id
large_string | submitter
large_string | authors
large_string | title
large_string | comments
large_string | journal-ref
large_string | doi
large_string | report-no
large_string | categories
large_string | license
large_string | abstract
large_string | versions
large list | update_date
timestamp[ms] | authors_parsed
large list | prompt
large_string | label
string | prob
float64 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2209.14727 | Khloud Al Jallad | Khloud Al Jallad | FastPacket: Towards Pre-trained Packets Embedding based on FastText for
next-generation NIDS | arXiv admin note: text overlap with arXiv:2209.13961 | (2023), J Curr Trends Comp Sci Res, 2(1), 40-44 | 10.33140/JCTCSR | null | cs.CR cs.LG cs.NI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | New Attacks are increasingly used by attackers everyday but many of them are
not detected by Intrusion Detection Systems as most IDS ignore raw packet
information and only care about some basic statistical information extracted
from PCAP files. Using networking programs to extract fixed statistical
features from packets is good, but may not enough to detect nowadays
challenges. We think that it is time to utilize big data and deep learning for
automatic dynamic feature extraction from packets. It is time to get inspired
by deep learning pre-trained models in computer vision and natural language
processing, so security deep learning solutions will have its pre-trained
models on big datasets to be used in future researches. In this paper, we
proposed a new approach for embedding packets based on character-level
embeddings, inspired by FastText success on text data. We called this approach
FastPacket. Results are measured on subsets of CIC-IDS-2017 dataset, but we
expect promising results on big data pre-trained models. We suggest building
pre-trained FastPacket on MAWI big dataset and make it available to community,
similar to FastText. To be able to outperform currently used NIDS, to start a
new era of packet-level NIDS that can better detect complex attacks.
| [
{
"version": "v1",
"created": "Thu, 29 Sep 2022 12:47:11 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Jallad",
"Khloud Al",
""
]
] | TITLE: FastPacket: Towards Pre-trained Packets Embedding based on FastText for
next-generation NIDS
ABSTRACT: New Attacks are increasingly used by attackers everyday but many of them are
not detected by Intrusion Detection Systems as most IDS ignore raw packet
information and only care about some basic statistical information extracted
from PCAP files. Using networking programs to extract fixed statistical
features from packets is good, but may not enough to detect nowadays
challenges. We think that it is time to utilize big data and deep learning for
automatic dynamic feature extraction from packets. It is time to get inspired
by deep learning pre-trained models in computer vision and natural language
processing, so security deep learning solutions will have its pre-trained
models on big datasets to be used in future researches. In this paper, we
proposed a new approach for embedding packets based on character-level
embeddings, inspired by FastText success on text data. We called this approach
FastPacket. Results are measured on subsets of CIC-IDS-2017 dataset, but we
expect promising results on big data pre-trained models. We suggest building
pre-trained FastPacket on MAWI big dataset and make it available to community,
similar to FastText. To be able to outperform currently used NIDS, to start a
new era of packet-level NIDS that can better detect complex attacks.
| no_new_dataset | 0.945298 |
2210.00413 | Yi-Jun Cao | Yijun Cao, Xianshi Zhang, Fuya Luo, Chuan Lin, and Yongjie Li | Unsupervised Visual Odometry and Action Integration for PointGoal
Navigation in Indoor Environment | 12 pages, 6 figures | IEEE Transactions on Circuits and Systems for Video Technology,
2023 | 10.1109/TCSVT.2023.3263484 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | PointGoal navigation in indoor environment is a fundamental task for personal
robots to navigate to a specified point. Recent studies solved this PointGoal
navigation task with near-perfect success rate in photo-realistically simulated
environments, under the assumptions with noiseless actuation and most
importantly, perfect localization with GPS and compass sensors. However,
accurate GPS signalis difficult to be obtained in real indoor environment. To
improve the PointGoal navigation accuracy without GPS signal, we use visual
odometry (VO) and propose a novel action integration module (AIM) trained in
unsupervised manner. Sepecifically, unsupervised VO computes the relative pose
of the agent from the re-projection error of two adjacent frames, and then
replaces the accurate GPS signal with the path integration. The pseudo position
estimated by VO is used to train action integration which assists agent to
update their internal perception of location and helps improve the success rate
of navigation. The training and inference process only use RGB, depth,
collision as well as self-action information. The experiments show that the
proposed system achieves satisfactory results and outperforms the partially
supervised learning algorithms on the popular Gibson dataset.
| [
{
"version": "v1",
"created": "Sun, 2 Oct 2022 03:12:03 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 09:18:10 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Cao",
"Yijun",
""
],
[
"Zhang",
"Xianshi",
""
],
[
"Luo",
"Fuya",
""
],
[
"Lin",
"Chuan",
""
],
[
"Li",
"Yongjie",
""
]
] | TITLE: Unsupervised Visual Odometry and Action Integration for PointGoal
Navigation in Indoor Environment
ABSTRACT: PointGoal navigation in indoor environment is a fundamental task for personal
robots to navigate to a specified point. Recent studies solved this PointGoal
navigation task with near-perfect success rate in photo-realistically simulated
environments, under the assumptions with noiseless actuation and most
importantly, perfect localization with GPS and compass sensors. However,
accurate GPS signalis difficult to be obtained in real indoor environment. To
improve the PointGoal navigation accuracy without GPS signal, we use visual
odometry (VO) and propose a novel action integration module (AIM) trained in
unsupervised manner. Sepecifically, unsupervised VO computes the relative pose
of the agent from the re-projection error of two adjacent frames, and then
replaces the accurate GPS signal with the path integration. The pseudo position
estimated by VO is used to train action integration which assists agent to
update their internal perception of location and helps improve the success rate
of navigation. The training and inference process only use RGB, depth,
collision as well as self-action information. The experiments show that the
proposed system achieves satisfactory results and outperforms the partially
supervised learning algorithms on the popular Gibson dataset.
| no_new_dataset | 0.948394 |
2210.00935 | Gijs Bellaard | Gijs Bellaard, Daan L. J. Bon, Gautam Pai, Bart M. N. Smets, Remco
Duits | Analysis of (sub-)Riemannian PDE-G-CNNs | 29 pages, 21 figures | null | null | null | cs.LG cs.CV math.DG | http://creativecommons.org/licenses/by/4.0/ | Group equivariant convolutional neural networks (G-CNNs) have been
successfully applied in geometric deep learning. Typically, G-CNNs have the
advantage over CNNs that they do not waste network capacity on training
symmetries that should have been hard-coded in the network. The recently
introduced framework of PDE-based G-CNNs (PDE-G-CNNs) generalises G-CNNs.
PDE-G-CNNs have the core advantages that they simultaneously 1) reduce network
complexity, 2) increase classification performance, and 3) provide geometric
interpretability. Their implementations primarily consist of linear and
morphological convolutions with kernels.
In this paper we show that the previously suggested approximative
morphological kernels do not always accurately approximate the exact kernels
accurately. More specifically, depending on the spatial anisotropy of the
Riemannian metric, we argue that one must resort to sub-Riemannian
approximations. We solve this problem by providing a new approximative kernel
that works regardless of the anisotropy. We provide new theorems with better
error estimates of the approximative kernels, and prove that they all carry the
same reflectional symmetries as the exact ones.
