataset.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 18:09:43 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Telukunta",
"Mukund",
""
],
[
"Nadendla",
"Venkata Sriram Siddhardh",
""
]
] | TITLE: Towards Inclusive Fairness Evaluation via Eliciting Disagreement
Feedback from Non-Expert Stakeholders
ABSTRACT: Traditional algorithmic fairness notions rely on label feedback, which can
only be elicited from expert critics. However, in most practical applications,
several non-expert stakeholders also play a major role in the system and can
have distinctive opinions about the decision making philosophy. For example, in
kidney placement programs, transplant surgeons are very wary about accepting
kidney offers for black patients due to genetic reasons. However, non-expert
stakeholders in kidney placement programs (e.g. patients, donors and their
family members) may misinterpret such decisions from the perspective of social
discrimination. This paper evaluates group fairness notions from the viewpoint
of non-expert stakeholders, who can only provide binary
\emph{agreement/disagreement feedback} regarding the decision in context.
Specifically, two types of group fairness notions have been identified: (i)
\emph{definite notions} (e.g. calibration), which can be evaluated exactly
using disagreement feedback, and (ii) \emph{indefinite notions} (e.g. equal
opportunity) which suffer from uncertainty due to lack of label feedback. In
the case of indefinite notions, bounds are presented based on disagreement
rates, and an estimate is constructed based on established bounds. The efficacy
of all our findings are validated empirically on real human feedback dataset.
|
2304.03812 | Xu Cheng Dr. | Yifan Yin, Xu Cheng, Fan Shi, Xiufeng Liu, Huan Huo, Shengyong Chen | High-order Spatial Interactions Enhanced Lightweight Model for Optical
Remote Sensing Image-based Small Ship Detection | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Accurate and reliable optical remote sensing image-based small-ship detection
is crucial for maritime surveillance systems, but existing methods often
struggle with balancing detection performance and computational complexity. In
this paper, we propose a novel lightweight framework called
\textit{HSI-ShipDetectionNet} that is based on high-order spatial interactions
and is suitable for deployment on resource-limited platforms, such as
satellites and unmanned aerial vehicles. HSI-ShipDetectionNet includes a
prediction branch specifically for tiny ships and a lightweight hybrid
attention block for reduced complexity. Additionally, the use of a high-order
spatial interactions module improves advanced feature understanding and
modeling ability. Our model is evaluated using the public Kaggle marine ship
detection dataset and compared with multiple state-of-the-art models including
small object detection models, lightweight detection models, and ship detection
models. The results show that HSI-ShipDetectionNet outperforms the other models
in terms of recall, and mean average precision (mAP) while being lightweight
and suitable for deployment on resource-limited platforms.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 18:40:49 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Yin",
"Yifan",
""
],
[
"Cheng",
"Xu",
""
],
[
"Shi",
"Fan",
""
],
[
"Liu",
"Xiufeng",
""
],
[
"Huo",
"Huan",
""
],
[
"Chen",
"Shengyong",
""
]
] | TITLE: High-order Spatial Interactions Enhanced Lightweight Model for Optical
Remote Sensing Image-based Small Ship Detection
ABSTRACT: Accurate and reliable optical remote sensing image-based small-ship detection
is crucial for maritime surveillance systems, but existing methods often
struggle with balancing detection performance and computational complexity. In
this paper, we propose a novel lightweight framework called
\textit{HSI-ShipDetectionNet} that is based on high-order spatial interactions
and is suitable for deployment on resource-limited platforms, such as
satellites and unmanned aerial vehicles. HSI-ShipDetectionNet includes a
prediction branch specifically for tiny ships and a lightweight hybrid
attention block for reduced complexity. Additionally, the use of a high-order
spatial interactions module improves advanced feature understanding and
modeling ability. Our model is evaluated using the public Kaggle marine ship
detection dataset and compared with multiple state-of-the-art models including
small object detection models, lightweight detection models, and ship detection
models. The results show that HSI-ShipDetectionNet outperforms the other models
in terms of recall, and mean average precision (mAP) while being lightweight
and suitable for deployment on resource-limited platforms.
|
2304.03816 | Sarah Fakhoury | Sarah Fakhoury, Saikat Chakraborty, Madan Musuvathi, and Shuvendu K.
Lahiri | Towards Generating Functionally Correct Code Edits from Natural Language
Issue Descriptions | null | null | null | null | cs.SE cs.LG | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs), such as OpenAI's Codex, have demonstrated their
potential to generate code from natural language descriptions across a wide
range of programming tasks. Several benchmarks have recently emerged to
evaluate the ability of LLMs to generate functionally correct code from natural
language intent with respect to a set of hidden test cases. This has enabled
the research community to identify significant and reproducible advancements in
LLM capabilities. However, there is currently a lack of benchmark datasets for
assessing the ability of LLMs to generate functionally correct code edits based
on natural language descriptions of intended changes. This paper aims to
address this gap by motivating the problem NL2Fix of translating natural
language descriptions of code changes (namely bug fixes described in Issue
reports in repositories) into correct code fixes. To this end, we introduce
Defects4J-NL2Fix, a dataset of 283 Java programs from the popular Defects4J
dataset augmented with high-level descriptions of bug fixes, and empirically
evaluate the performance of several state-of-the-art LLMs for the this task.
Results show that these LLMS together are capable of generating plausible fixes
for 64.6% of the bugs, and the best LLM-based technique can achieve up to
21.20% top-1 and 35.68% top-5 accuracy on this benchmark.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 18:58:33 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Fakhoury",
"Sarah",
""
],
[
"Chakraborty",
"Saikat",
""
],
[
"Musuvathi",
"Madan",
""
],
[
"Lahiri",
"Shuvendu K.",
""
]
] | TITLE: Towards Generating Functionally Correct Code Edits from Natural Language
Issue Descriptions
ABSTRACT: Large language models (LLMs), such as OpenAI's Codex, have demonstrated their
potential to generate code from natural language descriptions across a wide
range of programming tasks. Several benchmarks have recently emerged to
evaluate the ability of LLMs to generate functionally correct code from natural
language intent with respect to a set of hidden test cases. This has enabled
the research community to identify significant and reproducible advancements in
LLM capabilities. However, there is currently a lack of benchmark datasets for
assessing the ability of LLMs to generate functionally correct code edits based
on natural language descriptions of intended changes. This paper aims to
address this gap by motivating the problem NL2Fix of translating natural
language descriptions of code changes (namely bug fixes described in Issue
reports in repositories) into correct code fixes. To this end, we introduce
Defects4J-NL2Fix, a dataset of 283 Java programs from the popular Defects4J
dataset augmented with high-level descriptions of bug fixes, and empirically
evaluate the performance of several state-of-the-art LLMs for the this task.
Results show that these LLMS together are capable of generating plausible fixes
for 64.6% of the bugs, and the best LLM-based technique can achieve up to
21.20% top-1 and 35.68% top-5 accuracy on this benchmark.
|
2304.03844 | Zhenghang Yuan | Zhenghang Yuan, Lichao Mou, and Xiao Xiang Zhu | Multilingual Augmentation for Robust Visual Question Answering in Remote
Sensing Images | This paper was submitted to the JURSE 2023 conference on November 5,
2022 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Aiming at answering questions based on the content of remotely sensed images,
visual question answering for remote sensing data (RSVQA) has attracted much
attention nowadays. However, previous works in RSVQA have focused little on the
robustness of RSVQA. As we aim to enhance the reliability of RSVQA models, how
to learn robust representations against new words and different question
templates with the same meaning is the key challenge. With the proposed
augmented dataset, we are able to obtain more questions in addition to the
original ones with the same meaning. To make better use of this information, in
this study, we propose a contrastive learning strategy for training robust
RSVQA models against diverse question templates and words. Experimental results
demonstrate that the proposed augmented dataset is effective in improving the
robustness of the RSVQA model. In addition, the contrastive learning strategy
performs well on the low resolution (LR) dataset.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 21:06:58 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Yuan",
"Zhenghang",
""
],
[
"Mou",
"Lichao",
""
],
[
"Zhu",
"Xiao Xiang",
""
]
] | TITLE: Multilingual Augmentation for Robust Visual Question Answering in Remote
Sensing Images
ABSTRACT: Aiming at answering questions based on the content of remotely sensed images,
visual question answering for remote sensing data (RSVQA) has attracted much
attention nowadays. However, previous works in RSVQA have focused little on the
robustness of RSVQA. As we aim to enhance the reliability of RSVQA models, how
to learn robust representations against new words and different question
templates with the same meaning is the key challenge. With the proposed
augmented dataset, we are able to obtain more questions in addition to the
original ones with the same meaning. To make better use of this information, in
this study, we propose a contrastive learning strategy for training robust
RSVQA models against diverse question templates and words. Experimental results
demonstrate that the proposed augmented dataset is effective in improving the
robustness of the RSVQA model. In addition, the contrastive learning strategy
performs well on the low resolution (LR) dataset.
|
2304.03854 | Kevin Cao | Kevin Cao, Kevin Leach | Revisiting Deep Learning for Variable Type Recovery | In The 31st International Conference on Program Comprehension(ICPC
2023 RENE) | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Compiled binary executables are often the only available artifact in reverse
engineering, malware analysis, and software systems maintenance. Unfortunately,
the lack of semantic information like variable types makes comprehending
binaries difficult. In efforts to improve the comprehensibility of binaries,
researchers have recently used machine learning techniques to predict semantic
information contained in the original source code. Chen et al. implemented
DIRTY, a Transformer-based Encoder-Decoder architecture capable of augmenting
decompiled code with variable names and types by leveraging decompiler output
tokens and variable size information. Chen et al. were able to demonstrate a
substantial increase in name and type extraction accuracy on Hex-Rays
decompiler outputs compared to existing static analysis and AI-based
techniques. We extend the original DIRTY results by re-training the DIRTY model
on a dataset produced by the open-source Ghidra decompiler. Although Chen et
al. concluded that Ghidra was not a suitable decompiler candidate due to its
difficulty in parsing and incorporating DWARF symbols during analysis, we
demonstrate that straightforward parsing of variable data generated by Ghidra
results in similar retyping performance. We hope this work inspires further
interest and adoption of the Ghidra decompiler for use in research projects.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 22:28:28 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Cao",
"Kevin",
""
],
[
"Leach",
"Kevin",
""
]
] | TITLE: Revisiting Deep Learning for Variable Type Recovery
ABSTRACT: Compiled binary executables are often the only available artifact in reverse
engineering, malware analysis, and software systems maintenance. Unfortunately,
the lack of semantic information like variable types makes comprehending
binaries difficult. In efforts to improve the comprehensibility of binaries,
researchers have recently used machine learning techniques to predict semantic
information contained in the original source code. Chen et al. implemented
DIRTY, a Transformer-based Encoder-Decoder architecture capable of augmenting
decompiled code with variable names and types by leveraging decompiler output
tokens and variable size information. Chen et al. were able to demonstrate a
substantial increase in name and type extraction accuracy on Hex-Rays
decompiler outputs compared to existing static analysis and AI-based
techniques. We extend the original DIRTY results by re-training the DIRTY model
on a dataset produced by the open-source Ghidra decompiler. Although Chen et
al. concluded that Ghidra was not a suitable decompiler candidate due to its
difficulty in parsing and incorporating DWARF symbols during analysis, we
demonstrate that straightforward parsing of variable data generated by Ghidra
results in similar retyping performance. We hope this work inspires further
interest and adoption of the Ghidra decompiler for use in research projects.
|
2304.03867 | Sridhar Sola Mr. | Sridhar Sola and Darshan Gera | Masked Student Dataset of Expressions | Thirteenth Indian Conference on Computer Vision, Graphics and Image
Processing, ACM, 2022, Gandhinagar, India | null | null | null | cs.CV cs.HC | http://creativecommons.org/licenses/by/4.0/ | Facial expression recognition (FER) algorithms work well in constrained
environments with little or no occlusion of the face. However, real-world face
occlusion is prevalent, most notably with the need to use a face mask in the
current Covid-19 scenario. While there are works on the problem of occlusion in
FER, little has been done before on the particular face mask scenario.
Moreover, the few works in this area largely use synthetically created masked
FER datasets. Motivated by these challenges posed by the pandemic to FER, we
present a novel dataset, the Masked Student Dataset of Expressions or MSD-E,
consisting of 1,960 real-world non-masked and masked facial expression images
collected from 142 individuals. Along with the issue of obfuscated facial
features, we illustrate how other subtler issues in masked FER are represented
in our dataset. We then provide baseline results using ResNet-18, finding that
its performance dips in the non-masked case when trained for FER in the
presence of masks. To tackle this, we test two training paradigms: contrastive
learning and knowledge distillation, and find that they increase the model's
performance in the masked scenario while maintaining its non-masked
performance. We further visualise our results using t-SNE plots and Grad-CAM,
demonstrating that these paradigms capitalise on the limited features available
in the masked scenario. Finally, we benchmark SOTA methods on MSD-E.
| [
{
"version": "v1",
"created": "Fri, 7 Apr 2023 23:43:21 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Sola",
"Sridhar",
""
],
[
"Gera",
"Darshan",
""
]
] | TITLE: Masked Student Dataset of Expressions
ABSTRACT: Facial expression recognition (FER) algorithms work well in constrained
environments with little or no occlusion of the face. However, real-world face
occlusion is prevalent, most notably with the need to use a face mask in the
current Covid-19 scenario. While there are works on the problem of occlusion in
FER, little has been done before on the particular face mask scenario.