We test the effectiveness of multiple approximative kernels within the
PDE-G-CNN framework on two datasets, and observe an improvement with the new
approximative kernels. We report that the PDE-G-CNNs again allow for a
considerable reduction of network complexity while having comparable or better
performance than G-CNNs and CNNs on the two datasets. Moreover, PDE-G-CNNs have
the advantage of better geometric interpretability over G-CNNs, as the
morphological kernels are related to association fields from neurogeometry.
| [
{
"version": "v1",
"created": "Mon, 3 Oct 2022 13:47:47 GMT"
},
{
"version": "v2",
"created": "Fri, 21 Oct 2022 16:34:13 GMT"
},
{
"version": "v3",
"created": "Fri, 25 Nov 2022 14:24:10 GMT"
},
{
"version": "v4",
"created": "Mon, 3 Apr 2023 11:40:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Bellaard",
"Gijs",
""
],
[
"Bon",
"Daan L. J.",
""
],
[
"Pai",
"Gautam",
""
],
[
"Smets",
"Bart M. N.",
""
],
[
"Duits",
"Remco",
""
]
] | TITLE: Analysis of (sub-)Riemannian PDE-G-CNNs
ABSTRACT: Group equivariant convolutional neural networks (G-CNNs) have been
successfully applied in geometric deep learning. Typically, G-CNNs have the
advantage over CNNs that they do not waste network capacity on training
symmetries that should have been hard-coded in the network. The recently
introduced framework of PDE-based G-CNNs (PDE-G-CNNs) generalises G-CNNs.
PDE-G-CNNs have the core advantages that they simultaneously 1) reduce network
complexity, 2) increase classification performance, and 3) provide geometric
interpretability. Their implementations primarily consist of linear and
morphological convolutions with kernels.
In this paper we show that the previously suggested approximative
morphological kernels do not always accurately approximate the exact kernels
accurately. More specifically, depending on the spatial anisotropy of the
Riemannian metric, we argue that one must resort to sub-Riemannian
approximations. We solve this problem by providing a new approximative kernel
that works regardless of the anisotropy. We provide new theorems with better
error estimates of the approximative kernels, and prove that they all carry the
same reflectional symmetries as the exact ones.
We test the effectiveness of multiple approximative kernels within the
PDE-G-CNN framework on two datasets, and observe an improvement with the new
approximative kernels. We report that the PDE-G-CNNs again allow for a
considerable reduction of network complexity while having comparable or better
performance than G-CNNs and CNNs on the two datasets. Moreover, PDE-G-CNNs have
the advantage of better geometric interpretability over G-CNNs, as the
morphological kernels are related to association fields from neurogeometry.
| no_new_dataset | 0.952175 |
2210.01115 | Adrian Bulat | Adrian Bulat and Georgios Tzimiropoulos | LASP: Text-to-Text Optimization for Language-Aware Soft Prompting of
Vision & Language Models | Accepted at CVPR 2023 | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Soft prompt learning has recently emerged as one of the methods of choice for
adapting V&L models to a downstream task using a few training examples.
However, current methods significantly overfit the training data, suffering
from large accuracy degradation when tested on unseen classes from the same
domain. To this end, in this paper, we make the following 4 contributions: (1)
To alleviate base class overfitting, we propose a novel Language-Aware Soft
Prompting (LASP) learning method by means of a text-to-text cross-entropy loss
that maximizes the probability of the learned prompts to be correctly
classified with respect to pre-defined hand-crafted textual prompts. (2) To
increase the representation capacity of the prompts, we propose grouped LASP
where each group of prompts is optimized with respect to a separate subset of
textual prompts. (3) We identify a visual-language misalignment introduced by
prompt learning and LASP, and more importantly, propose a re-calibration
mechanism to address it. (4) We show that LASP is inherently amenable to
including, during training, virtual classes, i.e. class names for which no
visual samples are available, further increasing the robustness of the learned
prompts. Through evaluations on 11 datasets, we show that our approach (a)
significantly outperforms all prior works on soft prompting, and (b) matches
and surpasses, for the first time, the accuracy on novel classes obtained by
hand-crafted prompts and CLIP for 8 out of 11 test datasets. Code will be made
available at https://www.adrianbulat.com/lasp
| [
{
"version": "v1",
"created": "Mon, 3 Oct 2022 17:56:35 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Apr 2023 18:03:06 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Bulat",
"Adrian",
""
],
[
"Tzimiropoulos",
"Georgios",
""
]
] | TITLE: LASP: Text-to-Text Optimization for Language-Aware Soft Prompting of
Vision & Language Models
ABSTRACT: Soft prompt learning has recently emerged as one of the methods of choice for
adapting V&L models to a downstream task using a few training examples.
However, current methods significantly overfit the training data, suffering
from large accuracy degradation when tested on unseen classes from the same
domain. To this end, in this paper, we make the following 4 contributions: (1)
To alleviate base class overfitting, we propose a novel Language-Aware Soft
Prompting (LASP) learning method by means of a text-to-text cross-entropy loss
that maximizes the probability of the learned prompts to be correctly
classified with respect to pre-defined hand-crafted textual prompts. (2) To
increase the representation capacity of the prompts, we propose grouped LASP
where each group of prompts is optimized with respect to a separate subset of
textual prompts. (3) We identify a visual-language misalignment introduced by
prompt learning and LASP, and more importantly, propose a re-calibration
mechanism to address it. (4) We show that LASP is inherently amenable to
including, during training, virtual classes, i.e. class names for which no
visual samples are available, further increasing the robustness of the learned
prompts. Through evaluations on 11 datasets, we show that our approach (a)
significantly outperforms all prior works on soft prompting, and (b) matches
and surpasses, for the first time, the accuracy on novel classes obtained by
hand-crafted prompts and CLIP for 8 out of 11 test datasets. Code will be made
available at https://www.adrianbulat.com/lasp
| no_new_dataset | 0.950778 |
2210.01445 | Nick Pepper Dr | Nick Pepper and Marc Thomas and George De Ath and Enrico Oliver and
Richard Cannon and Richard Everson and Tim Dodwell | A Probabilistic Model for Aircraft in Climb using Monotonic Functional
Gaussian Process Emulators | null | null | 10.1098/rspa.2022.0607 | null | cs.CE | http://creativecommons.org/licenses/by/4.0/ | Ensuring vertical separation is a key means of maintaining safe separation
between aircraft in congested airspace. Aircraft trajectories are modelled in
the presence of significant epistemic uncertainty, leading to discrepancies
between observed trajectories and the predictions of deterministic models,
hampering the task of planning to ensure safe separation. In this paper a
probabilistic model is presented, for the purpose of emulating the trajectories
of aircraft in climb and bounding the uncertainty of the predicted trajectory.
A monotonic, functional representation exploits the spatio-temporal
correlations in the radar observations. Through the use of Gaussian Process
Emulators, features that parameterise the climb are mapped directly to
functional outputs, providing a fast approximation, while ensuring that the
resulting trajectory is monotonic. The model was applied as a probabilistic
digital twin for aircraft in climb and baselined against BADA, a deterministic
model widely used in industry. When applied to an unseen test dataset, the
probabilistic model was found to provide a mean prediction that was 21% more
accurate, with a 34% sharper forecast.
| [
{
"version": "v1",
"created": "Tue, 4 Oct 2022 08:14:58 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Pepper",
"Nick",
""
],
[
"Thomas",
"Marc",
""
],
[
"De Ath",
"George",
""
],
[
"Oliver",
"Enrico",
""
],
[
"Cannon",
"Richard",
""
],
[
"Everson",
"Richard",
""
],
[
"Dodwell",
"Tim",
""
]
] | TITLE: A Probabilistic Model for Aircraft in Climb using Monotonic Functional
Gaussian Process Emulators
ABSTRACT: Ensuring vertical separation is a key means of maintaining safe separation
between aircraft in congested airspace. Aircraft trajectories are modelled in
the presence of significant epistemic uncertainty, leading to discrepancies
between observed trajectories and the predictions of deterministic models,
hampering the task of planning to ensure safe separation. In this paper a
probabilistic model is presented, for the purpose of emulating the trajectories
of aircraft in climb and bounding the uncertainty of the predicted trajectory.
A monotonic, functional representation exploits the spatio-temporal
correlations in the radar observations. Through the use of Gaussian Process
Emulators, features that parameterise the climb are mapped directly to
functional outputs, providing a fast approximation, while ensuring that the
resulting trajectory is monotonic. The model was applied as a probabilistic
digital twin for aircraft in climb and baselined against BADA, a deterministic
model widely used in industry. When applied to an unseen test dataset, the
probabilistic model was found to provide a mean prediction that was 21% more
accurate, with a 34% sharper forecast.
| no_new_dataset | 0.951459 |
2210.03117 | Hanoona Bangalath Rasheed Ms | Muhammad Uzair Khattak, Hanoona Rasheed, Muhammad Maaz, Salman Khan,
Fahad Shahbaz Khan | MaPLe: Multi-modal Prompt Learning | Accepted at CVPR2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Pre-trained vision-language (V-L) models such as CLIP have shown excellent
generalization ability to downstream tasks. However, they are sensitive to the
choice of input text prompts and require careful selection of prompt templates
to perform well. Inspired by the Natural Language Processing (NLP) literature,
recent CLIP adaptation approaches learn prompts as the textual inputs to
fine-tune CLIP for downstream tasks. We note that using prompting to adapt
representations in a single branch of CLIP (language or vision) is sub-optimal
since it does not allow the flexibility to dynamically adjust both
representation spaces on a downstream task. In this work, we propose
Multi-modal Prompt Learning (MaPLe) for both vision and language branches to
improve alignment between the vision and language representations. Our design
promotes strong coupling between the vision-language prompts to ensure mutual
synergy and discourages learning independent uni-modal solutions. Further, we
learn separate prompts across different early stages to progressively model the
stage-wise feature relationships to allow rich context learning. We evaluate
the effectiveness of our approach on three representative tasks of
generalization to novel classes, new target datasets and unseen domain shifts.
Compared with the state-of-the-art method Co-CoOp, MaPLe exhibits favorable
performance and achieves an absolute gain of 3.45% on novel classes and 2.72%
on overall harmonic-mean, averaged over 11 diverse image recognition datasets.
Our code and pre-trained models are available at
https://github.com/muzairkhattak/multimodal-prompt-learning.
| [
{
"version": "v1",
"created": "Thu, 6 Oct 2022 17:59:56 GMT"
},
{
"version": "v2",
"created": "Sat, 25 Mar 2023 22:10:13 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Apr 2023 06:47:44 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Khattak",
"Muhammad Uzair",
""
],
[
"Rasheed",
"Hanoona",
""
],
[
"Maaz",
"Muhammad",
""
],
[
"Khan",
"Salman",
""
],
[
"Khan",
"Fahad Shahbaz",
""
]
] | TITLE: MaPLe: Multi-modal Prompt Learning
ABSTRACT: Pre-trained vision-language (V-L) models such as CLIP have shown excellent
generalization ability to downstream tasks. However, they are sensitive to the
choice of input text prompts and require careful selection of prompt templates
to perform well. Inspired by the Natural Language Processing (NLP) literature,
recent CLIP adaptation approaches learn prompts as the textual inputs to
fine-tune CLIP for downstream tasks. We note that using prompting to adapt
representations in a single branch of CLIP (language or vision) is sub-optimal
since it does not allow the flexibility to dynamically adjust both
representation spaces on a downstream task. In this work, we propose
Multi-modal Prompt Learning (MaPLe) for both vision and language branches to
improve alignment between the vision and language representations. Our design
promotes strong coupling between the vision-language prompts to ensure mutual
synergy and discourages learning independent uni-modal solutions. Further, we
learn separate prompts across different early stages to progressively model the
stage-wise feature relationships to allow rich context learning. We evaluate
the effectiveness of our approach on three representative tasks of
generalization to novel classes, new target datasets and unseen domain shifts.
Compared with the state-of-the-art method Co-CoOp, MaPLe exhibits favorable
performance and achieves an absolute gain of 3.45% on novel classes and 2.72%
on overall harmonic-mean, averaged over 11 diverse image recognition datasets.
Our code and pre-trained models are available at
https://github.com/muzairkhattak/multimodal-prompt-learning.
| no_new_dataset | 0.944536 |
2210.04150 | Feng Liang | Feng Liang, Bichen Wu, Xiaoliang Dai, Kunpeng Li, Yinan Zhao, Hang
Zhang, Peizhao Zhang, Peter Vajda, Diana Marculescu | Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP | CVPR 2023. Project page: https://jeff-liangf.github.io/projects/ovseg | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Open-vocabulary semantic segmentation aims to segment an image into semantic
regions according to text descriptions, which may not have been seen during
training. Recent two-stage methods first generate class-agnostic mask proposals
and then leverage pre-trained vision-language models, e.g., CLIP, to classify
masked regions. We identify the performance bottleneck of this paradigm to be
the pre-trained CLIP model, since it does not perform well on masked images. To
address this, we propose to finetune CLIP on a collection of masked image
regions and their corresponding text descriptions. We collect training data by
mining an existing image-caption dataset (e.g., COCO Captions), using CLIP to
match masked image regions to nouns in the image captions. Compared with the
more precise and manually annotated segmentation labels with fixed classes
(e.g., COCO-Stuff), we find our noisy but diverse dataset can better retain
CLIP's generalization ability. Along with finetuning the entire model, we
utilize the "blank" areas in masked images using a method we dub mask prompt
tuning. Experiments demonstrate mask prompt tuning brings significant
improvement without modifying any weights of CLIP, and it can further improve a
fully finetuned model. In particular, when trained on COCO and evaluated on
ADE20K-150, our best model achieves 29.6% mIoU, which is +8.5% higher than the
previous state-of-the-art. For the first time, open-vocabulary generalist
models match the performance of supervised specialist models in 2017 without
dataset-specific adaptations.
| [
{
"version": "v1",
"created": "Sun, 9 Oct 2022 02:57:32 GMT"
},
{
"version": "v2",
"created": "Tue, 14 Mar 2023 02:48:42 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Apr 2023 19:00:47 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liang",
"Feng",
""
],
[
"Wu",
"Bichen",
""
],
[
"Dai",
"Xiaoliang",
""
],
[
"Li",
"Kunpeng",
""
],
[
"Zhao",
"Yinan",
""
],
[
"Zhang",
"Hang",
""
],
[
"Zhang",
"Peizhao",
""
],
[
"Vajda",
"Peter",
""
],
[
"Marculescu",
"Diana",
""
]
] | TITLE: Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP
ABSTRACT: Open-vocabulary semantic segmentation aims to segment an image into semantic
regions according to text descriptions, which may not have been seen during
training. Recent two-stage methods first generate class-agnostic mask proposals
and then leverage pre-trained vision-language models, e.g., CLIP, to classify
masked regions. We identify the performance bottleneck of this paradigm to be
the pre-trained CLIP model, since it does not perform well on masked images. To
address this, we propose to finetune CLIP on a collection of masked image
regions and their corresponding text descriptions. We collect training data by
mining an existing image-caption dataset (e.g., COCO Captions), using CLIP to
match masked image regions to nouns in the image captions. Compared with the
more precise and manually annotated segmentation labels with fixed classes
(e.g., COCO-Stuff), we find our noisy but diverse dataset can better retain
CLIP's generalization ability. Along with finetuning the entire model, we
utilize the "blank" areas in masked images using a method we dub mask prompt
tuning. Experiments demonstrate mask prompt tuning brings significant
improvement without modifying any weights of CLIP, and it can further improve a
fully finetuned model. In particular, when trained on COCO and evaluated on
ADE20K-150, our best model achieves 29.6% mIoU, which is +8.5% higher than the
previous state-of-the-art. For the first time, open-vocabulary generalist
models match the performance of supervised specialist models in 2017 without
dataset-specific adaptations.
| no_new_dataset | 0.943034 |
2210.14515 | Li Fu | Li Fu, Siqi Li, Qingtao Li, Liping Deng, Fangzhu Li, Lu Fan, Meng
Chen, Xiaodong He | UFO2: A unified pre-training framework for online and offline speech
recognition | Accepted by ICASSP 2023 | null | null | null | eess.AS cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a Unified pre-training Framework for Online and
Offline (UFO2) Automatic Speech Recognition (ASR), which 1) simplifies the two
separate training workflows for online and offline modes into one process, and
2) improves the Word Error Rate (WER) performance with limited utterance
annotating. Specifically, we extend the conventional offline-mode
Self-Supervised Learning (SSL)-based ASR approach to a unified manner, where
the model training is conditioned on both the full-context and dynamic-chunked
inputs. To enhance the pre-trained representation model, stop-gradient
operation is applied to decouple the online-mode objectives to the quantizer.