Moreover, the few works in this area largely use synthetically created masked
FER datasets. Motivated by these challenges posed by the pandemic to FER, we
present a novel dataset, the Masked Student Dataset of Expressions or MSD-E,
consisting of 1,960 real-world non-masked and masked facial expression images
collected from 142 individuals. Along with the issue of obfuscated facial
features, we illustrate how other subtler issues in masked FER are represented
in our dataset. We then provide baseline results using ResNet-18, finding that
its performance dips in the non-masked case when trained for FER in the
presence of masks. To tackle this, we test two training paradigms: contrastive
learning and knowledge distillation, and find that they increase the model's
performance in the masked scenario while maintaining its non-masked
performance. We further visualise our results using t-SNE plots and Grad-CAM,
demonstrating that these paradigms capitalise on the limited features available
in the masked scenario. Finally, we benchmark SOTA methods on MSD-E.
|
2304.03879 | Jinming Li | Jinming Li, Wentao Zhang, Tian Wang, Guanglei Xiong, Alan Lu, Gerard
Medioni | GPT4Rec: A Generative Framework for Personalized Recommendation and User
Interests Interpretation | null | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in Natural Language Processing (NLP) have led to the
development of NLP-based recommender systems that have shown superior
performance. However, current models commonly treat items as mere IDs and adopt
discriminative modeling, resulting in limitations of (1) fully leveraging the
content information of items and the language modeling capabilities of NLP
models; (2) interpreting user interests to improve relevance and diversity; and
(3) adapting practical circumstances such as growing item inventories. To
address these limitations, we present GPT4Rec, a novel and flexible generative
framework inspired by search engines. It first generates hypothetical "search
queries" given item titles in a user's history, and then retrieves items for
recommendation by searching these queries. The framework overcomes previous
limitations by learning both user and item embeddings in the language space. To
well-capture user interests with different aspects and granularity for
improving relevance and diversity, we propose a multi-query generation
technique with beam search. The generated queries naturally serve as
interpretable representations of user interests and can be searched to
recommend cold-start items. With GPT-2 language model and BM25 search engine,
our framework outperforms state-of-the-art methods by $75.7\%$ and $22.2\%$ in
Recall@K on two public datasets. Experiments further revealed that multi-query
generation with beam search improves both the diversity of retrieved items and
the coverage of a user's multi-interests. The adaptiveness and interpretability
of generated queries are discussed with qualitative case studies.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 00:30:08 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Li",
"Jinming",
""
],
[
"Zhang",
"Wentao",
""
],
[
"Wang",
"Tian",
""
],
[
"Xiong",
"Guanglei",
""
],
[
"Lu",
"Alan",
""
],
[
"Medioni",
"Gerard",
""
]
] | TITLE: GPT4Rec: A Generative Framework for Personalized Recommendation and User
Interests Interpretation
ABSTRACT: Recent advancements in Natural Language Processing (NLP) have led to the
development of NLP-based recommender systems that have shown superior
performance. However, current models commonly treat items as mere IDs and adopt
discriminative modeling, resulting in limitations of (1) fully leveraging the
content information of items and the language modeling capabilities of NLP
models; (2) interpreting user interests to improve relevance and diversity; and
(3) adapting practical circumstances such as growing item inventories. To
address these limitations, we present GPT4Rec, a novel and flexible generative
framework inspired by search engines. It first generates hypothetical "search
queries" given item titles in a user's history, and then retrieves items for
recommendation by searching these queries. The framework overcomes previous
limitations by learning both user and item embeddings in the language space. To
well-capture user interests with different aspects and granularity for
improving relevance and diversity, we propose a multi-query generation
technique with beam search. The generated queries naturally serve as
interpretable representations of user interests and can be searched to
recommend cold-start items. With GPT-2 language model and BM25 search engine,
our framework outperforms state-of-the-art methods by $75.7\%$ and $22.2\%$ in
Recall@K on two public datasets. Experiments further revealed that multi-query
generation with beam search improves both the diversity of retrieved items and
the coverage of a user's multi-interests. The adaptiveness and interpretability
of generated queries are discussed with qualitative case studies.
|
2304.03891 | Jiangxia Cao | Jiangxia Cao, Xin Cong, Jiawei Sheng, Tingwen Liu, Bin Wang | Contrastive Cross-Domain Sequential Recommendation | This paper has been accepted by CIKM 2022 | null | 10.1145/3511808.3557262 | null | cs.IR cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Cross-Domain Sequential Recommendation (CDSR) aims to predict future
interactions based on user's historical sequential interactions from multiple
domains. Generally, a key challenge of CDSR is how to mine precise cross-domain
user preference based on the intra-sequence and inter-sequence item
interactions. Existing works first learn single-domain user preference only
with intra-sequence item interactions, and then build a transferring module to
obtain cross-domain user preference. However, such a pipeline and implicit
solution can be severely limited by the bottleneck of the designed transferring
module, and ignores to consider inter-sequence item relationships. In this
paper, we propose C^2DSR to tackle the above problems to capture precise user
preferences. The main idea is to simultaneously leverage the intra- and inter-
sequence item relationships, and jointly learn the single- and cross- domain
user preferences. Specifically, we first utilize a graph neural network to mine
inter-sequence item collaborative relationship, and then exploit sequential
attentive encoder to capture intra-sequence item sequential relationship. Based
on them, we devise two different sequential training objectives to obtain user
single-domain and cross-domain representations. Furthermore, we present a novel
contrastive cross-domain infomax objective to enhance the correlation between
single- and cross- domain user representations by maximizing their mutual
information. To validate the effectiveness of C^2DSR, we first re-split four
e-comerce datasets, and then conduct extensive experiments to demonstrate the
effectiveness of our approach C^2DSR.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 02:19:40 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Cao",
"Jiangxia",
""
],
[
"Cong",
"Xin",
""
],
[
"Sheng",
"Jiawei",
""
],
[
"Liu",
"Tingwen",
""
],
[
"Wang",
"Bin",
""
]
] | TITLE: Contrastive Cross-Domain Sequential Recommendation
ABSTRACT: Cross-Domain Sequential Recommendation (CDSR) aims to predict future
interactions based on user's historical sequential interactions from multiple
domains. Generally, a key challenge of CDSR is how to mine precise cross-domain
user preference based on the intra-sequence and inter-sequence item
interactions. Existing works first learn single-domain user preference only
with intra-sequence item interactions, and then build a transferring module to
obtain cross-domain user preference. However, such a pipeline and implicit
solution can be severely limited by the bottleneck of the designed transferring
module, and ignores to consider inter-sequence item relationships. In this
paper, we propose C^2DSR to tackle the above problems to capture precise user
preferences. The main idea is to simultaneously leverage the intra- and inter-
sequence item relationships, and jointly learn the single- and cross- domain
user preferences. Specifically, we first utilize a graph neural network to mine
inter-sequence item collaborative relationship, and then exploit sequential
attentive encoder to capture intra-sequence item sequential relationship. Based
on them, we devise two different sequential training objectives to obtain user
single-domain and cross-domain representations. Furthermore, we present a novel
contrastive cross-domain infomax objective to enhance the correlation between
single- and cross- domain user representations by maximizing their mutual
information. To validate the effectiveness of C^2DSR, we first re-split four
e-comerce datasets, and then conduct extensive experiments to demonstrate the
effectiveness of our approach C^2DSR.
|
2304.03899 | Zhen Wu | Zhen Wu, Yizhe Lu, Xinyu Dai | An Empirical Study and Improvement for Speech Emotion Recognition | Accepted by ICASSP 2023 | null | null | null | cs.CL cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal speech emotion recognition aims to detect speakers' emotions from
audio and text. Prior works mainly focus on exploiting advanced networks to
model and fuse different modality information to facilitate performance, while
neglecting the effect of different fusion strategies on emotion recognition. In
this work, we consider a simple yet important problem: how to fuse audio and
text modality information is more helpful for this multimodal task. Further, we
propose a multimodal emotion recognition model improved by perspective loss.
Empirical results show our method obtained new state-of-the-art results on the
IEMOCAP dataset. The in-depth analysis explains why the improved model can
achieve improvements and outperforms baselines.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 03:24:06 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Wu",
"Zhen",
""
],
[
"Lu",
"Yizhe",
""
],
[
"Dai",
"Xinyu",
""
]
] | TITLE: An Empirical Study and Improvement for Speech Emotion Recognition
ABSTRACT: Multimodal speech emotion recognition aims to detect speakers' emotions from
audio and text. Prior works mainly focus on exploiting advanced networks to
model and fuse different modality information to facilitate performance, while
neglecting the effect of different fusion strategies on emotion recognition. In
this work, we consider a simple yet important problem: how to fuse audio and
text modality information is more helpful for this multimodal task. Further, we
propose a multimodal emotion recognition model improved by perspective loss.
Empirical results show our method obtained new state-of-the-art results on the
IEMOCAP dataset. The in-depth analysis explains why the improved model can
achieve improvements and outperforms baselines.
|
2304.03903 | Tignting Liao | Tingting Liao and Xiaomei Zhang and Yuliang Xiu and Hongwei Yi and
Xudong Liu and Guo-Jun Qi and Yong Zhang and Xuan Wang and Xiangyu Zhu and
Zhen Lei | High-Fidelity Clothed Avatar Reconstruction from a Single Image | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents a framework for efficient 3D clothed avatar
reconstruction. By combining the advantages of the high accuracy of
optimization-based methods and the efficiency of learning-based methods, we
propose a coarse-to-fine way to realize a high-fidelity clothed avatar
reconstruction (CAR) from a single image. At the first stage, we use an
implicit model to learn the general shape in the canonical space of a person in
a learning-based way, and at the second stage, we refine the surface detail by
estimating the non-rigid deformation in the posed space in an optimization way.
A hyper-network is utilized to generate a good initialization so that the
convergence o f the optimization process is greatly accelerated. Extensive
experiments on various datasets show that the proposed CAR successfully
produces high-fidelity avatars for arbitrarily clothed humans in real scenes.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 04:01:04 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Liao",
"Tingting",
""
],
[
"Zhang",
"Xiaomei",
""
],
[
"Xiu",
"Yuliang",
""
],
[
"Yi",
"Hongwei",
""
],
[
"Liu",
"Xudong",
""
],
[
"Qi",
"Guo-Jun",
""
],
[
"Zhang",
"Yong",
""
],
[
"Wang",
"Xuan",
""
],
[
"Zhu",
"Xiangyu",
""
],
[
"Lei",
"Zhen",
""
]
] | TITLE: High-Fidelity Clothed Avatar Reconstruction from a Single Image
ABSTRACT: This paper presents a framework for efficient 3D clothed avatar
reconstruction. By combining the advantages of the high accuracy of
optimization-based methods and the efficiency of learning-based methods, we
propose a coarse-to-fine way to realize a high-fidelity clothed avatar
reconstruction (CAR) from a single image. At the first stage, we use an
implicit model to learn the general shape in the canonical space of a person in
a learning-based way, and at the second stage, we refine the surface detail by
estimating the non-rigid deformation in the posed space in an optimization way.
A hyper-network is utilized to generate a good initialization so that the
convergence o f the optimization process is greatly accelerated. Extensive
experiments on various datasets show that the proposed CAR successfully
produces high-fidelity avatars for arbitrarily clothed humans in real scenes.
|
2304.03941 | Miguel Xochicale | Michelle Iskandar, Harvey Mannering, Zhanxiang Sun, Jacqueline
Matthew, Hamideh Kerdegari, Laura Peralta, Miguel Xochicale | Towards Realistic Ultrasound Fetal Brain Imaging Synthesis | 3 pages, 1 figure | null | null | null | eess.IV cs.CV cs.LG physics.med-ph | http://creativecommons.org/licenses/by-sa/4.0/ | Prenatal ultrasound imaging is the first-choice modality to assess fetal
health. Medical image datasets for AI and ML methods must be diverse (i.e.
diagnoses, diseases, pathologies, scanners, demographics, etc), however there
are few public ultrasound fetal imaging datasets due to insufficient amounts of
clinical data, patient privacy, rare occurrence of abnormalities in general
practice, and limited experts for data collection and validation. To address
such data scarcity, we proposed generative adversarial networks (GAN)-based
models, diffusion-super-resolution-GAN and transformer-based-GAN, to synthesise
images of fetal ultrasound brain planes from one public dataset. We reported
that GAN-based methods can generate 256x256 pixel size of fetal ultrasound
trans-cerebellum brain image plane with stable training losses, resulting in
lower FID values for diffusion-super-resolution-GAN (average 7.04 and lower FID
5.09 at epoch 10) than the FID values of transformer-based-GAN (average 36.02
and lower 28.93 at epoch 60). The results of this work illustrate the potential
of GAN-based methods to synthesise realistic high-resolution ultrasound images,
leading to future work with other fetal brain planes, anatomies, devices and
the need of a pool of experts to evaluate synthesised images. Code, data and
other resources to reproduce this work are available at
\url{https://github.com/budai4medtech/midl2023}.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 07:07:20 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Iskandar",
"Michelle",
""
],
[
"Mannering",
"Harvey",
""
],
[
"Sun",
"Zhanxiang",
""
],
[
"Matthew",
"Jacqueline",
""
],
[
"Kerdegari",
"Hamideh",
""
],
[
"Peralta",
"Laura",
""
],
[
"Xochicale",
"Miguel",
""
]
] | TITLE: Towards Realistic Ultrasound Fetal Brain Imaging Synthesis
ABSTRACT: Prenatal ultrasound imaging is the first-choice modality to assess fetal
health. Medical image datasets for AI and ML methods must be diverse (i.e.
diagnoses, diseases, pathologies, scanners, demographics, etc), however there
are few public ultrasound fetal imaging datasets due to insufficient amounts of
clinical data, patient privacy, rare occurrence of abnormalities in general
practice, and limited experts for data collection and validation. To address
such data scarcity, we proposed generative adversarial networks (GAN)-based
models, diffusion-super-resolution-GAN and transformer-based-GAN, to synthesise
images of fetal ultrasound brain planes from one public dataset. We reported
that GAN-based methods can generate 256x256 pixel size of fetal ultrasound
trans-cerebellum brain image plane with stable training losses, resulting in
lower FID values for diffusion-super-resolution-GAN (average 7.04 and lower FID
5.09 at epoch 10) than the FID values of transformer-based-GAN (average 36.02
and lower 28.93 at epoch 60). The results of this work illustrate the potential
of GAN-based methods to synthesise realistic high-resolution ultrasound images,
leading to future work with other fetal brain planes, anatomies, devices and
the need of a pool of experts to evaluate synthesised images. Code, data and
other resources to reproduce this work are available at
\url{https://github.com/budai4medtech/midl2023}.
|
2304.03945 | Zhifeng Wang | Linqing Li, Zhifeng Wang | Knowledge Relation Rank Enhanced Heterogeneous Learning Interaction
Modeling for Neural Graph Forgetting Knowledge Tracing | 11 pages, 3 figures | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, knowledge tracing models have been applied in educational data
mining such as the Self-attention knowledge tracing model(SAKT), which models
the relationship between exercises and Knowledge concepts(Kcs). However,
relation modeling in traditional Knowledge tracing models only considers the
static question-knowledge relationship and knowledge-knowledge relationship and
treats these relationships with equal importance. This kind of relation
modeling is difficult to avoid the influence of subjective labeling and
considers the relationship between exercises and KCs, or KCs and KCs
separately. In this work, a novel knowledge tracing model, named Knowledge
Relation Rank Enhanced Heterogeneous Learning Interaction Modeling for Neural
Graph Forgetting Knowledge Tracing(NGFKT), is proposed to reduce the impact of
the subjective labeling by calibrating the skill relation matrix and the
Q-matrix and apply the Graph Convolutional Network(GCN) to model the
heterogeneous interactions between students, exercises, and skills.
Specifically, the skill relation matrix and Q-matrix are generated by the
Knowledge Relation Importance Rank Calibration method(KRIRC). Then the
calibrated skill relation matrix, Q-matrix, and the heterogeneous interactions
are treated as the input of the GCN to generate the exercise embedding and
skill embedding. Next, the exercise embedding, skill embedding, item
difficulty, and contingency table are incorporated to generate an exercise
relation matrix as the inputs of the Position-Relation-Forgetting attention
mechanism. Finally, the Position-Relation-Forgetting attention mechanism is
applied to make the predictions. Experiments are conducted on the two public
educational datasets and results indicate that the NGFKT model outperforms all
baseline models in terms of AUC, ACC, and Performance Stability(PS).
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 07:29:53 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Li",
"Linqing",
""
],
[
"Wang",
"Zhifeng",
""
]
] | TITLE: Knowledge Relation Rank Enhanced Heterogeneous Learning Interaction
Modeling for Neural Graph Forgetting Knowledge Tracing
ABSTRACT: Recently, knowledge tracing models have been applied in educational data
mining such as the Self-attention knowledge tracing model(SAKT), which models
the relationship between exercises and Knowledge concepts(Kcs). However,
relation modeling in traditional Knowledge tracing models only considers the
static question-knowledge relationship and knowledge-knowledge relationship and
treats these relationships with equal importance. This kind of relation
modeling is difficult to avoid the influence of subjective labeling and
considers the relationship between exercises and KCs, or KCs and KCs
separately. In this work, a novel knowledge tracing model, named Knowledge
Relation Rank Enhanced Heterogeneous Learning Interaction Modeling for Neural
Graph Forgetting Knowledge Tracing(NGFKT), is proposed to reduce the impact of
the subjective labeling by calibrating the skill relation matrix and the
Q-matrix and apply the Graph Convolutional Network(GCN) to model the
heterogeneous interactions between students, exercises, and skills.
Specifically, the skill relation matrix and Q-matrix are generated by the
Knowledge Relation Importance Rank Calibration method(KRIRC). Then the
calibrated skill relation matrix, Q-matrix, and the heterogeneous interactions
are treated as the input of the GCN to generate the exercise embedding and
skill embedding. Next, the exercise embedding, skill embedding, item
difficulty, and contingency table are incorporated to generate an exercise
relation matrix as the inputs of the Position-Relation-Forgetting attention
mechanism. Finally, the Position-Relation-Forgetting attention mechanism is
applied to make the predictions. Experiments are conducted on the two public
educational datasets and results indicate that the NGFKT model outperforms all
baseline models in terms of AUC, ACC, and Performance Stability(PS).
|
2304.03947 | Jing Long | Jing Long, Tong Chen, Nguyen Quoc Viet Hung, Guandong Xu, Kai Zheng,
Hongzhi Yin | Model-Agnostic Decentralized Collaborative Learning for On-Device POI
Recommendation | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | As an indispensable personalized service in Location-based Social Networks
(LBSNs), the next Point-of-Interest (POI) recommendation aims to help people
discover attractive and interesting places. Currently, most POI recommenders
are based on the conventional centralized paradigm that heavily relies on the
cloud to train the recommendation models with large volumes of collected users'
sensitive check-in data. Although a few recent works have explored on-device
frameworks for resilient and privacy-preserving POI recommendations, they
invariably hold the assumption of model homogeneity for parameters/gradients
aggregation and collaboration. However, users' mobile devices in the real world
have various hardware configurations (e.g., compute resources), leading to
heterogeneous on-device models with different architectures and sizes. In light
of this, We propose a novel on-device POI recommendation framework, namely
Model-Agnostic Collaborative learning for on-device POI recommendation (MAC),
allowing users to customize their own model structures (e.g., dimension \&
number of hidden layers). To counteract the sparsity of on-device user data, we
propose to pre-select neighbors for collaboration based on physical distances,
category-level preferences, and social networks. To assimilate knowledge from
the above-selected neighbors in an efficient and secure way, we adopt the
knowledge distillation framework with mutual information maximization. Instead
of sharing sensitive models/gradients, clients in MAC only share their soft
decisions on a preloaded reference dataset. To filter out low-quality
neighbors, we propose two sampling strategies, performance-triggered sampling
and similarity-based sampling, to speed up the training process and obtain
optimal recommenders. In addition, we design two novel approaches to generate
more effective reference datasets while protecting users' privacy.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 07:36:37 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Long",
"Jing",
""
],
[
"Chen",
"Tong",
""
],
[
"Hung",
"Nguyen Quoc Viet",
""
],
[
"Xu",
"Guandong",
""
],
[
"Zheng",
"Kai",
""
],
[
"Yin",
"Hongzhi",
""
]
] | TITLE: Model-Agnostic Decentralized Collaborative Learning for On-Device POI
Recommendation
ABSTRACT: As an indispensable personalized service in Location-based Social Networks
(LBSNs), the next Point-of-Interest (POI) recommendation aims to help people
discover attractive and interesting places. Currently, most POI recommenders
are based on the conventional centralized paradigm that heavily relies on the
cloud to train the recommendation models with large volumes of collected users'
sensitive check-in data. Although a few recent works have explored on-device
frameworks for resilient and privacy-preserving POI recommendations, they
invariably hold the assumption of model homogeneity for parameters/gradients
aggregation and collaboration. However, users' mobile devices in the real world
have various hardware configurations (e.g., compute resources), leading to
heterogeneous on-device models with different architectures and sizes. In light
of this, We propose a novel on-device POI recommendation framework, namely
Model-Agnostic Collaborative learning for on-device POI recommendation (MAC),
allowing users to customize their own model structures (e.g., dimension \&
number of hidden layers). To counteract the sparsity of on-device user data, we
propose to pre-select neighbors for collaboration based on physical distances,
category-level preferences, and social networks. To assimilate knowledge from
the above-selected neighbors in an efficient and secure way, we adopt the
knowledge distillation framework with mutual information maximization. Instead
of sharing sensitive models/gradients, clients in MAC only share their soft
decisions on a preloaded reference dataset. To filter out low-quality
neighbors, we propose two sampling strategies, performance-triggered sampling
and similarity-based sampling, to speed up the training process and obtain
optimal recommenders. In addition, we design two novel approaches to generate
more effective reference datasets while protecting users' privacy.
|
2304.03952 | Rendani Mbuvha | Rendani Mbuvha, David I. Adelani, Tendani Mutavhatsindi, Tshimangadzo
Rakhuhu, Aluwani Mauda, Tshifhiwa Joshua Maumela, Andisani Masindi, Seani
Rananga, Vukosi Marivate, and Tshilidzi Marwala | MphayaNER: Named Entity Recognition for Tshivenda | Accepted at AfricaNLP Workshop at ICLR 2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Named Entity Recognition (NER) plays a vital role in various Natural Language
Processing tasks such as information retrieval, text classification, and
question answering. However, NER can be challenging, especially in low-resource
languages with limited annotated datasets and tools. This paper adds to the
effort of addressing these challenges by introducing MphayaNER, the first
Tshivenda NER corpus in the news domain. We establish NER baselines by
\textit{fine-tuning} state-of-the-art models on MphayaNER. The study also
explores zero-shot transfer between Tshivenda and other related Bantu
languages, with chiShona and Kiswahili showing the best results. Augmenting
MphayaNER with chiShona data was also found to improve model performance
significantly. Both MphayaNER and the baseline models are made publicly
available.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 08:03:58 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Mbuvha",
"Rendani",
""
],
[
"Adelani",
"David I.",
""
],
[
"Mutavhatsindi",
"Tendani",
""
],
[
"Rakhuhu",
"Tshimangadzo",
""
],
[
"Mauda",
"Aluwani",
""
],
[
"Maumela",
"Tshifhiwa Joshua",
""
],
[
"Masindi",
"Andisani",
""
],
[
"Rananga",
"Seani",
""
],
[
"Marivate",
"Vukosi",
""
],
[
"Marwala",
"Tshilidzi",
""
]
] | TITLE: MphayaNER: Named Entity Recognition for Tshivenda
ABSTRACT: Named Entity Recognition (NER) plays a vital role in various Natural Language
Processing tasks such as information retrieval, text classification, and
question answering. However, NER can be challenging, especially in low-resource
languages with limited annotated datasets and tools. This paper adds to the
effort of addressing these challenges by introducing MphayaNER, the first
Tshivenda NER corpus in the news domain. We establish NER baselines by
\textit{fine-tuning} state-of-the-art models on MphayaNER. The study also
explores zero-shot transfer between Tshivenda and other related Bantu
languages, with chiShona and Kiswahili showing the best results. Augmenting
MphayaNER with chiShona data was also found to improve model performance
significantly. Both MphayaNER and the baseline models are made publicly
available.
|
2304.03955 | Dashan Gao | Dashan Gao and Yunce Zhao and Yinghua Yao and Zeqi Zhang and Bifei Mao
and Xin Yao | Robust Deep Learning Models Against Semantic-Preserving Adversarial
Attack | Paper accepted by the 2023 International Joint Conference on Neural
Networks (IJCNN 2023) | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | Deep learning models can be fooled by small $l_p$-norm adversarial
perturbations and natural perturbations in terms of attributes. Although the
robustness against each perturbation has been explored, it remains a challenge
to address the robustness against joint perturbations effectively. In this
paper, we study the robustness of deep learning models against joint
perturbations by proposing a novel attack mechanism named Semantic-Preserving
Adversarial (SPA) attack, which can then be used to enhance adversarial
training. Specifically, we introduce an attribute manipulator to generate
natural and human-comprehensible perturbations and a noise generator to
generate diverse adversarial noises. Based on such combined noises, we optimize
both the attribute value and the diversity variable to generate
jointly-perturbed samples. For robust training, we adversarially train the deep
learning model against the generated joint perturbations. Empirical results on
four benchmarks show that the SPA attack causes a larger performance decline
with small $l_{\infty}$ norm-ball constraints compared to existing approaches.
Furthermore, our SPA-enhanced training outperforms existing defense methods
against such joint perturbations.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 08:28:36 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Gao",
"Dashan",
""
],
[
"Zhao",
"Yunce",
""
],
[
"Yao",
"Yinghua",
""
],
[
"Zhang",
"Zeqi",
""
],
[
"Mao",
"Bifei",
""
],
[
"Yao",
"Xin",
""
]
] | TITLE: Robust Deep Learning Models Against Semantic-Preserving Adversarial
Attack
ABSTRACT: Deep learning models can be fooled by small $l_p$-norm adversarial
perturbations and natural perturbations in terms of attributes. Although the
robustness against each perturbation has been explored, it remains a challenge
to address the robustness against joint perturbations effectively. In this
paper, we study the robustness of deep learning models against joint
perturbations by proposing a novel attack mechanism named Semantic-Preserving
Adversarial (SPA) attack, which can then be used to enhance adversarial
training. Specifically, we introduce an attribute manipulator to generate
natural and human-comprehensible perturbations and a noise generator to
generate diverse adversarial noises. Based on such combined noises, we optimize
both the attribute value and the diversity variable to generate
jointly-perturbed samples. For robust training, we adversarially train the deep
learning model against the generated joint perturbations. Empirical results on
four benchmarks show that the SPA attack causes a larger performance decline
with small $l_{\infty}$ norm-ball constraints compared to existing approaches.
Furthermore, our SPA-enhanced training outperforms existing defense methods
against such joint perturbations.
|
2304.03977 | Shengbang Tong | Shengbang Tong, Yubei Chen, Yi Ma, Yann Lecun | EMP-SSL: Towards Self-Supervised Learning in One Training Epoch | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, self-supervised learning (SSL) has achieved tremendous success in
learning image representation. Despite the empirical success, most
self-supervised learning methods are rather "inefficient" learners, typically
taking hundreds of training epochs to fully converge. In this work, we show
that the key towards efficient self-supervised learning is to increase the
number of crops from each image instance. Leveraging one of the
state-of-the-art SSL method, we introduce a simplistic form of self-supervised
learning method called Extreme-Multi-Patch Self-Supervised-Learning (EMP-SSL)
that does not rely on many heuristic techniques for SSL such as weight sharing
between the branches, feature-wise normalization, output quantization, and stop
gradient, etc, and reduces the training epochs by two orders of magnitude. We
show that the proposed method is able to converge to 85.1% on CIFAR-10, 58.5%
on CIFAR-100, 38.1% on Tiny ImageNet and 58.5% on ImageNet-100 in just one
epoch. Furthermore, the proposed method achieves 91.5% on CIFAR-10, 70.1% on
CIFAR-100, 51.5% on Tiny ImageNet and 78.9% on ImageNet-100 with linear probing
in less than ten training epochs. In addition, we show that EMP-SSL shows
significantly better transferability to out-of-domain datasets compared to
baseline SSL methods. We will release the code in
https://github.com/tsb0601/EMP-SSL.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 10:09:30 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Tong",
"Shengbang",
""
],
[
"Chen",
"Yubei",
""
],
[
"Ma",
"Yi",
""
],
[
"Lecun",
"Yann",
""
]
] | TITLE: EMP-SSL: Towards Self-Supervised Learning in One Training Epoch
ABSTRACT: Recently, self-supervised learning (SSL) has achieved tremendous success in
learning image representation. Despite the empirical success, most
self-supervised learning methods are rather "inefficient" learners, typically
taking hundreds of training epochs to fully converge. In this work, we show
that the key towards efficient self-supervised learning is to increase the
number of crops from each image instance. Leveraging one of the
state-of-the-art SSL method, we introduce a simplistic form of self-supervised
learning method called Extreme-Multi-Patch Self-Supervised-Learning (EMP-SSL)
that does not rely on many heuristic techniques for SSL such as weight sharing
between the branches, feature-wise normalization, output quantization, and stop
gradient, etc, and reduces the training epochs by two orders of magnitude. We
show that the proposed method is able to converge to 85.1% on CIFAR-10, 58.5%
on CIFAR-100, 38.1% on Tiny ImageNet and 58.5% on ImageNet-100 in just one
epoch. Furthermore, the proposed method achieves 91.5% on CIFAR-10, 70.1% on
CIFAR-100, 51.5% on Tiny ImageNet and 78.9% on ImageNet-100 with linear probing
in less than ten training epochs. In addition, we show that EMP-SSL shows
significantly better transferability to out-of-domain datasets compared to
baseline SSL methods. We will release the code in
https://github.com/tsb0601/EMP-SSL.
|
2304.03984 | Shangfei Zheng | Shangfei Zheng, Hongzhi Yin, Tong Chen, Quoc Viet Hung Nguyen, Wei
Chen, Lei Zhao | DREAM: Adaptive Reinforcement Learning based on Attention Mechanism for
Temporal Knowledge Graph Reasoning | 11 pages | null | null | null | cs.AI cs.IR | http://creativecommons.org/licenses/by/4.0/ | Temporal knowledge graphs (TKGs) model the temporal evolution of events and
have recently attracted increasing attention. Since TKGs are intrinsically
incomplete, it is necessary to reason out missing elements. Although existing
TKG reasoning methods have the ability to predict missing future events, they
fail to generate explicit reasoning paths and lack explainability. As
reinforcement learning (RL) for multi-hop reasoning on traditional knowledge
graphs starts showing superior explainability and performance in recent
advances, it has opened up opportunities for exploring RL techniques on TKG
reasoning. However, the performance of RL-based TKG reasoning methods is
limited due to: (1) lack of ability to capture temporal evolution and semantic
dependence jointly; (2) excessive reliance on manually designed rewards. To
overcome these challenges, we propose an adaptive reinforcement learning model
based on attention mechanism (DREAM) to predict missing elements in the future.