Moreover, in both the pre-training and the downstream fine-tuning stages, joint
losses are proposed to train the unified model with full-weight sharing for the
two modes. Experimental results on the LibriSpeech dataset show that UFO2
outperforms the SSL-based baseline method by 29.7% and 18.2% relative WER
reduction in offline and online modes, respectively.
| [
{
"version": "v1",
"created": "Wed, 26 Oct 2022 06:59:02 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 08:26:37 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Fu",
"Li",
""
],
[
"Li",
"Siqi",
""
],
[
"Li",
"Qingtao",
""
],
[
"Deng",
"Liping",
""
],
[
"Li",
"Fangzhu",
""
],
[
"Fan",
"Lu",
""
],
[
"Chen",
"Meng",
""
],
[
"He",
"Xiaodong",
""
]
] | TITLE: UFO2: A unified pre-training framework for online and offline speech
recognition
ABSTRACT: In this paper, we propose a Unified pre-training Framework for Online and
Offline (UFO2) Automatic Speech Recognition (ASR), which 1) simplifies the two
separate training workflows for online and offline modes into one process, and
2) improves the Word Error Rate (WER) performance with limited utterance
annotating. Specifically, we extend the conventional offline-mode
Self-Supervised Learning (SSL)-based ASR approach to a unified manner, where
the model training is conditioned on both the full-context and dynamic-chunked
inputs. To enhance the pre-trained representation model, stop-gradient
operation is applied to decouple the online-mode objectives to the quantizer.
Moreover, in both the pre-training and the downstream fine-tuning stages, joint
losses are proposed to train the unified model with full-weight sharing for the
two modes. Experimental results on the LibriSpeech dataset show that UFO2
outperforms the SSL-based baseline method by 29.7% and 18.2% relative WER
reduction in offline and online modes, respectively.
| no_new_dataset | 0.946941 |
2210.15943 | Jongwoo Park | Jongwoo Park, Kumara Kahatapitiya, Donghyun Kim, Shivchander
Sudalairaj, Quanfu Fan, Michael S. Ryoo | Grafting Vision Transformers | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vision Transformers (ViTs) have recently become the state-of-the-art across
many computer vision tasks. In contrast to convolutional networks (CNNs), ViTs
enable global information sharing even within shallow layers of a network,
i.e., among high-resolution features. However, this perk was later overlooked
with the success of pyramid architectures such as Swin Transformer, which show
better performance-complexity trade-offs. In this paper, we present a simple
and efficient add-on component (termed GrafT) that considers global
dependencies and multi-scale information throughout the network, in both high-
and low-resolution features alike. It has the flexibility of branching out at
arbitrary depths and shares most of the parameters and computations of the
backbone. GrafT shows consistent gains over various well-known models which
includes both hybrid and pure Transformer types, both homogeneous and pyramid
structures, and various self-attention methods. In particular, it largely
benefits mobile-size models by providing high-level semantics. On the
ImageNet-1k dataset, GrafT delivers +3.9%, +1.4%, and +1.9% top-1 accuracy
improvement to DeiT-T, Swin-T, and MobileViT-XXS, respectively. Our code and
models will be made available.
| [
{
"version": "v1",
"created": "Fri, 28 Oct 2022 07:07:13 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 14:16:14 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Park",
"Jongwoo",
""
],
[
"Kahatapitiya",
"Kumara",
""
],
[
"Kim",
"Donghyun",
""
],
[
"Sudalairaj",
"Shivchander",
""
],
[
"Fan",
"Quanfu",
""
],
[
"Ryoo",
"Michael S.",
""
]
] | TITLE: Grafting Vision Transformers
ABSTRACT: Vision Transformers (ViTs) have recently become the state-of-the-art across
many computer vision tasks. In contrast to convolutional networks (CNNs), ViTs
enable global information sharing even within shallow layers of a network,
i.e., among high-resolution features. However, this perk was later overlooked
with the success of pyramid architectures such as Swin Transformer, which show
better performance-complexity trade-offs. In this paper, we present a simple
and efficient add-on component (termed GrafT) that considers global
dependencies and multi-scale information throughout the network, in both high-
and low-resolution features alike. It has the flexibility of branching out at
arbitrary depths and shares most of the parameters and computations of the
backbone. GrafT shows consistent gains over various well-known models which
includes both hybrid and pure Transformer types, both homogeneous and pyramid
structures, and various self-attention methods. In particular, it largely
benefits mobile-size models by providing high-level semantics. On the
ImageNet-1k dataset, GrafT delivers +3.9%, +1.4%, and +1.9% top-1 accuracy
improvement to DeiT-T, Swin-T, and MobileViT-XXS, respectively. Our code and
models will be made available.
| no_new_dataset | 0.942981 |
2211.00895 | Jongho Choi | Jongho Choi, Kyogu Lee | Pop2Piano : Pop Audio-based Piano Cover Generation | null | null | null | null | cs.SD cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | Piano covers of pop music are enjoyed by many people. However, the task of
automatically generating piano covers of pop music is still understudied. This
is partly due to the lack of synchronized {Pop, Piano Cover} data pairs, which
made it challenging to apply the latest data-intensive deep learning-based
methods. To leverage the power of the data-driven approach, we make a large
amount of paired and synchronized {Pop, Piano Cover} data using an automated
pipeline. In this paper, we present Pop2Piano, a Transformer network that
generates piano covers given waveforms of pop music. To the best of our
knowledge, this is the first model to generate a piano cover directly from pop
audio without using melody and chord extraction modules. We show that
Pop2Piano, trained with our dataset, is capable of producing plausible piano
covers.
| [
{
"version": "v1",
"created": "Wed, 2 Nov 2022 05:42:22 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 06:02:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Choi",
"Jongho",
""
],
[
"Lee",
"Kyogu",
""
]
] | TITLE: Pop2Piano : Pop Audio-based Piano Cover Generation
ABSTRACT: Piano covers of pop music are enjoyed by many people. However, the task of
automatically generating piano covers of pop music is still understudied. This
is partly due to the lack of synchronized {Pop, Piano Cover} data pairs, which
made it challenging to apply the latest data-intensive deep learning-based
methods. To leverage the power of the data-driven approach, we make a large
amount of paired and synchronized {Pop, Piano Cover} data using an automated
pipeline. In this paper, we present Pop2Piano, a Transformer network that
generates piano covers given waveforms of pop music. To the best of our
knowledge, this is the first model to generate a piano cover directly from pop
audio without using melody and chord extraction modules. We show that
Pop2Piano, trained with our dataset, is capable of producing plausible piano
covers.
| new_dataset | 0.700946 |
2211.03387 | Jing Li | Qidan Zhu, Jing Li, Fei Yuan, Quan Gan | Temporal superimposed crossover module for effective continuous sign
language | 10 pages, 7 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ultimate goal of continuous sign language recognition(CSLR) is to
facilitate the communication between special people and normal people, which
requires a certain degree of real-time and deploy-ability of the model.