Specifically, the model contains two components: (1) a multi-faceted attention
representation learning method that captures semantic dependence and temporal
evolution jointly; (2) an adaptive RL framework that conducts multi-hop
reasoning by adaptively learning the reward functions. Experimental results
demonstrate DREAM outperforms state-of-the-art models on public dataset
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 10:57:37 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Zheng",
"Shangfei",
""
],
[
"Yin",
"Hongzhi",
""
],
[
"Chen",
"Tong",
""
],
[
"Nguyen",
"Quoc Viet Hung",
""
],
[
"Chen",
"Wei",
""
],
[
"Zhao",
"Lei",
""
]
] | TITLE: DREAM: Adaptive Reinforcement Learning based on Attention Mechanism for
Temporal Knowledge Graph Reasoning
ABSTRACT: Temporal knowledge graphs (TKGs) model the temporal evolution of events and
have recently attracted increasing attention. Since TKGs are intrinsically
incomplete, it is necessary to reason out missing elements. Although existing
TKG reasoning methods have the ability to predict missing future events, they
fail to generate explicit reasoning paths and lack explainability. As
reinforcement learning (RL) for multi-hop reasoning on traditional knowledge
graphs starts showing superior explainability and performance in recent
advances, it has opened up opportunities for exploring RL techniques on TKG
reasoning. However, the performance of RL-based TKG reasoning methods is
limited due to: (1) lack of ability to capture temporal evolution and semantic
dependence jointly; (2) excessive reliance on manually designed rewards. To
overcome these challenges, we propose an adaptive reinforcement learning model
based on attention mechanism (DREAM) to predict missing elements in the future.
Specifically, the model contains two components: (1) a multi-faceted attention
representation learning method that captures semantic dependence and temporal
evolution jointly; (2) an adaptive RL framework that conducts multi-hop
reasoning by adaptively learning the reward functions. Experimental results
demonstrate DREAM outperforms state-of-the-art models on public dataset
|
2304.03994 | Ruiqi Wu | Rui-Qi Wu, Zheng-Peng Duan, Chun-Le Guo, Zhi Chai, Chong-Yi Li | RIDCP: Revitalizing Real Image Dehazing via High-Quality Codebook Priors | Acceptted by CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Existing dehazing approaches struggle to process real-world hazy images owing
to the lack of paired real data and robust priors. In this work, we present a
new paradigm for real image dehazing from the perspectives of synthesizing more
realistic hazy data and introducing more robust priors into the network.
Specifically, (1) instead of adopting the de facto physical scattering model,
we rethink the degradation of real hazy images and propose a phenomenological
pipeline considering diverse degradation types. (2) We propose a Real Image
Dehazing network via high-quality Codebook Priors (RIDCP). Firstly, a VQGAN is
pre-trained on a large-scale high-quality dataset to obtain the discrete
codebook, encapsulating high-quality priors (HQPs). After replacing the
negative effects brought by haze with HQPs, the decoder equipped with a novel
normalized feature alignment module can effectively utilize high-quality
features and produce clean results. However, although our degradation pipeline
drastically mitigates the domain gap between synthetic and real data, it is
still intractable to avoid it, which challenges HQPs matching in the wild.
Thus, we re-calculate the distance when matching the features to the HQPs by a
controllable matching operation, which facilitates finding better counterparts.
We provide a recommendation to control the matching based on an explainable
solution. Users can also flexibly adjust the enhancement degree as per their
preference. Extensive experiments verify the effectiveness of our data
synthesis pipeline and the superior performance of RIDCP in real image
dehazing.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 12:12:24 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Wu",
"Rui-Qi",
""
],
[
"Duan",
"Zheng-Peng",
""
],
[
"Guo",
"Chun-Le",
""
],
[
"Chai",
"Zhi",
""
],
[
"Li",
"Chong-Yi",
""
]
] | TITLE: RIDCP: Revitalizing Real Image Dehazing via High-Quality Codebook Priors
ABSTRACT: Existing dehazing approaches struggle to process real-world hazy images owing
to the lack of paired real data and robust priors. In this work, we present a
new paradigm for real image dehazing from the perspectives of synthesizing more
realistic hazy data and introducing more robust priors into the network.
Specifically, (1) instead of adopting the de facto physical scattering model,
we rethink the degradation of real hazy images and propose a phenomenological
pipeline considering diverse degradation types. (2) We propose a Real Image
Dehazing network via high-quality Codebook Priors (RIDCP). Firstly, a VQGAN is
pre-trained on a large-scale high-quality dataset to obtain the discrete
codebook, encapsulating high-quality priors (HQPs). After replacing the
negative effects brought by haze with HQPs, the decoder equipped with a novel
normalized feature alignment module can effectively utilize high-quality
features and produce clean results. However, although our degradation pipeline
drastically mitigates the domain gap between synthetic and real data, it is
still intractable to avoid it, which challenges HQPs matching in the wild.
Thus, we re-calculate the distance when matching the features to the HQPs by a
controllable matching operation, which facilitates finding better counterparts.
We provide a recommendation to control the matching based on an explainable
solution. Users can also flexibly adjust the enhancement degree as per their
preference. Extensive experiments verify the effectiveness of our data
synthesis pipeline and the superior performance of RIDCP in real image
dehazing.
|
2304.04005 | Seyed Mohammad Hosien Abedy Nejad | Seyed Mohammad Hossein Abedy Nejad, Mohammad Amin Behzadi, Abdolrahim
Taheri | A new transformation for embedded convolutional neural network approach
toward real-time servo motor overload fault-detection | null | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Overloading in DC servo motors is a major concern in industries, as many
companies face the problem of finding expert operators, and also human
monitoring may not be an effective solution. Therefore, this paper proposed an
embedded Artificial intelligence (AI) approach using a Convolutional Neural
Network (CNN) using a new transformation to extract faults from real-time input
signals without human interference. Our main purpose is to extract as many as
possible features from the input signal to achieve a relaxed dataset that
results in an effective but compact network to provide real-time fault
detection even in a low-memory microcontroller. Besides, fault detection method
a synchronous dual-motor system is also proposed to take action in faulty
events. To fulfill this intention, a one-dimensional input signal from the
output current of each DC servo motor is monitored and transformed into a 3d
stack of data and then the CNN is implemented into the processor to detect any
fault corresponding to overloading, finally experimental setup results in
99.9997% accuracy during testing for a model with nearly 8000 parameters. In
addition, the proposed dual-motor system could achieve overload reduction and
provide a fault-tolerant system and it is shown that this system also takes
advantage of less energy consumption.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 13:36:33 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Nejad",
"Seyed Mohammad Hossein Abedy",
""
],
[
"Behzadi",
"Mohammad Amin",
""
],
[
"Taheri",
"Abdolrahim",
""
]
] | TITLE: A new transformation for embedded convolutional neural network approach
toward real-time servo motor overload fault-detection
ABSTRACT: Overloading in DC servo motors is a major concern in industries, as many
companies face the problem of finding expert operators, and also human
monitoring may not be an effective solution. Therefore, this paper proposed an
embedded Artificial intelligence (AI) approach using a Convolutional Neural
Network (CNN) using a new transformation to extract faults from real-time input
signals without human interference. Our main purpose is to extract as many as
possible features from the input signal to achieve a relaxed dataset that
results in an effective but compact network to provide real-time fault
detection even in a low-memory microcontroller. Besides, fault detection method
a synchronous dual-motor system is also proposed to take action in faulty
events. To fulfill this intention, a one-dimensional input signal from the
output current of each DC servo motor is monitored and transformed into a 3d
stack of data and then the CNN is implemented into the processor to detect any
fault corresponding to overloading, finally experimental setup results in
99.9997% accuracy during testing for a model with nearly 8000 parameters. In
addition, the proposed dual-motor system could achieve overload reduction and
provide a fault-tolerant system and it is shown that this system also takes
advantage of less energy consumption.
|
2304.04026 | Marek \v{S}uppa | D\'avid \v{S}uba and Marek \v{S}uppa and Jozef Kub\'ik and Endre
Hamerlik and Martin Tak\'a\v{c} | WikiGoldSK: Annotated Dataset, Baselines and Few-Shot Learning
Experiments for Slovak Named Entity Recognition | BSNLP 2023 Workshop at EACL 2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | Named Entity Recognition (NER) is a fundamental NLP tasks with a wide range
of practical applications. The performance of state-of-the-art NER methods
depends on high quality manually anotated datasets which still do not exist for
some languages. In this work we aim to remedy this situation in Slovak by
introducing WikiGoldSK, the first sizable human labelled Slovak NER dataset. We
benchmark it by evaluating state-of-the-art multilingual Pretrained Language
Models and comparing it to the existing silver-standard Slovak NER dataset. We
also conduct few-shot experiments and show that training on a sliver-standard
dataset yields better results. To enable future work that can be based on
Slovak NER, we release the dataset, code, as well as the trained models
publicly under permissible licensing terms at
https://github.com/NaiveNeuron/WikiGoldSK.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 14:37:52 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Šuba",
"Dávid",
""
],
[
"Šuppa",
"Marek",
""
],
[
"Kubík",
"Jozef",
""
],
[
"Hamerlik",
"Endre",
""
],
[
"Takáč",
"Martin",
""
]
] | TITLE: WikiGoldSK: Annotated Dataset, Baselines and Few-Shot Learning
Experiments for Slovak Named Entity Recognition
ABSTRACT: Named Entity Recognition (NER) is a fundamental NLP tasks with a wide range
of practical applications. The performance of state-of-the-art NER methods
depends on high quality manually anotated datasets which still do not exist for
some languages. In this work we aim to remedy this situation in Slovak by
introducing WikiGoldSK, the first sizable human labelled Slovak NER dataset. We
benchmark it by evaluating state-of-the-art multilingual Pretrained Language
Models and comparing it to the existing silver-standard Slovak NER dataset. We
also conduct few-shot experiments and show that training on a sliver-standard
dataset yields better results. To enable future work that can be based on
Slovak NER, we release the dataset, code, as well as the trained models
publicly under permissible licensing terms at
https://github.com/NaiveNeuron/WikiGoldSK.
|
2304.04046 | Haiwei Xie | Haiwei Xie, Federica Bellizio, Jochen L. Cremer, Goran Strbac | Regularised Learning with Selected Physics for Power System Dynamics | null | null | null | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the increasing system stability issues caused by the technological
revolutions of power system equipment, the assessment of the dynamic security
of the systems for changing operating conditions (OCs) is nowadays crucial. To
address the computational time problem of conventional dynamic security
assessment tools, many machine learning (ML) approaches have been proposed and
well-studied in this context. However, these learned models only rely on data,
and thus miss resourceful information offered by the physical system. To this
end, this paper focuses on combining the power system dynamical model together
with the conventional ML. Going beyond the classic Physics Informed Neural
Networks (PINNs), this paper proposes Selected Physics Informed Neural Networks
(SPINNs) to predict the system dynamics for varying OCs. A two-level structure
of feed-forward NNs is proposed, where the first NN predicts the generator bus
rotor angles (system states) and the second NN learns to adapt to varying OCs.
We show a case study on an IEEE-9 bus system that considering selected physics
in model training reduces the amount of needed training data. Moreover, the
trained model effectively predicted long-term dynamics that were beyond the
time scale of the collected training dataset (extrapolation).
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 15:36:32 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Xie",
"Haiwei",
""
],
[
"Bellizio",
"Federica",
""
],
[
"Cremer",
"Jochen L.",
""
],
[
"Strbac",
"Goran",
""
]
] | TITLE: Regularised Learning with Selected Physics for Power System Dynamics
ABSTRACT: Due to the increasing system stability issues caused by the technological
revolutions of power system equipment, the assessment of the dynamic security
of the systems for changing operating conditions (OCs) is nowadays crucial. To
address the computational time problem of conventional dynamic security
assessment tools, many machine learning (ML) approaches have been proposed and
well-studied in this context. However, these learned models only rely on data,
and thus miss resourceful information offered by the physical system. To this
end, this paper focuses on combining the power system dynamical model together
with the conventional ML. Going beyond the classic Physics Informed Neural
Networks (PINNs), this paper proposes Selected Physics Informed Neural Networks
(SPINNs) to predict the system dynamics for varying OCs. A two-level structure
of feed-forward NNs is proposed, where the first NN predicts the generator bus
rotor angles (system states) and the second NN learns to adapt to varying OCs.
We show a case study on an IEEE-9 bus system that considering selected physics
in model training reduces the amount of needed training data. Moreover, the
trained model effectively predicted long-term dynamics that were beyond the
time scale of the collected training dataset (extrapolation).
|
2304.04051 | George Watkins | George Watkins, Giovanni Montana, and Juergen Branke | Generating a Graph Colouring Heuristic with Deep Q-Learning and Graph
Neural Networks | 15 pages, 6 figures, to be published in LION17 conference proceedings | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The graph colouring problem consists of assigning labels, or colours, to the
vertices of a graph such that no two adjacent vertices share the same colour.
In this work we investigate whether deep reinforcement learning can be used to
discover a competitive construction heuristic for graph colouring. Our proposed
approach, ReLCol, uses deep Q-learning together with a graph neural network for
feature extraction, and employs a novel way of parameterising the graph that
results in improved performance. Using standard benchmark graphs with varied
topologies, we empirically evaluate the benefits and limitations of the
heuristic learned by ReLCol relative to existing construction algorithms, and
demonstrate that reinforcement learning is a promising direction for further
research on the graph colouring problem.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 15:41:01 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Watkins",
"George",
""
],
[
"Montana",
"Giovanni",
""
],
[
"Branke",
"Juergen",
""
]
] | TITLE: Generating a Graph Colouring Heuristic with Deep Q-Learning and Graph
Neural Networks
ABSTRACT: The graph colouring problem consists of assigning labels, or colours, to the
vertices of a graph such that no two adjacent vertices share the same colour.
In this work we investigate whether deep reinforcement learning can be used to
discover a competitive construction heuristic for graph colouring. Our proposed
approach, ReLCol, uses deep Q-learning together with a graph neural network for
feature extraction, and employs a novel way of parameterising the graph that
results in improved performance. Using standard benchmark graphs with varied
topologies, we empirically evaluate the benefits and limitations of the
heuristic learned by ReLCol relative to existing construction algorithms, and
demonstrate that reinforcement learning is a promising direction for further
research on the graph colouring problem.
|
2304.04060 | Thomas Le | Phuong D. Nguyen, Thinh D. Le, Duong Q. Nguyen, Binh Nguyen, H.
Nguyen-Xuan | Application of Self-Supervised Learning to MICA Model for Reconstructing
Imperfect 3D Facial Structures | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this study, we emphasize the integration of a pre-trained MICA model with
an imperfect face dataset, employing a self-supervised learning approach. We
present an innovative method for regenerating flawed facial structures,
yielding 3D printable outputs that effectively support physicians in their
patient treatment process. Our results highlight the model's capacity for
concealing scars and achieving comprehensive facial reconstructions without
discernible scarring. By capitalizing on pre-trained models and necessitating
only a few hours of supplementary training, our methodology adeptly devises an
optimal model for reconstructing damaged and imperfect facial features.
Harnessing contemporary 3D printing technology, we institute a standardized
protocol for fabricating realistic, camouflaging mask models for patients in a
laboratory environment.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 16:13:30 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Nguyen",
"Phuong D.",
""
],
[
"Le",
"Thinh D.",
""
],
[
"Nguyen",
"Duong Q.",
""
],
[
"Nguyen",
"Binh",
""
],
[
"Nguyen-Xuan",
"H.",
""
]
] | TITLE: Application of Self-Supervised Learning to MICA Model for Reconstructing
Imperfect 3D Facial Structures
ABSTRACT: In this study, we emphasize the integration of a pre-trained MICA model with
an imperfect face dataset, employing a self-supervised learning approach. We
present an innovative method for regenerating flawed facial structures,
yielding 3D printable outputs that effectively support physicians in their
patient treatment process. Our results highlight the model's capacity for
concealing scars and achieving comprehensive facial reconstructions without
discernible scarring. By capitalizing on pre-trained models and necessitating
only a few hours of supplementary training, our methodology adeptly devises an
optimal model for reconstructing damaged and imperfect facial features.
Harnessing contemporary 3D printing technology, we institute a standardized
protocol for fabricating realistic, camouflaging mask models for patients in a
laboratory environment.
|
2304.04068 | Javad Peymanfard | Javad Peymanfard, Ali Lashini, Samin Heydarian, Hossein Zeinali,
Nasser Mozayani | Word-level Persian Lipreading Dataset | null | In 2022 12th International Conference on Computer and Knowledge
Engineering (ICCKE) (pp. 225-230). IEEE | 10.1109/ICCKE57176.2022.9960105 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Lip-reading has made impressive progress in recent years, driven by advances
in deep learning. Nonetheless, the prerequisite such advances is a suitable
dataset. This paper provides a new in-the-wild dataset for Persian word-level
lipreading containing 244,000 videos from approximately 1,800 speakers. We
evaluated the state-of-the-art method in this field and used a novel approach
for word-level lip-reading. In this method, we used the AV-HuBERT model for
feature extraction and obtained significantly better performance on our
dataset.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 17:00:35 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Peymanfard",
"Javad",
""
],
[
"Lashini",
"Ali",
""
],
[
"Heydarian",
"Samin",
""
],
[
"Zeinali",
"Hossein",
""
],
[
"Mozayani",
"Nasser",
""
]
] | TITLE: Word-level Persian Lipreading Dataset
ABSTRACT: Lip-reading has made impressive progress in recent years, driven by advances
in deep learning. Nonetheless, the prerequisite such advances is a suitable
dataset. This paper provides a new in-the-wild dataset for Persian word-level
lipreading containing 244,000 videos from approximately 1,800 speakers. We
evaluated the state-of-the-art method in this field and used a novel approach
for word-level lip-reading. In this method, we used the AV-HuBERT model for
feature extraction and obtained significantly better performance on our
dataset.
|
2304.04118 | Muskan Garg | Muskan Garg | Multi-class Categorization of Reasons behind Mental Disturbance in Long
Texts | null | null | null | null | cs.CL cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Motivated with recent advances in inferring users' mental state in social
media posts, we identify and formulate the problem of finding causal indicators
behind mental illness in self-reported text. In the past, we witness the
presence of rule-based studies for causal explanation analysis on curated
Facebook data. The investigation on transformer-based model for multi-class
causal categorization in Reddit posts point to a problem of using long-text
which contains as many as 4000 words. Developing end-to-end transformer-based
models subject to the limitation of maximum-length in a given instance. To
handle this problem, we use Longformer and deploy its encoding on
transformer-based classifier. The experimental results show that Longformer
achieves new state-of-the-art results on M-CAMS, a publicly available dataset
with 62\% F1-score. Cause-specific analysis and ablation study prove the
effectiveness of Longformer. We believe our work facilitates causal analysis of
depression and suicide risk on social media data, and shows potential for
application on other mental health conditions.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 22:44:32 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Garg",
"Muskan",
""
]
] | TITLE: Multi-class Categorization of Reasons behind Mental Disturbance in Long
Texts
ABSTRACT: Motivated with recent advances in inferring users' mental state in social
media posts, we identify and formulate the problem of finding causal indicators
behind mental illness in self-reported text. In the past, we witness the
presence of rule-based studies for causal explanation analysis on curated
Facebook data. The investigation on transformer-based model for multi-class
causal categorization in Reddit posts point to a problem of using long-text
which contains as many as 4000 words. Developing end-to-end transformer-based
models subject to the limitation of maximum-length in a given instance. To
handle this problem, we use Longformer and deploy its encoding on
transformer-based classifier. The experimental results show that Longformer
achieves new state-of-the-art results on M-CAMS, a publicly available dataset
with 62\% F1-score. Cause-specific analysis and ablation study prove the
effectiveness of Longformer. We believe our work facilitates causal analysis of
depression and suicide risk on social media data, and shows potential for
application on other mental health conditions.
|
2304.04120 | Shanglin Zhou | Shanglin Zhou, Mikhail A. Bragin, Lynn Pepin, Deniz Gurevin, Fei Miao,
Caiwen Ding | Surrogate Lagrangian Relaxation: A Path To Retrain-free Deep Neural
Network Pruning | arXiv admin note: text overlap with arXiv:2012.10079 | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Network pruning is a widely used technique to reduce computation cost and
model size for deep neural networks. However, the typical three-stage pipeline
significantly increases the overall training time. In this paper, we develop a
systematic weight-pruning optimization approach based on Surrogate Lagrangian
relaxation, which is tailored to overcome difficulties caused by the discrete
nature of the weight-pruning problem. We prove that our method ensures fast
convergence of the model compression problem, and the convergence of the SLR is
accelerated by using quadratic penalties. Model parameters obtained by SLR
during the training phase are much closer to their optimal values as compared
to those obtained by other state-of-the-art methods. We evaluate our method on
image classification tasks using CIFAR-10 and ImageNet with state-of-the-art
MLP-Mixer, Swin Transformer, and VGG-16, ResNet-18, ResNet-50 and ResNet-110,
MobileNetV2. We also evaluate object detection and segmentation tasks on COCO,
KITTI benchmark, and TuSimple lane detection dataset using a variety of models.
Experimental results demonstrate that our SLR-based weight-pruning optimization
approach achieves a higher compression rate than state-of-the-art methods under
the same accuracy requirement and also can achieve higher accuracy under the
same compression rate requirement. Under classification tasks, our SLR approach
converges to the desired accuracy $3\times$ faster on both of the datasets.
Under object detection and segmentation tasks, SLR also converges $2\times$
faster to the desired accuracy. Further, our SLR achieves high model accuracy
even at the hard-pruning stage without retraining, which reduces the
traditional three-stage pruning into a two-stage process. Given a limited
budget of retraining epochs, our approach quickly recovers the model's
accuracy.
| [
{
"version": "v1",
"created": "Sat, 8 Apr 2023 22:48:30 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Zhou",
"Shanglin",
""
],
[
"Bragin",
"Mikhail A.",
""
],
[
"Pepin",
"Lynn",
""
],
[
"Gurevin",
"Deniz",
""
],
[
"Miao",
"Fei",
""
],
[
"Ding",
"Caiwen",
""
]
] | TITLE: Surrogate Lagrangian Relaxation: A Path To Retrain-free Deep Neural
Network Pruning
ABSTRACT: Network pruning is a widely used technique to reduce computation cost and
model size for deep neural networks. However, the typical three-stage pipeline
significantly increases the overall training time. In this paper, we develop a
systematic weight-pruning optimization approach based on Surrogate Lagrangian
relaxation, which is tailored to overcome difficulties caused by the discrete
nature of the weight-pruning problem. We prove that our method ensures fast
convergence of the model compression problem, and the convergence of the SLR is
accelerated by using quadratic penalties. Model parameters obtained by SLR
during the training phase are much closer to their optimal values as compared
to those obtained by other state-of-the-art methods. We evaluate our method on
image classification tasks using CIFAR-10 and ImageNet with state-of-the-art
MLP-Mixer, Swin Transformer, and VGG-16, ResNet-18, ResNet-50 and ResNet-110,
MobileNetV2. We also evaluate object detection and segmentation tasks on COCO,
KITTI benchmark, and TuSimple lane detection dataset using a variety of models.
Experimental results demonstrate that our SLR-based weight-pruning optimization
approach achieves a higher compression rate than state-of-the-art methods under
the same accuracy requirement and also can achieve higher accuracy under the
same compression rate requirement. Under classification tasks, our SLR approach
converges to the desired accuracy $3\times$ faster on both of the datasets.
Under object detection and segmentation tasks, SLR also converges $2\times$
faster to the desired accuracy. Further, our SLR achieves high model accuracy
even at the hard-pruning stage without retraining, which reduces the
traditional three-stage pruning into a two-stage process. Given a limited
budget of retraining epochs, our approach quickly recovers the model's
accuracy.
|
2304.04140 | Jie Yang | Jie Yang, Chaoqun Wang, Zhen Li, Junle Wang, Ruimao Zhang | Semantic Human Parsing via Scalable Semantic Transfer over Multiple
Label Domains | Accepted to CVPR23 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper presents Scalable Semantic Transfer (SST), a novel training
paradigm, to explore how to leverage the mutual benefits of the data from
different label domains (i.e. various levels of label granularity) to train a
powerful human parsing network. In practice, two common application scenarios
are addressed, termed universal parsing and dedicated parsing, where the former
aims to learn homogeneous human representations from multiple label domains and
switch predictions by only using different segmentation heads, and the latter
aims to learn a specific domain prediction while distilling the semantic
knowledge from other domains. The proposed SST has the following appealing
benefits: (1) it can capably serve as an effective training scheme to embed
semantic associations of human body parts from multiple label domains into the
human representation learning process; (2) it is an extensible semantic
transfer framework without predetermining the overall relations of multiple
label domains, which allows continuously adding human parsing datasets to
promote the training. (3) the relevant modules are only used for auxiliary
training and can be removed during inference, eliminating the extra reasoning
cost. Experimental results demonstrate SST can effectively achieve promising
universal human parsing performance as well as impressive improvements compared
to its counterparts on three human parsing benchmarks (i.e.,
PASCAL-Person-Part, ATR, and CIHP). Code is available at
https://github.com/yangjie-cv/SST.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 02:44:29 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Yang",
"Jie",
""
],
[
"Wang",
"Chaoqun",
""
],
[
"Li",
"Zhen",
""
],
[
"Wang",
"Junle",
""
],
[
"Zhang",
"Ruimao",
""
]
] | TITLE: Semantic Human Parsing via Scalable Semantic Transfer over Multiple
Label Domains
ABSTRACT: This paper presents Scalable Semantic Transfer (SST), a novel training
paradigm, to explore how to leverage the mutual benefits of the data from
different label domains (i.e. various levels of label granularity) to train a
powerful human parsing network. In practice, two common application scenarios
are addressed, termed universal parsing and dedicated parsing, where the former
aims to learn homogeneous human representations from multiple label domains and
switch predictions by only using different segmentation heads, and the latter
aims to learn a specific domain prediction while distilling the semantic
knowledge from other domains. The proposed SST has the following appealing
benefits: (1) it can capably serve as an effective training scheme to embed
semantic associations of human body parts from multiple label domains into the
human representation learning process; (2) it is an extensible semantic
transfer framework without predetermining the overall relations of multiple
label domains, which allows continuously adding human parsing datasets to
promote the training. (3) the relevant modules are only used for auxiliary
training and can be removed during inference, eliminating the extra reasoning
cost. Experimental results demonstrate SST can effectively achieve promising
universal human parsing performance as well as impressive improvements compared
to its counterparts on three human parsing benchmarks (i.e.,
PASCAL-Person-Part, ATR, and CIHP). Code is available at
https://github.com/yangjie-cv/SST.
|
2304.04147 | Ravi Vadlamani | Polaki Durga Prasad, Yelleti Vivek, Vadlamani Ravi | FedPNN: One-shot Federated Classification via Evolving Clustering Method
and Probabilistic Neural Network hybrid | 27 pages, 13 figures, 7 tables | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Protecting data privacy is paramount in the fields such as finance, banking,
and healthcare. Federated Learning (FL) has attracted widespread attention due
to its decentralized, distributed training and the ability to protect the
privacy while obtaining a global shared model. However, FL presents challenges
such as communication overhead, and limited resource capability. This motivated
us to propose a two-stage federated learning approach toward the objective of
privacy protection, which is a first-of-its-kind study as follows: (i) During
the first stage, the synthetic dataset is generated by employing two different
distributions as noise to the vanilla conditional tabular generative
adversarial neural network (CTGAN) resulting in modified CTGAN, and (ii) In the
second stage, the Federated Probabilistic Neural Network (FedPNN) is developed
and employed for building globally shared classification model. We also
employed synthetic dataset metrics to check the quality of the generated
synthetic dataset. Further, we proposed a meta-clustering algorithm whereby the
cluster centers obtained from the clients are clustered at the server for
training the global model. Despite PNN being a one-pass learning classifier,
its complexity depends on the training data size. Therefore, we employed a
modified evolving clustering method (ECM), another one-pass algorithm to
cluster the training data thereby increasing the speed further. Moreover, we
conducted sensitivity analysis by varying Dthr, a hyperparameter of ECM at the
server and client, one at a time. The effectiveness of our approach is
validated on four finance and medical datasets.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 03:23:37 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Prasad",
"Polaki Durga",
""
],
[
"Vivek",
"Yelleti",
""
],
[
"Ravi",
"Vadlamani",
""
]
] | TITLE: FedPNN: One-shot Federated Classification via Evolving Clustering Method
and Probabilistic Neural Network hybrid
ABSTRACT: Protecting data privacy is paramount in the fields such as finance, banking,
and healthcare. Federated Learning (FL) has attracted widespread attention due
to its decentralized, distributed training and the ability to protect the
privacy while obtaining a global shared model. However, FL presents challenges
such as communication overhead, and limited resource capability. This motivated
us to propose a two-stage federated learning approach toward the objective of
privacy protection, which is a first-of-its-kind study as follows: (i) During
the first stage, the synthetic dataset is generated by employing two different
distributions as noise to the vanilla conditional tabular generative
adversarial neural network (CTGAN) resulting in modified CTGAN, and (ii) In the
second stage, the Federated Probabilistic Neural Network (FedPNN) is developed
and employed for building globally shared classification model. We also
employed synthetic dataset metrics to check the quality of the generated
synthetic dataset. Further, we proposed a meta-clustering algorithm whereby the
cluster centers obtained from the clients are clustered at the server for
training the global model. Despite PNN being a one-pass learning classifier,
its complexity depends on the training data size. Therefore, we employed a
modified evolving clustering method (ECM), another one-pass algorithm to
cluster the training data thereby increasing the speed further. Moreover, we
conducted sensitivity analysis by varying Dthr, a hyperparameter of ECM at the
server and client, one at a time. The effectiveness of our approach is
validated on four finance and medical datasets.
|
2304.04151 | Yan Luo | Yan Luo, Haoyi Duan, Ye Liu, Fu-lai Chung | Timestamps as Prompts for Geography-Aware Location Recommendation | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Location recommendation plays a vital role in improving users' travel
experience. The timestamp of the POI to be predicted is of great significance,
since a user will go to different places at different times. However, most
existing methods either do not use this kind of temporal information, or just
implicitly fuse it with other contextual information. In this paper, we revisit
the problem of location recommendation and point out that explicitly modeling
temporal information is a great help when the model needs to predict not only
the next location but also further locations. In addition, state-of-the-art
methods do not make effective use of geographic information and suffer from the
hard boundary problem when encoding geographic information by gridding. To this
end, a Temporal Prompt-based and Geography-aware (TPG) framework is proposed.