However, in the previous research on CSLR, little attention has been paid to
the real-time and deploy-ability. In order to improve the real-time and
deploy-ability of the model, this paper proposes a zero parameter, zero
computation temporal superposition crossover module(TSCM), and combines it with
2D convolution to form a "TSCM+2D convolution" hybrid convolution, which
enables 2D convolution to have strong spatial-temporal modelling capability
with zero parameter increase and lower deployment cost compared with other
spatial-temporal convolutions. The overall CSLR model based on TSCM is built on
the improved ResBlockT network in this paper. The hybrid convolution of
"TSCM+2D convolution" is applied to the ResBlock of the ResNet network to form
the new ResBlockT, and random gradient stop and multi-level CTC loss are
introduced to train the model, which reduces the final recognition WER while
reducing the training memory usage, and extends the ResNet network from image
classification task to video recognition task. In addition, this study is the
first in CSLR to use only 2D convolution extraction of sign language video
temporal-spatial features for end-to-end learning for recognition. Experiments
on two large-scale continuous sign language datasets demonstrate the
effectiveness of the proposed method and achieve highly competitive results.
| [
{
"version": "v1",
"created": "Mon, 7 Nov 2022 09:33:42 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Jan 2023 03:20:44 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Apr 2023 10:34:13 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Zhu",
"Qidan",
""
],
[
"Li",
"Jing",
""
],
[
"Yuan",
"Fei",
""
],
[
"Gan",
"Quan",
""
]
] | TITLE: Temporal superimposed crossover module for effective continuous sign
language
ABSTRACT: The ultimate goal of continuous sign language recognition(CSLR) is to
facilitate the communication between special people and normal people, which
requires a certain degree of real-time and deploy-ability of the model.
However, in the previous research on CSLR, little attention has been paid to
the real-time and deploy-ability. In order to improve the real-time and
deploy-ability of the model, this paper proposes a zero parameter, zero
computation temporal superposition crossover module(TSCM), and combines it with
2D convolution to form a "TSCM+2D convolution" hybrid convolution, which
enables 2D convolution to have strong spatial-temporal modelling capability
with zero parameter increase and lower deployment cost compared with other
spatial-temporal convolutions. The overall CSLR model based on TSCM is built on
the improved ResBlockT network in this paper. The hybrid convolution of
"TSCM+2D convolution" is applied to the ResBlock of the ResNet network to form
the new ResBlockT, and random gradient stop and multi-level CTC loss are
introduced to train the model, which reduces the final recognition WER while
reducing the training memory usage, and extends the ResNet network from image
classification task to video recognition task. In addition, this study is the
first in CSLR to use only 2D convolution extraction of sign language video
temporal-spatial features for end-to-end learning for recognition. Experiments
on two large-scale continuous sign language datasets demonstrate the
effectiveness of the proposed method and achieve highly competitive results.
| no_new_dataset | 0.949902 |
2211.05776 | Lu Qi | Lu Qi, Jason Kuen, Weidong Guo, Tiancheng Shen, Jiuxiang Gu, Jiaya
Jia, Zhe Lin, Ming-Hsuan Yang | High-Quality Entity Segmentation | The project webiste: http://luqi.info/entityv2.github.io/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Dense image segmentation tasks e.g., semantic, panoptic) are useful for image
editing, but existing methods can hardly generalize well in an in-the-wild
setting where there are unrestricted image domains, classes, and image
resolution and quality variations. Motivated by these observations, we
construct a new entity segmentation dataset, with a strong focus on
high-quality dense segmentation in the wild. The dataset contains images
spanning diverse image domains and entities, along with plentiful
high-resolution images and high-quality mask annotations for training and
testing. Given the high-quality and -resolution nature of the dataset, we
propose CropFormer which is designed to tackle the intractability of
instance-level segmentation on high-resolution images. It improves mask
prediction by fusing high-res image crops that provide more fine-grained image
details and the full image. CropFormer is the first query-based Transformer
architecture that can effectively fuse mask predictions from multiple image
views, by learning queries that effectively associate the same entities across
the full image and its crop. With CropFormer, we achieve a significant AP gain
of $1.9$ on the challenging entity segmentation task. Furthermore, CropFormer
consistently improves the accuracy of traditional segmentation tasks and
datasets. The dataset and code will be released at
http://luqi.info/entityv2.github.io/.
| [
{
"version": "v1",
"created": "Thu, 10 Nov 2022 18:58:22 GMT"
},
{
"version": "v2",
"created": "Sat, 12 Nov 2022 04:10:32 GMT"
},
{
"version": "v3",
"created": "Sun, 2 Apr 2023 22:01:17 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Qi",
"Lu",
""
],
[
"Kuen",
"Jason",
""
],
[
"Guo",
"Weidong",
""
],
[
"Shen",
"Tiancheng",
""
],
[
"Gu",
"Jiuxiang",
""
],
[
"Jia",
"Jiaya",
""
],
[
"Lin",
"Zhe",
""
],
[
"Yang",
"Ming-Hsuan",
""
]
] | TITLE: High-Quality Entity Segmentation
ABSTRACT: Dense image segmentation tasks e.g., semantic, panoptic) are useful for image
editing, but existing methods can hardly generalize well in an in-the-wild
setting where there are unrestricted image domains, classes, and image
resolution and quality variations. Motivated by these observations, we
construct a new entity segmentation dataset, with a strong focus on
high-quality dense segmentation in the wild. The dataset contains images
spanning diverse image domains and entities, along with plentiful
high-resolution images and high-quality mask annotations for training and
testing. Given the high-quality and -resolution nature of the dataset, we
propose CropFormer which is designed to tackle the intractability of
instance-level segmentation on high-resolution images. It improves mask
prediction by fusing high-res image crops that provide more fine-grained image
details and the full image. CropFormer is the first query-based Transformer
architecture that can effectively fuse mask predictions from multiple image
views, by learning queries that effectively associate the same entities across
the full image and its crop. With CropFormer, we achieve a significant AP gain
of $1.9$ on the challenging entity segmentation task. Furthermore, CropFormer
consistently improves the accuracy of traditional segmentation tasks and
datasets. The dataset and code will be released at
http://luqi.info/entityv2.github.io/.
| new_dataset | 0.957912 |
2211.07816 | Shuqi Ke | Shuqi Ke, Chao Huang, Xin Liu | Quantifying the Impact of Label Noise on Federated Learning | Accepted by The AAAI 2023 Workshop on Representation Learning for
Responsible Human-Centric AI | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Federated Learning (FL) is a distributed machine learning paradigm where
clients collaboratively train a model using their local (human-generated)
datasets. While existing studies focus on FL algorithm development to tackle
data heterogeneity across clients, the important issue of data quality (e.g.,
label noise) in FL is overlooked. This paper aims to fill this gap by providing
a quantitative study on the impact of label noise on FL. We derive an upper
bound for the generalization error that is linear in the clients' label noise
level. Then we conduct experiments on MNIST and CIFAR-10 datasets using various
FL algorithms. Our empirical results show that the global model accuracy
linearly decreases as the noise level increases, which is consistent with our
theoretical analysis. We further find that label noise slows down the
convergence of FL training, and the global model tends to overfit when the
noise level is high.
| [
{
"version": "v1",
"created": "Tue, 15 Nov 2022 00:40:55 GMT"
},
{
"version": "v2",
"created": "Wed, 11 Jan 2023 08:14:51 GMT"
},
{
"version": "v3",
"created": "Fri, 13 Jan 2023 09:42:42 GMT"
},
{
"version": "v4",
"created": "Mon, 23 Jan 2023 12:49:42 GMT"
},
{
"version": "v5",
"created": "Sat, 28 Jan 2023 09:25:08 GMT"
},
{
"version": "v6",
"created": "Sun, 12 Feb 2023 19:08:09 GMT"
},
{
"version": "v7",
"created": "Mon, 3 Apr 2023 09:45:32 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ke",
"Shuqi",
""
],
[
"Huang",
"Chao",
""
],
[
"Liu",
"Xin",
""
]
] | TITLE: Quantifying the Impact of Label Noise on Federated Learning
ABSTRACT: Federated Learning (FL) is a distributed machine learning paradigm where
clients collaboratively train a model using their local (human-generated)
datasets. While existing studies focus on FL algorithm development to tackle
data heterogeneity across clients, the important issue of data quality (e.g.,
label noise) in FL is overlooked. This paper aims to fill this gap by providing
a quantitative study on the impact of label noise on FL. We derive an upper
bound for the generalization error that is linear in the clients' label noise
level. Then we conduct experiments on MNIST and CIFAR-10 datasets using various
FL algorithms. Our empirical results show that the global model accuracy
linearly decreases as the noise level increases, which is consistent with our
theoretical analysis. We further find that label noise slows down the
convergence of FL training, and the global model tends to overfit when the
noise level is high.
| no_new_dataset | 0.948346 |
2211.10624 | Jiaxin Deng | Jiaxin Deng, Dong Shen, Haojie Pan, Xiangyu Wu, Ximan Liu, Gaofeng
Meng, Fan Yang, Size Li, Ruiji Fu, Zhongyuan Wang | A Unified Model for Video Understanding and Knowledge Embedding with
Heterogeneous Knowledge Graph Dataset | Accepted by ICMR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Video understanding is an important task in short video business platforms
and it has a wide application in video recommendation and classification. Most
of the existing video understanding works only focus on the information that
appeared within the video content, including the video frames, audio and text.
However, introducing common sense knowledge from the external Knowledge Graph
(KG) dataset is essential for video understanding when referring to the content
which is less relevant to the video. Owing to the lack of video knowledge graph
dataset, the work which integrates video understanding and KG is rare. In this
paper, we propose a heterogeneous dataset that contains the multi-modal video
entity and fruitful common sense relations. This dataset also provides multiple
novel video inference tasks like the Video-Relation-Tag (VRT) and
Video-Relation-Video (VRV) tasks. Furthermore, based on this dataset, we
propose an end-to-end model that jointly optimizes the video understanding
objective with knowledge graph embedding, which can not only better inject
factual knowledge into video understanding but also generate effective
multi-modal entity embedding for KG. Comprehensive experiments indicate that
combining video understanding embedding with factual knowledge benefits the
content-based video retrieval performance. Moreover, it also helps the model
generate better knowledge graph embedding which outperforms traditional
KGE-based methods on VRT and VRV tasks with at least 42.36% and 17.73%
improvement in HITS@10.
| [
{
"version": "v1",
"created": "Sat, 19 Nov 2022 09:00:45 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Apr 2023 03:10:21 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Deng",
"Jiaxin",
""
],
[
"Shen",
"Dong",
""
],
[
"Pan",
"Haojie",
""
],
[
"Wu",
"Xiangyu",
""
],
[
"Liu",
"Ximan",
""
],
[
"Meng",
"Gaofeng",
""
],
[
"Yang",
"Fan",
""
],
[
"Li",
"Size",
""
],
[
"Fu",
"Ruiji",
""
],
[
"Wang",
"Zhongyuan",
""
]
] | TITLE: A Unified Model for Video Understanding and Knowledge Embedding with
Heterogeneous Knowledge Graph Dataset
ABSTRACT: Video understanding is an important task in short video business platforms
and it has a wide application in video recommendation and classification. Most
of the existing video understanding works only focus on the information that
appeared within the video content, including the video frames, audio and text.
However, introducing common sense knowledge from the external Knowledge Graph
(KG) dataset is essential for video understanding when referring to the content
which is less relevant to the video. Owing to the lack of video knowledge graph
dataset, the work which integrates video understanding and KG is rare. In this
paper, we propose a heterogeneous dataset that contains the multi-modal video
entity and fruitful common sense relations. This dataset also provides multiple
novel video inference tasks like the Video-Relation-Tag (VRT) and
Video-Relation-Video (VRV) tasks. Furthermore, based on this dataset, we
propose an end-to-end model that jointly optimizes the video understanding
objective with knowledge graph embedding, which can not only better inject
factual knowledge into video understanding but also generate effective
multi-modal entity embedding for KG. Comprehensive experiments indicate that
combining video understanding embedding with factual knowledge benefits the
content-based video retrieval performance. Moreover, it also helps the model
generate better knowledge graph embedding which outperforms traditional
KGE-based methods on VRT and VRV tasks with at least 42.36% and 17.73%
improvement in HITS@10.
| new_dataset | 0.971293 |
2211.10844 | Zheng Xu | Zheng Xu, Maxwell Collins, Yuxiao Wang, Liviu Panait, Sewoong Oh, Sean
Augenstein, Ting Liu, Florian Schroff, H. Brendan McMahan | Learning to Generate Image Embeddings with User-level Differential
Privacy | CVPR camera ready. Addressed reviewer comments. Switched from
add-or-remove-one DP to substitute-one DP | null | null | null | cs.LG cs.CR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Small on-device models have been successfully trained with user-level
differential privacy (DP) for next word prediction and image classification
tasks in the past. However, existing methods can fail when directly applied to
learn embedding models using supervised training data with a large class space.
To achieve user-level DP for large image-to-embedding feature extractors, we
propose DP-FedEmb, a variant of federated learning algorithms with per-user
sensitivity control and noise addition, to train from user-partitioned data
centralized in the datacenter. DP-FedEmb combines virtual clients, partial
aggregation, private local fine-tuning, and public pretraining to achieve
strong privacy utility trade-offs. We apply DP-FedEmb to train image embedding
models for faces, landmarks and natural species, and demonstrate its superior
utility under same privacy budget on benchmark datasets DigiFace, EMNIST, GLD
and iNaturalist. We further illustrate it is possible to achieve strong
user-level DP guarantees of $\epsilon<4$ while controlling the utility drop
within 5%, when millions of users can participate in training.
| [
{
"version": "v1",
"created": "Sun, 20 Nov 2022 01:59:37 GMT"
},
{
"version": "v2",
"created": "Fri, 31 Mar 2023 22:55:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Xu",
"Zheng",
""
],
[
"Collins",
"Maxwell",
""
],
[
"Wang",
"Yuxiao",
""
],
[
"Panait",
"Liviu",
""
],
[
"Oh",
"Sewoong",
""
],
[
"Augenstein",
"Sean",
""
],
[
"Liu",
"Ting",
""
],
[
"Schroff",
"Florian",
""
],
[
"McMahan",
"H. Brendan",
""
]
] | TITLE: Learning to Generate Image Embeddings with User-level Differential
Privacy
ABSTRACT: Small on-device models have been successfully trained with user-level
differential privacy (DP) for next word prediction and image classification
tasks in the past. However, existing methods can fail when directly applied to
learn embedding models using supervised training data with a large class space.
To achieve user-level DP for large image-to-embedding feature extractors, we
propose DP-FedEmb, a variant of federated learning algorithms with per-user
sensitivity control and noise addition, to train from user-partitioned data
centralized in the datacenter. DP-FedEmb combines virtual clients, partial
aggregation, private local fine-tuning, and public pretraining to achieve
strong privacy utility trade-offs. We apply DP-FedEmb to train image embedding
models for faces, landmarks and natural species, and demonstrate its superior
utility under same privacy budget on benchmark datasets DigiFace, EMNIST, GLD
and iNaturalist. We further illustrate it is possible to achieve strong
user-level DP guarantees of $\epsilon<4$ while controlling the utility drop
within 5%, when millions of users can participate in training.
| no_new_dataset | 0.94887 |
2211.11152 | Shengkun Tang | Shengkun Tang, Yaqing Wang, Zhenglun Kong, Tianchi Zhang, Yao Li,
Caiwen Ding, Yanzhi Wang, Yi Liang, Dongkuan Xu | You Need Multiple Exiting: Dynamic Early Exiting for Accelerating
Unified Vision Language Model | null | null | null | null | cs.CV cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Large-scale Transformer models bring significant improvements for various
downstream vision language tasks with a unified architecture. The performance
improvements come with increasing model size, resulting in slow inference speed
and increased cost for severing. While some certain predictions benefit from
the full complexity of the large-scale model, not all of inputs need the same
amount of computation to conduct, potentially leading to computation resource
waste. To handle this challenge, early exiting is proposed to adaptively
allocate computational power in term of input complexity to improve inference
efficiency. The existing early exiting strategies usually adopt output
confidence based on intermediate layers as a proxy of input complexity to incur
the decision of skipping following layers. However, such strategies cannot
apply to encoder in the widely-used unified architecture with both encoder and
decoder due to difficulty of output confidence estimation in the encoder. It is
suboptimal in term of saving computation power to ignore the early exiting in
encoder component. To handle this challenge, we propose a novel early exiting
strategy for unified visual language models, which allows dynamically skip the
layers in encoder and decoder simultaneously in term of input layer-wise
similarities with multiple times of early exiting, namely \textbf{MuE}. By
decomposing the image and text modalities in the encoder, MuE is flexible and
can skip different layers in term of modalities, advancing the inference
efficiency while minimizing performance drop. Experiments on the SNLI-VE and MS
COCO datasets show that the proposed approach MuE can reduce expected inference
time by up to 50\% and 40\% while maintaining 99\% and 96\% performance
respectively.