The temporal prompt is firstly designed to incorporate temporal information of
any further check-in. A shifted window mechanism is then devised to augment
geographic data for addressing the hard boundary problem. Via extensive
comparisons with existing methods and ablation studies on five real-world
datasets, we demonstrate the effectiveness and superiority of the proposed
method under various settings. Most importantly, our proposed model has the
superior ability of interval prediction. In particular, the model can predict
the location that a user wants to go to at a certain time while the most recent
check-in behavioral data is masked, or it can predict specific future check-in
(not just the next one) at a given timestamp.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 03:58:08 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Luo",
"Yan",
""
],
[
"Duan",
"Haoyi",
""
],
[
"Liu",
"Ye",
""
],
[
"Chung",
"Fu-lai",
""
]
] | TITLE: Timestamps as Prompts for Geography-Aware Location Recommendation
ABSTRACT: Location recommendation plays a vital role in improving users' travel
experience. The timestamp of the POI to be predicted is of great significance,
since a user will go to different places at different times. However, most
existing methods either do not use this kind of temporal information, or just
implicitly fuse it with other contextual information. In this paper, we revisit
the problem of location recommendation and point out that explicitly modeling
temporal information is a great help when the model needs to predict not only
the next location but also further locations. In addition, state-of-the-art
methods do not make effective use of geographic information and suffer from the
hard boundary problem when encoding geographic information by gridding. To this
end, a Temporal Prompt-based and Geography-aware (TPG) framework is proposed.
The temporal prompt is firstly designed to incorporate temporal information of
any further check-in. A shifted window mechanism is then devised to augment
geographic data for addressing the hard boundary problem. Via extensive
comparisons with existing methods and ablation studies on five real-world
datasets, we demonstrate the effectiveness and superiority of the proposed
method under various settings. Most importantly, our proposed model has the
superior ability of interval prediction. In particular, the model can predict
the location that a user wants to go to at a certain time while the most recent
check-in behavioral data is masked, or it can predict specific future check-in
(not just the next one) at a given timestamp.
|
2304.04152 | Qijiong Liu | Tiandeng Wu, Qijiong Liu, Yi Cao, Yao Huang, Xiao-Ming Wu, Jiandong
Ding | Continual Graph Convolutional Network for Text Classification | 9 pages, 4 figures, AAAI 2023 accepted paper | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Graph convolutional network (GCN) has been successfully applied to capture
global non-consecutive and long-distance semantic information for text
classification. However, while GCN-based methods have shown promising results
in offline evaluations, they commonly follow a seen-token-seen-document
paradigm by constructing a fixed document-token graph and cannot make
inferences on new documents. It is a challenge to deploy them in online systems
to infer steaming text data. In this work, we present a continual GCN model
(ContGCN) to generalize inferences from observed documents to unobserved
documents. Concretely, we propose a new all-token-any-document paradigm to
dynamically update the document-token graph in every batch during both the
training and testing phases of an online system. Moreover, we design an
occurrence memory module and a self-supervised contrastive learning objective
to update ContGCN in a label-free manner. A 3-month A/B test on Huawei public
opinion analysis system shows ContGCN achieves 8.86% performance gain compared
with state-of-the-art methods. Offline experiments on five public datasets also
show ContGCN can improve inference quality. The source code will be released at
https://github.com/Jyonn/ContGCN.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 03:59:48 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Wu",
"Tiandeng",
""
],
[
"Liu",
"Qijiong",
""
],
[
"Cao",
"Yi",
""
],
[
"Huang",
"Yao",
""
],
[
"Wu",
"Xiao-Ming",
""
],
[
"Ding",
"Jiandong",
""
]
] | TITLE: Continual Graph Convolutional Network for Text Classification
ABSTRACT: Graph convolutional network (GCN) has been successfully applied to capture
global non-consecutive and long-distance semantic information for text
classification. However, while GCN-based methods have shown promising results
in offline evaluations, they commonly follow a seen-token-seen-document
paradigm by constructing a fixed document-token graph and cannot make
inferences on new documents. It is a challenge to deploy them in online systems
to infer steaming text data. In this work, we present a continual GCN model
(ContGCN) to generalize inferences from observed documents to unobserved
documents. Concretely, we propose a new all-token-any-document paradigm to
dynamically update the document-token graph in every batch during both the
training and testing phases of an online system. Moreover, we design an
occurrence memory module and a self-supervised contrastive learning objective
to update ContGCN in a label-free manner. A 3-month A/B test on Huawei public
opinion analysis system shows ContGCN achieves 8.86% performance gain compared
with state-of-the-art methods. Offline experiments on five public datasets also
show ContGCN can improve inference quality. The source code will be released at
https://github.com/Jyonn/ContGCN.
|
2304.04158 | Haiyan Zhao | Haiyan Zhao, Tianyi Zhou, Guodong Long, Jing Jiang, Chengqi Zhang | Does Continual Learning Equally Forget All Parameters? | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Distribution shift (e.g., task or domain shift) in continual learning (CL)
usually results in catastrophic forgetting of neural networks. Although it can
be alleviated by repeatedly replaying buffered data, the every-step replay is
time-consuming. In this paper, we study which modules in neural networks are
more prone to forgetting by investigating their training dynamics during CL.
Our proposed metrics show that only a few modules are more task-specific and
sensitively alter between tasks, while others can be shared across tasks as
common knowledge. Hence, we attribute forgetting mainly to the former and find
that finetuning them only on a small buffer at the end of any CL method can
bring non-trivial improvement. Due to the small number of finetuned parameters,
such ``Forgetting Prioritized Finetuning (FPF)'' is efficient in computation.
We further propose a more efficient and simpler method that entirely removes
the every-step replay and replaces them by only $k$-times of FPF periodically
triggered during CL. Surprisingly, this ``$k$-FPF'' performs comparably to FPF
and outperforms the SOTA CL methods but significantly reduces their
computational overhead and cost. In experiments on several benchmarks of class-
and domain-incremental CL, FPF consistently improves existing CL methods by a
large margin, and $k$-FPF further excels in efficiency without degrading the
accuracy. We also empirically studied the impact of buffer size, epochs per
task, and finetuning modules on the cost and accuracy of our methods.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 04:36:24 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Zhao",
"Haiyan",
""
],
[
"Zhou",
"Tianyi",
""
],
[
"Long",
"Guodong",
""
],
[
"Jiang",
"Jing",
""
],
[
"Zhang",
"Chengqi",
""
]
] | TITLE: Does Continual Learning Equally Forget All Parameters?
ABSTRACT: Distribution shift (e.g., task or domain shift) in continual learning (CL)
usually results in catastrophic forgetting of neural networks. Although it can
be alleviated by repeatedly replaying buffered data, the every-step replay is
time-consuming. In this paper, we study which modules in neural networks are
more prone to forgetting by investigating their training dynamics during CL.
Our proposed metrics show that only a few modules are more task-specific and
sensitively alter between tasks, while others can be shared across tasks as
common knowledge. Hence, we attribute forgetting mainly to the former and find
that finetuning them only on a small buffer at the end of any CL method can
bring non-trivial improvement. Due to the small number of finetuned parameters,
such ``Forgetting Prioritized Finetuning (FPF)'' is efficient in computation.
We further propose a more efficient and simpler method that entirely removes
the every-step replay and replaces them by only $k$-times of FPF periodically
triggered during CL. Surprisingly, this ``$k$-FPF'' performs comparably to FPF
and outperforms the SOTA CL methods but significantly reduces their
computational overhead and cost. In experiments on several benchmarks of class-
and domain-incremental CL, FPF consistently improves existing CL methods by a
large margin, and $k$-FPF further excels in efficiency without degrading the
accuracy. We also empirically studied the impact of buffer size, epochs per
task, and finetuning modules on the cost and accuracy of our methods.
|
2304.04161 | Zanobya Khan | Zanoby N.Khan | Detection of COVID19 in Chest X-Ray Images Using Transfer Learning | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | COVID19 is a highly contagious disease infected millions of people worldwide.
With limited testing components, screening tools such as chest radiography can
assist the clinicians in the diagnosis and assessing the progress of disease.
The performance of deep learning-based systems for diagnosis of COVID-19
disease in radiograph images has been encouraging. This paper investigates the
concept of transfer learning using two of the most well-known VGGNet
architectures, namely VGG-16 and VGG-19. The classifier block and
hyperparameters are fine-tuned to adopt the models for automatic detection of
Covid-19 in chest x-ray images. We generated two different datasets to evaluate
the performance of the proposed system for the identification of positive
Covid-19 instances in a multiclass and binary classification problems. The
experimental outcome demonstrates the usefulness of transfer learning for
small-sized datasets particularly in the field of medical imaging, not only to
prevent over-fitting and convergence problems but also to attain optimal
classification performance as well.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 05:02:04 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Khan",
"Zanoby N.",
""
]
] | TITLE: Detection of COVID19 in Chest X-Ray Images Using Transfer Learning
ABSTRACT: COVID19 is a highly contagious disease infected millions of people worldwide.
With limited testing components, screening tools such as chest radiography can
assist the clinicians in the diagnosis and assessing the progress of disease.
The performance of deep learning-based systems for diagnosis of COVID-19
disease in radiograph images has been encouraging. This paper investigates the
concept of transfer learning using two of the most well-known VGGNet
architectures, namely VGG-16 and VGG-19. The classifier block and
hyperparameters are fine-tuned to adopt the models for automatic detection of
Covid-19 in chest x-ray images. We generated two different datasets to evaluate
the performance of the proposed system for the identification of positive
Covid-19 instances in a multiclass and binary classification problems. The
experimental outcome demonstrates the usefulness of transfer learning for
small-sized datasets particularly in the field of medical imaging, not only to
prevent over-fitting and convergence problems but also to attain optimal
classification performance as well.
|
2304.04168 | Beini Xie | Beini Xie, Heng Chang, Ziwei Zhang, Xin Wang, Daixin Wang, Zhiqiang
Zhang, Rex Ying, Wenwu Zhu | Adversarially Robust Neural Architecture Search for Graph Neural
Networks | Accepted as a conference paper at CVPR 2023 | null | null | null | cs.LG cs.CR cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph Neural Networks (GNNs) obtain tremendous success in modeling relational
data. Still, they are prone to adversarial attacks, which are massive threats
to applying GNNs to risk-sensitive domains. Existing defensive methods neither
guarantee performance facing new data/tasks or adversarial attacks nor provide
insights to understand GNN robustness from an architectural perspective. Neural
Architecture Search (NAS) has the potential to solve this problem by automating
GNN architecture designs. Nevertheless, current graph NAS approaches lack
robust design and are vulnerable to adversarial attacks. To tackle these
challenges, we propose a novel Robust Neural Architecture search framework for
GNNs (G-RNA). Specifically, we design a robust search space for the
message-passing mechanism by adding graph structure mask operations into the
search space, which comprises various defensive operation candidates and allows
us to search for defensive GNNs. Furthermore, we define a robustness metric to
guide the search procedure, which helps to filter robust architectures. In this
way, G-RNA helps understand GNN robustness from an architectural perspective
and effectively searches for optimal adversarial robust GNNs. Extensive
experimental results on benchmark datasets show that G-RNA significantly
outperforms manually designed robust GNNs and vanilla graph NAS baselines by
12.1% to 23.4% under adversarial attacks.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 06:00:50 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Xie",
"Beini",
""
],
[
"Chang",
"Heng",
""
],
[
"Zhang",
"Ziwei",
""
],
[
"Wang",
"Xin",
""
],
[
"Wang",
"Daixin",
""
],
[
"Zhang",
"Zhiqiang",
""
],
[
"Ying",
"Rex",
""
],
[
"Zhu",
"Wenwu",
""
]
] | TITLE: Adversarially Robust Neural Architecture Search for Graph Neural
Networks
ABSTRACT: Graph Neural Networks (GNNs) obtain tremendous success in modeling relational
data. Still, they are prone to adversarial attacks, which are massive threats
to applying GNNs to risk-sensitive domains. Existing defensive methods neither
guarantee performance facing new data/tasks or adversarial attacks nor provide
insights to understand GNN robustness from an architectural perspective. Neural
Architecture Search (NAS) has the potential to solve this problem by automating
GNN architecture designs. Nevertheless, current graph NAS approaches lack
robust design and are vulnerable to adversarial attacks. To tackle these
challenges, we propose a novel Robust Neural Architecture search framework for
GNNs (G-RNA). Specifically, we design a robust search space for the
message-passing mechanism by adding graph structure mask operations into the
search space, which comprises various defensive operation candidates and allows
us to search for defensive GNNs. Furthermore, we define a robustness metric to
guide the search procedure, which helps to filter robust architectures. In this
way, G-RNA helps understand GNN robustness from an architectural perspective
and effectively searches for optimal adversarial robust GNNs. Extensive
experimental results on benchmark datasets show that G-RNA significantly
outperforms manually designed robust GNNs and vanilla graph NAS baselines by
12.1% to 23.4% under adversarial attacks.
|
2304.04171 | Weiwei Sun | Weiwei Sun, Lingyong Yan, Zheng Chen, Shuaiqiang Wang, Haichao Zhu,
Pengjie Ren, Zhumin Chen, Dawei Yin, Maarten de Rijke and Zhaochun Ren | Learning to Tokenize for Generative Retrieval | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Conventional document retrieval techniques are mainly based on the
index-retrieve paradigm. It is challenging to optimize pipelines based on this
paradigm in an end-to-end manner. As an alternative, generative retrieval
represents documents as identifiers (docid) and retrieves documents by
generating docids, enabling end-to-end modeling of document retrieval tasks.
However, it is an open question how one should define the document identifiers.
Current approaches to the task of defining document identifiers rely on fixed
rule-based docids, such as the title of a document or the result of clustering
BERT embeddings, which often fail to capture the complete semantic information
of a document. We propose GenRet, a document tokenization learning method to
address the challenge of defining document identifiers for generative
retrieval. GenRet learns to tokenize documents into short discrete
representations (i.e., docids) via a discrete auto-encoding approach. Three
components are included in GenRet: (i) a tokenization model that produces
docids for documents; (ii) a reconstruction model that learns to reconstruct a
document based on a docid; and (iii) a sequence-to-sequence retrieval model
that generates relevant document identifiers directly for a designated query.
By using an auto-encoding framework, GenRet learns semantic docids in a fully
end-to-end manner. We also develop a progressive training scheme to capture the
autoregressive nature of docids and to stabilize training. We conduct
experiments on the NQ320K, MS MARCO, and BEIR datasets to assess the
effectiveness of GenRet. GenRet establishes the new state-of-the-art on the
NQ320K dataset. Especially, compared to generative retrieval baselines, GenRet
can achieve significant improvements on the unseen documents. GenRet also
outperforms comparable baselines on MS MARCO and BEIR, demonstrating the
method's generalizability.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 06:18:19 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Sun",
"Weiwei",
""
],
[
"Yan",
"Lingyong",
""
],
[
"Chen",
"Zheng",
""
],
[
"Wang",
"Shuaiqiang",
""
],
[
"Zhu",
"Haichao",
""
],
[
"Ren",
"Pengjie",
""
],
[
"Chen",
"Zhumin",
""
],
[
"Yin",
"Dawei",
""
],
[
"de Rijke",
"Maarten",
""
],
[
"Ren",
"Zhaochun",
""
]
] | TITLE: Learning to Tokenize for Generative Retrieval
ABSTRACT: Conventional document retrieval techniques are mainly based on the
index-retrieve paradigm. It is challenging to optimize pipelines based on this
paradigm in an end-to-end manner. As an alternative, generative retrieval
represents documents as identifiers (docid) and retrieves documents by
generating docids, enabling end-to-end modeling of document retrieval tasks.
However, it is an open question how one should define the document identifiers.
Current approaches to the task of defining document identifiers rely on fixed
rule-based docids, such as the title of a document or the result of clustering
BERT embeddings, which often fail to capture the complete semantic information
of a document. We propose GenRet, a document tokenization learning method to
address the challenge of defining document identifiers for generative
retrieval. GenRet learns to tokenize documents into short discrete
representations (i.e., docids) via a discrete auto-encoding approach. Three
components are included in GenRet: (i) a tokenization model that produces
docids for documents; (ii) a reconstruction model that learns to reconstruct a
document based on a docid; and (iii) a sequence-to-sequence retrieval model
that generates relevant document identifiers directly for a designated query.
By using an auto-encoding framework, GenRet learns semantic docids in a fully
end-to-end manner. We also develop a progressive training scheme to capture the
autoregressive nature of docids and to stabilize training. We conduct
experiments on the NQ320K, MS MARCO, and BEIR datasets to assess the
effectiveness of GenRet. GenRet establishes the new state-of-the-art on the
NQ320K dataset. Especially, compared to generative retrieval baselines, GenRet
can achieve significant improvements on the unseen documents. GenRet also
outperforms comparable baselines on MS MARCO and BEIR, demonstrating the
method's generalizability.
|
2304.04179 | Yulu Gao | Yulu Gao, Chonghao Sima, Shaoshuai Shi, Shangzhe Di, Si Liu, Hongyang
Li | Sparse Dense Fusion for 3D Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the prevalence of multimodal learning, camera-LiDAR fusion has gained
popularity in 3D object detection. Although multiple fusion approaches have
been proposed, they can be classified into either sparse-only or dense-only
fashion based on the feature representation in the fusion module. In this
paper, we analyze them in a common taxonomy and thereafter observe two
challenges: 1) sparse-only solutions preserve 3D geometric prior and yet lose
rich semantic information from the camera, and 2) dense-only alternatives
retain the semantic continuity but miss the accurate geometric information from
LiDAR. By analyzing these two formulations, we conclude that the information
loss is inevitable due to their design scheme. To compensate for the
information loss in either manner, we propose Sparse Dense Fusion (SDF), a
complementary framework that incorporates both sparse-fusion and dense-fusion
modules via the Transformer architecture. Such a simple yet effective
sparse-dense fusion structure enriches semantic texture and exploits spatial
structure information simultaneously. Through our SDF strategy, we assemble two
popular methods with moderate performance and outperform baseline by 4.3% in
mAP and 2.5% in NDS, ranking first on the nuScenes benchmark. Extensive
ablations demonstrate the effectiveness of our method and empirically align our
analysis.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 07:10:34 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Gao",
"Yulu",
""
],
[
"Sima",
"Chonghao",
""
],
[
"Shi",
"Shaoshuai",
""
],
[
"Di",
"Shangzhe",
""
],
[
"Liu",
"Si",
""
],
[
"Li",
"Hongyang",
""
]
] | TITLE: Sparse Dense Fusion for 3D Object Detection
ABSTRACT: With the prevalence of multimodal learning, camera-LiDAR fusion has gained
popularity in 3D object detection. Although multiple fusion approaches have
been proposed, they can be classified into either sparse-only or dense-only
fashion based on the feature representation in the fusion module. In this
paper, we analyze them in a common taxonomy and thereafter observe two
challenges: 1) sparse-only solutions preserve 3D geometric prior and yet lose
rich semantic information from the camera, and 2) dense-only alternatives
retain the semantic continuity but miss the accurate geometric information from
LiDAR. By analyzing these two formulations, we conclude that the information
loss is inevitable due to their design scheme. To compensate for the
information loss in either manner, we propose Sparse Dense Fusion (SDF), a
complementary framework that incorporates both sparse-fusion and dense-fusion
modules via the Transformer architecture. Such a simple yet effective
sparse-dense fusion structure enriches semantic texture and exploits spatial
structure information simultaneously. Through our SDF strategy, we assemble two
popular methods with moderate performance and outperform baseline by 4.3% in
mAP and 2.5% in NDS, ranking first on the nuScenes benchmark. Extensive
ablations demonstrate the effectiveness of our method and empirically align our
analysis.
|
2304.04185 | Yinhao Li | Yinhao Li, Jinrong Yang, Jianjian Sun, Han Bao, Zheng Ge, Li Xiao | BEVStereo++: Accurate Depth Estimation in Multi-view 3D Object Detection
via Dynamic Temporal Stereo | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bounded by the inherent ambiguity of depth perception, contemporary
multi-view 3D object detection methods fall into the performance bottleneck.
Intuitively, leveraging temporal multi-view stereo (MVS) technology is the
natural knowledge for tackling this ambiguity. However, traditional attempts of
MVS has two limitations when applying to 3D object detection scenes: 1) The
affinity measurement among all views suffers expensive computational cost; 2)
It is difficult to deal with outdoor scenarios where objects are often mobile.
To this end, we propose BEVStereo++: by introducing a dynamic temporal stereo
strategy, BEVStereo++ is able to cut down the harm that is brought by
introducing temporal stereo when dealing with those two scenarios. Going one
step further, we apply Motion Compensation Module and long sequence Frame
Fusion to BEVStereo++, which shows further performance boosting and error
reduction. Without bells and whistles, BEVStereo++ achieves
state-of-the-art(SOTA) on both Waymo and nuScenes dataset.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 08:04:26 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Li",
"Yinhao",
""
],
[
"Yang",
"Jinrong",
""
],
[
"Sun",
"Jianjian",
""
],
[
"Bao",
"Han",
""
],
[
"Ge",
"Zheng",
""
],
[
"Xiao",
"Li",
""
]
] | TITLE: BEVStereo++: Accurate Depth Estimation in Multi-view 3D Object Detection
via Dynamic Temporal Stereo
ABSTRACT: Bounded by the inherent ambiguity of depth perception, contemporary
multi-view 3D object detection methods fall into the performance bottleneck.
Intuitively, leveraging temporal multi-view stereo (MVS) technology is the
natural knowledge for tackling this ambiguity. However, traditional attempts of
MVS has two limitations when applying to 3D object detection scenes: 1) The
affinity measurement among all views suffers expensive computational cost; 2)
It is difficult to deal with outdoor scenarios where objects are often mobile.
To this end, we propose BEVStereo++: by introducing a dynamic temporal stereo
strategy, BEVStereo++ is able to cut down the harm that is brought by
introducing temporal stereo when dealing with those two scenarios. Going one
step further, we apply Motion Compensation Module and long sequence Frame
Fusion to BEVStereo++, which shows further performance boosting and error
reduction. Without bells and whistles, BEVStereo++ achieves
state-of-the-art(SOTA) on both Waymo and nuScenes dataset.
|
2304.04200 | Qiu Changjie | Changjie Qiu, Zhiyong Wang, Xiuhong Lin, Yu Zang, Cheng Wang, Weiquan
Liu | DSMNet: Deep High-precision 3D Surface Modeling from Sparse Point Cloud
Frames | To be published in IEEE Geoscience and Remote Sensing Letters (GRSL) | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing point cloud modeling datasets primarily express the modeling
precision by pose or trajectory precision rather than the point cloud modeling
effect itself. Under this demand, we first independently construct a set of
LiDAR system with an optical stage, and then we build a HPMB dataset based on
the constructed LiDAR system, a High-Precision, Multi-Beam, real-world dataset.
Second, we propose an modeling evaluation method based on HPMB for object-level
modeling to overcome this limitation. In addition, the existing point cloud
modeling methods tend to generate continuous skeletons of the global
environment, hence lacking attention to the shape of complex objects. To tackle
this challenge, we propose a novel learning-based joint framework, DSMNet, for
high-precision 3D surface modeling from sparse point cloud frames. DSMNet
comprises density-aware Point Cloud Registration (PCR) and geometry-aware Point
Cloud Sampling (PCS) to effectively learn the implicit structure feature of
sparse point clouds. Extensive experiments demonstrate that DSMNet outperforms
the state-of-the-art methods in PCS and PCR on Multi-View Partial Point Cloud
(MVP) database. Furthermore, the experiments on the open source KITTI and our
proposed HPMB datasets show that DSMNet can be generalized as a post-processing
of Simultaneous Localization And Mapping (SLAM), thereby improving modeling
precision in environments with sparse point clouds.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 09:23:06 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Qiu",
"Changjie",
""
],
[
"Wang",
"Zhiyong",
""
],
[
"Lin",
"Xiuhong",
""
],
[
"Zang",
"Yu",
""
],
[
"Wang",
"Cheng",
""
],
[
"Liu",
"Weiquan",
""
]
] | TITLE: DSMNet: Deep High-precision 3D Surface Modeling from Sparse Point Cloud
Frames
ABSTRACT: Existing point cloud modeling datasets primarily express the modeling
precision by pose or trajectory precision rather than the point cloud modeling
effect itself. Under this demand, we first independently construct a set of
LiDAR system with an optical stage, and then we build a HPMB dataset based on
the constructed LiDAR system, a High-Precision, Multi-Beam, real-world dataset.
Second, we propose an modeling evaluation method based on HPMB for object-level
modeling to overcome this limitation. In addition, the existing point cloud
modeling methods tend to generate continuous skeletons of the global
environment, hence lacking attention to the shape of complex objects. To tackle
this challenge, we propose a novel learning-based joint framework, DSMNet, for
high-precision 3D surface modeling from sparse point cloud frames. DSMNet
comprises density-aware Point Cloud Registration (PCR) and geometry-aware Point
Cloud Sampling (PCS) to effectively learn the implicit structure feature of
sparse point clouds. Extensive experiments demonstrate that DSMNet outperforms
the state-of-the-art methods in PCS and PCR on Multi-View Partial Point Cloud
(MVP) database. Furthermore, the experiments on the open source KITTI and our
proposed HPMB datasets show that DSMNet can be generalized as a post-processing
of Simultaneous Localization And Mapping (SLAM), thereby improving modeling
precision in environments with sparse point clouds.
|
2304.04205 | Jiawei Feng | Jiawei Feng and Ancong Wu and Wei-Shi Zheng | Shape-Erased Feature Learning for Visible-Infrared Person
Re-Identification | CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the modality gap between visible and infrared images with high visual
ambiguity, learning \textbf{diverse} modality-shared semantic concepts for
visible-infrared person re-identification (VI-ReID) remains a challenging
problem. Body shape is one of the significant modality-shared cues for VI-ReID.
To dig more diverse modality-shared cues, we expect that erasing
body-shape-related semantic concepts in the learned features can force the ReID
model to extract more and other modality-shared features for identification. To
this end, we propose shape-erased feature learning paradigm that decorrelates
modality-shared features in two orthogonal subspaces. Jointly learning
shape-related feature in one subspace and shape-erased features in the
orthogonal complement achieves a conditional mutual information maximization
between shape-erased feature and identity discarding body shape information,
thus enhancing the diversity of the learned representation explicitly.
Extensive experiments on SYSU-MM01, RegDB, and HITSZ-VCM datasets demonstrate
the effectiveness of our method.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 10:22:10 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Feng",
"Jiawei",
""
],
[
"Wu",
"Ancong",
""
],
[
"Zheng",
"Wei-Shi",
""
]
] | TITLE: Shape-Erased Feature Learning for Visible-Infrared Person
Re-Identification
ABSTRACT: Due to the modality gap between visible and infrared images with high visual
ambiguity, learning \textbf{diverse} modality-shared semantic concepts for
visible-infrared person re-identification (VI-ReID) remains a challenging
problem. Body shape is one of the significant modality-shared cues for VI-ReID.
To dig more diverse modality-shared cues, we expect that erasing
body-shape-related semantic concepts in the learned features can force the ReID
model to extract more and other modality-shared features for identification. To
this end, we propose shape-erased feature learning paradigm that decorrelates
modality-shared features in two orthogonal subspaces. Jointly learning
shape-related feature in one subspace and shape-erased features in the
orthogonal complement achieves a conditional mutual information maximization
between shape-erased feature and identity discarding body shape information,
thus enhancing the diversity of the learned representation explicitly.
Extensive experiments on SYSU-MM01, RegDB, and HITSZ-VCM datasets demonstrate
the effectiveness of our method.
|
2304.04211 | Jian Shi | Jian Shi and Ni Zhang | AGAD: Adversarial Generative Anomaly Detection | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomaly detection suffered from the lack of anomalies due to the diversity of
abnormalities and the difficulties of obtaining large-scale anomaly data.
Semi-supervised anomaly detection methods are often used to solely leverage
normal data to detect abnormalities that deviated from the learnt normality
distributions. Meanwhile, given the fact that limited anomaly data can be
obtained with a minor cost in practice, some researches also investigated
anomaly detection methods under supervised scenarios with limited anomaly data.
In order to address the lack of abnormal data for robust anomaly detection, we
propose Adversarial Generative Anomaly Detection (AGAD), a self-contrast-based
anomaly detection paradigm that learns to detect anomalies by generating
\textit{contextual adversarial information} from the massive normal examples.
Essentially, our method generates pseudo-anomaly data for both supervised and
semi-supervised anomaly detection scenarios. Extensive experiments are carried
out on multiple benchmark datasets and real-world datasets, the results show
significant improvement in both supervised and semi-supervised scenarios.
Importantly, our approach is data-efficient that can boost up the detection
accuracy with no more than 5% anomalous training data.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 10:40:02 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Shi",
"Jian",
""
],
[
"Zhang",
"Ni",
""
]
] | TITLE: AGAD: Adversarial Generative Anomaly Detection
ABSTRACT: Anomaly detection suffered from the lack of anomalies due to the diversity of
abnormalities and the difficulties of obtaining large-scale anomaly data.
Semi-supervised anomaly detection methods are often used to solely leverage
normal data to detect abnormalities that deviated from the learnt normality
distributions. Meanwhile, given the fact that limited anomaly data can be
obtained with a minor cost in practice, some researches also investigated
anomaly detection methods under supervised scenarios with limited anomaly data.
In order to address the lack of abnormal data for robust anomaly detection, we
propose Adversarial Generative Anomaly Detection (AGAD), a self-contrast-based
anomaly detection paradigm that learns to detect anomalies by generating
\textit{contextual adversarial information} from the massive normal examples.
Essentially, our method generates pseudo-anomaly data for both supervised and
semi-supervised anomaly detection scenarios. Extensive experiments are carried
out on multiple benchmark datasets and real-world datasets, the results show
significant improvement in both supervised and semi-supervised scenarios.