| [
{
"version": "v1",
"created": "Mon, 21 Nov 2022 02:32:25 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 06:41:13 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Tang",
"Shengkun",
""
],
[
"Wang",
"Yaqing",
""
],
[
"Kong",
"Zhenglun",
""
],
[
"Zhang",
"Tianchi",
""
],
[
"Li",
"Yao",
""
],
[
"Ding",
"Caiwen",
""
],
[
"Wang",
"Yanzhi",
""
],
[
"Liang",
"Yi",
""
],
[
"Xu",
"Dongkuan",
""
]
] | TITLE: You Need Multiple Exiting: Dynamic Early Exiting for Accelerating
Unified Vision Language Model
ABSTRACT: Large-scale Transformer models bring significant improvements for various
downstream vision language tasks with a unified architecture. The performance
improvements come with increasing model size, resulting in slow inference speed
and increased cost for severing. While some certain predictions benefit from
the full complexity of the large-scale model, not all of inputs need the same
amount of computation to conduct, potentially leading to computation resource
waste. To handle this challenge, early exiting is proposed to adaptively
allocate computational power in term of input complexity to improve inference
efficiency. The existing early exiting strategies usually adopt output
confidence based on intermediate layers as a proxy of input complexity to incur
the decision of skipping following layers. However, such strategies cannot
apply to encoder in the widely-used unified architecture with both encoder and
decoder due to difficulty of output confidence estimation in the encoder. It is
suboptimal in term of saving computation power to ignore the early exiting in
encoder component. To handle this challenge, we propose a novel early exiting
strategy for unified visual language models, which allows dynamically skip the
layers in encoder and decoder simultaneously in term of input layer-wise
similarities with multiple times of early exiting, namely \textbf{MuE}. By
decomposing the image and text modalities in the encoder, MuE is flexible and
can skip different layers in term of modalities, advancing the inference
efficiency while minimizing performance drop. Experiments on the SNLI-VE and MS
COCO datasets show that the proposed approach MuE can reduce expected inference
time by up to 50\% and 40\% while maintaining 99\% and 96\% performance
respectively.
| no_new_dataset | 0.95096 |
2211.11704 | Mohammad Mahdi Johari | Mohammad Mahdi Johari, Camilla Carta, Fran\c{c}ois Fleuret | ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of
Signed Distance Fields | CVPR 2023 Highlight. Project page: https://www.idiap.ch/paper/eslam/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We present ESLAM, an efficient implicit neural representation method for
Simultaneous Localization and Mapping (SLAM). ESLAM reads RGB-D frames with
unknown camera poses in a sequential manner and incrementally reconstructs the
scene representation while estimating the current camera position in the scene.
We incorporate the latest advances in Neural Radiance Fields (NeRF) into a SLAM
system, resulting in an efficient and accurate dense visual SLAM method. Our
scene representation consists of multi-scale axis-aligned perpendicular feature
planes and shallow decoders that, for each point in the continuous space,
decode the interpolated features into Truncated Signed Distance Field (TSDF)
and RGB values. Our extensive experiments on three standard datasets, Replica,
ScanNet, and TUM RGB-D show that ESLAM improves the accuracy of 3D
reconstruction and camera localization of state-of-the-art dense visual SLAM
methods by more than 50%, while it runs up to 10 times faster and does not
require any pre-training.
| [
{
"version": "v1",
"created": "Mon, 21 Nov 2022 18:25:14 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 08:54:42 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Johari",
"Mohammad Mahdi",
""
],
[
"Carta",
"Camilla",
""
],
[
"Fleuret",
"François",
""
]
] | TITLE: ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of
Signed Distance Fields
ABSTRACT: We present ESLAM, an efficient implicit neural representation method for
Simultaneous Localization and Mapping (SLAM). ESLAM reads RGB-D frames with
unknown camera poses in a sequential manner and incrementally reconstructs the
scene representation while estimating the current camera position in the scene.
We incorporate the latest advances in Neural Radiance Fields (NeRF) into a SLAM
system, resulting in an efficient and accurate dense visual SLAM method. Our
scene representation consists of multi-scale axis-aligned perpendicular feature
planes and shallow decoders that, for each point in the continuous space,
decode the interpolated features into Truncated Signed Distance Field (TSDF)
and RGB values. Our extensive experiments on three standard datasets, Replica,
ScanNet, and TUM RGB-D show that ESLAM improves the accuracy of 3D
reconstruction and camera localization of state-of-the-art dense visual SLAM
methods by more than 50%, while it runs up to 10 times faster and does not
require any pre-training.
| no_new_dataset | 0.945701 |
2211.12051 | Hao Shen | Hao Shen, Zhong-Qiu Zhao, Wandi Zhang | Adaptive Dynamic Filtering Network for Image Denoising | 9 pages, Accepted in AAAI Conference on Artificial Intelligence
(AAAI) 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In image denoising networks, feature scaling is widely used to enlarge the
receptive field size and reduce computational costs. This practice, however,
also leads to the loss of high-frequency information and fails to consider
within-scale characteristics. Recently, dynamic convolution has exhibited
powerful capabilities in processing high-frequency information (e.g., edges,
corners, textures), but previous works lack sufficient spatial contextual
information in filter generation. To alleviate these issues, we propose to
employ dynamic convolution to improve the learning of high-frequency and
multi-scale features. Specifically, we design a spatially enhanced kernel
generation (SEKG) module to improve dynamic convolution, enabling the learning
of spatial context information with a very low computational complexity. Based
on the SEKG module, we propose a dynamic convolution block (DCB) and a
multi-scale dynamic convolution block (MDCB). The former enhances the
high-frequency information via dynamic convolution and preserves low-frequency
information via skip connections. The latter utilizes shared adaptive dynamic
kernels and the idea of dilated convolution to achieve efficient multi-scale
feature extraction. The proposed multi-dimension feature integration (MFI)
mechanism further fuses the multi-scale features, providing precise and
contextually enriched feature representations. Finally, we build an efficient
denoising network with the proposed DCB and MDCB, named ADFNet. It achieves
better performance with low computational complexity on real-world and
synthetic Gaussian noisy datasets. The source code is available at
https://github.com/it-hao/ADFNet.