Importantly, our approach is data-efficient that can boost up the detection
accuracy with no more than 5% anomalous training data.
|
2304.04212 | David Beauchemin | David Beauchemin and Richard Khoury | RISC: Generating Realistic Synthetic Bilingual Insurance Contract | Accepted at Canadian AI conference 2023 | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper presents RISC, an open-source Python package data generator
(https://github.com/GRAAL-Research/risc). RISC generates look-alike automobile
insurance contracts based on the Quebec regulatory insurance form in French and
English. Insurance contracts are 90 to 100 pages long and use complex legal and
insurance-specific vocabulary for a layperson. Hence, they are a much more
complex class of documents than those in traditional NLP corpora. Therefore, we
introduce RISCBAC, a Realistic Insurance Synthetic Bilingual Automobile
Contract dataset based on the mandatory Quebec car insurance contract. The
dataset comprises 10,000 French and English unannotated insurance contracts.
RISCBAC enables NLP research for unsupervised automatic summarisation, question
answering, text simplification, machine translation and more. Moreover, it can
be further automatically annotated as a dataset for supervised tasks such as
NER
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 10:42:18 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Beauchemin",
"David",
""
],
[
"Khoury",
"Richard",
""
]
] | TITLE: RISC: Generating Realistic Synthetic Bilingual Insurance Contract
ABSTRACT: This paper presents RISC, an open-source Python package data generator
(https://github.com/GRAAL-Research/risc). RISC generates look-alike automobile
insurance contracts based on the Quebec regulatory insurance form in French and
English. Insurance contracts are 90 to 100 pages long and use complex legal and
insurance-specific vocabulary for a layperson. Hence, they are a much more
complex class of documents than those in traditional NLP corpora. Therefore, we
introduce RISCBAC, a Realistic Insurance Synthetic Bilingual Automobile
Contract dataset based on the mandatory Quebec car insurance contract. The
dataset comprises 10,000 French and English unannotated insurance contracts.
RISCBAC enables NLP research for unsupervised automatic summarisation, question
answering, text simplification, machine translation and more. Moreover, it can
be further automatically annotated as a dataset for supervised tasks such as
NER
|
2304.04222 | Xuanqi Gao | Xuanqi Gao, Juan Zhai, Shiqing Ma, Chao Shen, Yufei Chen, Shiwei Wang | CILIATE: Towards Fairer Class-based Incremental Learning by Dataset and
Training Refinement | null | null | null | null | cs.LG cs.AI cs.CY cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the model aging problem, Deep Neural Networks (DNNs) need updates to
adjust them to new data distributions. The common practice leverages
incremental learning (IL), e.g., Class-based Incremental Learning (CIL) that
updates output labels, to update the model with new data and a limited number
of old data. This avoids heavyweight training (from scratch) using conventional
methods and saves storage space by reducing the number of old data to store.
But it also leads to poor performance in fairness. In this paper, we show that
CIL suffers both dataset and algorithm bias problems, and existing solutions
can only partially solve the problem. We propose a novel framework, CILIATE,
that fixes both dataset and algorithm bias in CIL. It features a novel
differential analysis guided dataset and training refinement process that
identifies unique and important samples overlooked by existing CIL and enforces
the model to learn from them. Through this process, CILIATE improves the
fairness of CIL by 17.03%, 22.46%, and 31.79% compared to state-of-the-art
methods, iCaRL, BiC, and WA, respectively, based on our evaluation on three
popular datasets and widely used ResNet models.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 12:10:39 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Gao",
"Xuanqi",
""
],
[
"Zhai",
"Juan",
""
],
[
"Ma",
"Shiqing",
""
],
[
"Shen",
"Chao",
""
],
[
"Chen",
"Yufei",
""
],
[
"Wang",
"Shiwei",
""
]
] | TITLE: CILIATE: Towards Fairer Class-based Incremental Learning by Dataset and
Training Refinement
ABSTRACT: Due to the model aging problem, Deep Neural Networks (DNNs) need updates to
adjust them to new data distributions. The common practice leverages
incremental learning (IL), e.g., Class-based Incremental Learning (CIL) that
updates output labels, to update the model with new data and a limited number
of old data. This avoids heavyweight training (from scratch) using conventional
methods and saves storage space by reducing the number of old data to store.
But it also leads to poor performance in fairness. In this paper, we show that
CIL suffers both dataset and algorithm bias problems, and existing solutions
can only partially solve the problem. We propose a novel framework, CILIATE,
that fixes both dataset and algorithm bias in CIL. It features a novel
differential analysis guided dataset and training refinement process that
identifies unique and important samples overlooked by existing CIL and enforces
the model to learn from them. Through this process, CILIATE improves the
fairness of CIL by 17.03%, 22.46%, and 31.79% compared to state-of-the-art
methods, iCaRL, BiC, and WA, respectively, based on our evaluation on three
popular datasets and widely used ResNet models.
|
2304.04228 | Cong Wang | Yanru Xiao, Cong Wang, Xing Gao | Unsupervised Multi-Criteria Adversarial Detection in Deep Image
Retrieval | null | null | null | null | cs.CV cs.IR | http://creativecommons.org/licenses/by/4.0/ | The vulnerability in the algorithm supply chain of deep learning has imposed
new challenges to image retrieval systems in the downstream. Among a variety of
techniques, deep hashing is gaining popularity. As it inherits the algorithmic
backend from deep learning, a handful of attacks are recently proposed to
disrupt normal image retrieval. Unfortunately, the defense strategies in
softmax classification are not readily available to be applied in the image
retrieval domain. In this paper, we propose an efficient and unsupervised
scheme to identify unique adversarial behaviors in the hamming space. In
particular, we design three criteria from the perspectives of hamming distance,
quantization loss and denoising to defend against both untargeted and targeted
attacks, which collectively limit the adversarial space. The extensive
experiments on four datasets demonstrate 2-23% improvements of detection rates
with minimum computational overhead for real-time image queries.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 12:46:35 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Xiao",
"Yanru",
""
],
[
"Wang",
"Cong",
""
],
[
"Gao",
"Xing",
""
]
] | TITLE: Unsupervised Multi-Criteria Adversarial Detection in Deep Image
Retrieval
ABSTRACT: The vulnerability in the algorithm supply chain of deep learning has imposed
new challenges to image retrieval systems in the downstream. Among a variety of
techniques, deep hashing is gaining popularity. As it inherits the algorithmic
backend from deep learning, a handful of attacks are recently proposed to
disrupt normal image retrieval. Unfortunately, the defense strategies in
softmax classification are not readily available to be applied in the image
retrieval domain. In this paper, we propose an efficient and unsupervised
scheme to identify unique adversarial behaviors in the hamming space. In
particular, we design three criteria from the perspectives of hamming distance,
quantization loss and denoising to defend against both untargeted and targeted
attacks, which collectively limit the adversarial space. The extensive
experiments on four datasets demonstrate 2-23% improvements of detection rates
with minimum computational overhead for real-time image queries.
|
2304.04231 | Dingkang Liang | Dingkang Liang, Jiahao Xie, Zhikang Zou, Xiaoqing Ye, Wei Xu, Xiang
Bai | CrowdCLIP: Unsupervised Crowd Counting via Vision-Language Model | Accepted by CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Supervised crowd counting relies heavily on costly manual labeling, which is
difficult and expensive, especially in dense scenes. To alleviate the problem,
we propose a novel unsupervised framework for crowd counting, named CrowdCLIP.
The core idea is built on two observations: 1) the recent contrastive
pre-trained vision-language model (CLIP) has presented impressive performance
on various downstream tasks; 2) there is a natural mapping between crowd
patches and count text. To the best of our knowledge, CrowdCLIP is the first to
investigate the vision language knowledge to solve the counting problem.
Specifically, in the training stage, we exploit the multi-modal ranking loss by
constructing ranking text prompts to match the size-sorted crowd patches to
guide the image encoder learning. In the testing stage, to deal with the
diversity of image patches, we propose a simple yet effective progressive
filtering strategy to first select the highly potential crowd patches and then
map them into the language space with various counting intervals. Extensive
experiments on five challenging datasets demonstrate that the proposed
CrowdCLIP achieves superior performance compared to previous unsupervised
state-of-the-art counting methods. Notably, CrowdCLIP even surpasses some
popular fully-supervised methods under the cross-dataset setting. The source
code will be available at https://github.com/dk-liang/CrowdCLIP.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 12:56:54 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Liang",
"Dingkang",
""
],
[
"Xie",
"Jiahao",
""
],
[
"Zou",
"Zhikang",
""
],
[
"Ye",
"Xiaoqing",
""
],
[
"Xu",
"Wei",
""
],
[
"Bai",
"Xiang",
""
]
] | TITLE: CrowdCLIP: Unsupervised Crowd Counting via Vision-Language Model
ABSTRACT: Supervised crowd counting relies heavily on costly manual labeling, which is
difficult and expensive, especially in dense scenes. To alleviate the problem,
we propose a novel unsupervised framework for crowd counting, named CrowdCLIP.
The core idea is built on two observations: 1) the recent contrastive
pre-trained vision-language model (CLIP) has presented impressive performance
on various downstream tasks; 2) there is a natural mapping between crowd
patches and count text. To the best of our knowledge, CrowdCLIP is the first to
investigate the vision language knowledge to solve the counting problem.
Specifically, in the training stage, we exploit the multi-modal ranking loss by
constructing ranking text prompts to match the size-sorted crowd patches to
guide the image encoder learning. In the testing stage, to deal with the
diversity of image patches, we propose a simple yet effective progressive
filtering strategy to first select the highly potential crowd patches and then
map them into the language space with various counting intervals. Extensive
experiments on five challenging datasets demonstrate that the proposed
CrowdCLIP achieves superior performance compared to previous unsupervised
state-of-the-art counting methods. Notably, CrowdCLIP even surpasses some
popular fully-supervised methods under the cross-dataset setting. The source
code will be available at https://github.com/dk-liang/CrowdCLIP.
|
2304.04237 | Xuran Pan | Xuran Pan, Tianzhu Ye, Zhuofan Xia, Shiji Song, Gao Huang | Slide-Transformer: Hierarchical Vision Transformer with Local
Self-Attention | Accepted to CVPR2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Self-attention mechanism has been a key factor in the recent progress of
Vision Transformer (ViT), which enables adaptive feature extraction from global
contexts. However, existing self-attention methods either adopt sparse global
attention or window attention to reduce the computation complexity, which may
compromise the local feature learning or subject to some handcrafted designs.
In contrast, local attention, which restricts the receptive field of each query
to its own neighboring pixels, enjoys the benefits of both convolution and
self-attention, namely local inductive bias and dynamic feature selection.
Nevertheless, current local attention modules either use inefficient Im2Col
function or rely on specific CUDA kernels that are hard to generalize to
devices without CUDA support. In this paper, we propose a novel local attention
module, Slide Attention, which leverages common convolution operations to
achieve high efficiency, flexibility and generalizability. Specifically, we
first re-interpret the column-based Im2Col function from a new row-based
perspective and use Depthwise Convolution as an efficient substitution. On this
basis, we propose a deformed shifting module based on the re-parameterization
technique, which further relaxes the fixed key/value positions to deformed
features in the local region. In this way, our module realizes the local
attention paradigm in both efficient and flexible manner. Extensive experiments
show that our slide attention module is applicable to a variety of advanced
Vision Transformer models and compatible with various hardware devices, and
achieves consistently improved performances on comprehensive benchmarks. Code
is available at https://github.com/LeapLabTHU/Slide-Transformer.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 13:37:59 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Pan",
"Xuran",
""
],
[
"Ye",
"Tianzhu",
""
],
[
"Xia",
"Zhuofan",
""
],
[
"Song",
"Shiji",
""
],
[
"Huang",
"Gao",
""
]
] | TITLE: Slide-Transformer: Hierarchical Vision Transformer with Local
Self-Attention
ABSTRACT: Self-attention mechanism has been a key factor in the recent progress of
Vision Transformer (ViT), which enables adaptive feature extraction from global
contexts. However, existing self-attention methods either adopt sparse global
attention or window attention to reduce the computation complexity, which may
compromise the local feature learning or subject to some handcrafted designs.
In contrast, local attention, which restricts the receptive field of each query
to its own neighboring pixels, enjoys the benefits of both convolution and
self-attention, namely local inductive bias and dynamic feature selection.
Nevertheless, current local attention modules either use inefficient Im2Col
function or rely on specific CUDA kernels that are hard to generalize to
devices without CUDA support. In this paper, we propose a novel local attention
module, Slide Attention, which leverages common convolution operations to
achieve high efficiency, flexibility and generalizability. Specifically, we
first re-interpret the column-based Im2Col function from a new row-based
perspective and use Depthwise Convolution as an efficient substitution. On this
basis, we propose a deformed shifting module based on the re-parameterization
technique, which further relaxes the fixed key/value positions to deformed
features in the local region. In this way, our module realizes the local
attention paradigm in both efficient and flexible manner. Extensive experiments
show that our slide attention module is applicable to a variety of advanced
Vision Transformer models and compatible with various hardware devices, and
achieves consistently improved performances on comprehensive benchmarks. Code
is available at https://github.com/LeapLabTHU/Slide-Transformer.
|
2304.04238 | Linhao Qu | Linhao Qu, Minghong Duan, Zhiwei Yang, Manning Wang, Zhijian Song | Towards Arbitrary-scale Histopathology Image Super-resolution: An
Efficient Dual-branch Framework based on Implicit Self-texture Enhancement | null | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing super-resolution models for pathology images can only work in fixed
integer magnifications and have limited performance. Though implicit neural
network-based methods have shown promising results in arbitrary-scale
super-resolution of natural images, it is not effective to directly apply them
in pathology images, because pathology images have special fine-grained image
textures different from natural images. To address this challenge, we propose a
dual-branch framework with an efficient self-texture enhancement mechanism for
arbitrary-scale super-resolution of pathology images. Extensive experiments on
two public datasets show that our method outperforms both existing fixed-scale
and arbitrary-scale algorithms. To the best of our knowledge, this is the first
work to achieve arbitrary-scale super-resolution in the field of pathology
images. Codes will be available.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 13:38:18 GMT"
}
] | 2023-04-11T00:00:00 | [
[
"Qu",
"Linhao",
""
],
[
"Duan",
"Minghong",
""
],
[
"Yang",
"Zhiwei",
""
],
[
"Wang",
"Manning",
""
],
[
"Song",
"Zhijian",
""
]
] | TITLE: Towards Arbitrary-scale Histopathology Image Super-resolution: An
Efficient Dual-branch Framework based on Implicit Self-texture Enhancement
ABSTRACT: Existing super-resolution models for pathology images can only work in fixed
integer magnifications and have limited performance. Though implicit neural
network-based methods have shown promising results in arbitrary-scale
super-resolution of natural images, it is not effective to directly apply them
in pathology images, because pathology images have special fine-grained image
textures different from natural images. To address this challenge, we propose a
dual-branch framework with an efficient self-texture enhancement mechanism for
arbitrary-scale super-resolution of pathology images. Extensive experiments on
two public datasets show that our method outperforms both existing fixed-scale
and arbitrary-scale algorithms. To the best of our knowledge, this is the first
work to achieve arbitrary-scale super-resolution in the field of pathology
images. Codes will be available.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.