| [
{
"version": "v1",
"created": "Tue, 22 Nov 2022 06:54:27 GMT"
},
{
"version": "v2",
"created": "Sat, 26 Nov 2022 05:56:20 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Apr 2023 01:14:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Shen",
"Hao",
""
],
[
"Zhao",
"Zhong-Qiu",
""
],
[
"Zhang",
"Wandi",
""
]
] | TITLE: Adaptive Dynamic Filtering Network for Image Denoising
ABSTRACT: In image denoising networks, feature scaling is widely used to enlarge the
receptive field size and reduce computational costs. This practice, however,
also leads to the loss of high-frequency information and fails to consider
within-scale characteristics. Recently, dynamic convolution has exhibited
powerful capabilities in processing high-frequency information (e.g., edges,
corners, textures), but previous works lack sufficient spatial contextual
information in filter generation. To alleviate these issues, we propose to
employ dynamic convolution to improve the learning of high-frequency and
multi-scale features. Specifically, we design a spatially enhanced kernel
generation (SEKG) module to improve dynamic convolution, enabling the learning
of spatial context information with a very low computational complexity. Based
on the SEKG module, we propose a dynamic convolution block (DCB) and a
multi-scale dynamic convolution block (MDCB). The former enhances the
high-frequency information via dynamic convolution and preserves low-frequency
information via skip connections. The latter utilizes shared adaptive dynamic
kernels and the idea of dilated convolution to achieve efficient multi-scale
feature extraction. The proposed multi-dimension feature integration (MFI)
mechanism further fuses the multi-scale features, providing precise and
contextually enriched feature representations. Finally, we build an efficient
denoising network with the proposed DCB and MDCB, named ADFNet. It achieves
better performance with low computational complexity on real-world and
synthetic Gaussian noisy datasets. The source code is available at
https://github.com/it-hao/ADFNet.
| no_new_dataset | 0.948965 |
2212.04808 | Muhammad Anwaar Khalid | Muhammad Anwaar Khalid, Kanwal Zulfiqar, Ulfat Bashir, Areeba Shaheen,
Rida Iqbal, Zarnab Rizwan, Ghina Rizwan, Muhammad Moazam Fraz | CEPHA29: Automatic Cephalometric Landmark Detection Challenge 2023 | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Quantitative cephalometric analysis is the most widely used clinical and
research tool in modern orthodontics. Accurate localization of cephalometric
landmarks enables the quantification and classification of anatomical
abnormalities, however, the traditional manual way of marking these landmarks
is a very tedious job. Endeavours have constantly been made to develop
automated cephalometric landmark detection systems but they are inadequate for
orthodontic applications. The fundamental reason for this is that the amount of
publicly available datasets as well as the images provided for training in
these datasets are insufficient for an AI model to perform well. To facilitate
the development of robust AI solutions for morphometric analysis, we organise
the CEPHA29 Automatic Cephalometric Landmark Detection Challenge in conjunction
with IEEE International Symposium on Biomedical Imaging (ISBI 2023). In this
context, we provide the largest known publicly available dataset, consisting of
1000 cephalometric X-ray images. We hope that our challenge will not only
derive forward research and innovation in automatic cephalometric landmark
identification but will also signal the beginning of a new era in the
discipline.
| [
{
"version": "v1",
"created": "Fri, 9 Dec 2022 12:25:58 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 10:27:21 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Khalid",
"Muhammad Anwaar",
""
],
[
"Zulfiqar",
"Kanwal",
""
],
[
"Bashir",
"Ulfat",
""
],
[
"Shaheen",
"Areeba",
""
],
[
"Iqbal",
"Rida",
""
],
[
"Rizwan",
"Zarnab",
""
],
[
"Rizwan",
"Ghina",
""
],
[
"Fraz",
"Muhammad Moazam",
""
]
] | TITLE: CEPHA29: Automatic Cephalometric Landmark Detection Challenge 2023
ABSTRACT: Quantitative cephalometric analysis is the most widely used clinical and
research tool in modern orthodontics. Accurate localization of cephalometric
landmarks enables the quantification and classification of anatomical
abnormalities, however, the traditional manual way of marking these landmarks
is a very tedious job. Endeavours have constantly been made to develop
automated cephalometric landmark detection systems but they are inadequate for
orthodontic applications. The fundamental reason for this is that the amount of
publicly available datasets as well as the images provided for training in
these datasets are insufficient for an AI model to perform well. To facilitate
the development of robust AI solutions for morphometric analysis, we organise
the CEPHA29 Automatic Cephalometric Landmark Detection Challenge in conjunction
with IEEE International Symposium on Biomedical Imaging (ISBI 2023). In this
context, we provide the largest known publicly available dataset, consisting of
1000 cephalometric X-ray images. We hope that our challenge will not only
derive forward research and innovation in automatic cephalometric landmark
identification but will also signal the beginning of a new era in the
discipline.
| new_dataset | 0.966156 |
2212.04843 | Bruno Rossi | Martin Macak, Matus Stovcik, Tomas Rebok, Mouzhi Ge, Bruno Rossi,
Barbora Buhnova | CopAS: A Big Data Forensic Analytics System | null | null | null | null | cs.CR cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the advancing digitization of our society, network security has become
one of the critical concerns for most organizations. In this paper, we present
CopAS, a system targeted at Big Data forensics analysis, allowing network
operators to comfortably analyze and correlate large amounts of network data to
get insights about potentially malicious and suspicious events. We demonstrate
the practical usage of CopAS for insider attack detection on a publicly
available PCAP dataset and show how the system can be used to detect insiders
hiding their malicious activity in the large amounts of data streams generated
during the operations of an organization within the network.
| [
{
"version": "v1",
"created": "Fri, 9 Dec 2022 13:22:41 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 08:51:18 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Macak",
"Martin",
""
],
[
"Stovcik",
"Matus",
""
],
[
"Rebok",
"Tomas",
""
],
[
"Ge",
"Mouzhi",
""
],
[
"Rossi",
"Bruno",
""
],
[
"Buhnova",
"Barbora",
""
]
] | TITLE: CopAS: A Big Data Forensic Analytics System
ABSTRACT: With the advancing digitization of our society, network security has become
one of the critical concerns for most organizations. In this paper, we present
CopAS, a system targeted at Big Data forensics analysis, allowing network
operators to comfortably analyze and correlate large amounts of network data to
get insights about potentially malicious and suspicious events. We demonstrate
the practical usage of CopAS for insider attack detection on a publicly
available PCAP dataset and show how the system can be used to detect insiders
hiding their malicious activity in the large amounts of data streams generated
during the operations of an organization within the network.
| no_new_dataset | 0.945601 |
2212.06250 | Ahmed Abdelreheem Mr. | Ahmed Abdelreheem, Kyle Olszewski, Hsin-Ying Lee, Peter Wonka, Panos
Achlioptas | ScanEnts3D: Exploiting Phrase-to-3D-Object Correspondences for Improved
Visio-Linguistic Models in 3D Scenes | The project's webpage is https://scanents3d.github.io/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The two popular datasets ScanRefer [16] and ReferIt3D [3] connect natural
language to real-world 3D data. In this paper, we curate a large-scale and
complementary dataset extending both the aforementioned ones by associating all
objects mentioned in a referential sentence to their underlying instances
inside a 3D scene. Specifically, our Scan Entities in 3D (ScanEnts3D) dataset
provides explicit correspondences between 369k objects across 84k natural
referential sentences, covering 705 real-world scenes. Crucially, we show that
by incorporating intuitive losses that enable learning from this novel dataset,
we can significantly improve the performance of several recently introduced
neural listening architectures, including improving the SoTA in both the Nr3D
and ScanRefer benchmarks by 4.3% and 5.0%, respectively. Moreover, we
experiment with competitive baselines and recent methods for the task of
language generation and show that, as with neural listeners, 3D neural speakers
can also noticeably benefit by training with ScanEnts3D, including improving
the SoTA by 13.2 CIDEr points on the Nr3D benchmark. Overall, our carefully
conducted experimental studies strongly support the conclusion that, by
learning on ScanEnts3D, commonly used visio-linguistic 3D architectures can
become more efficient and interpretable in their generalization without needing
to provide these newly collected annotations at test time. The project's
webpage is https://scanents3d.github.io/ .
| [
{
"version": "v1",
"created": "Mon, 12 Dec 2022 21:25:58 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 12:13:27 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Abdelreheem",
"Ahmed",
""
],
[
"Olszewski",
"Kyle",
""
],
[
"Lee",
"Hsin-Ying",
""
],
[
"Wonka",
"Peter",
""
],
[
"Achlioptas",
"Panos",
""
]
] | TITLE: ScanEnts3D: Exploiting Phrase-to-3D-Object Correspondences for Improved
Visio-Linguistic Models in 3D Scenes
ABSTRACT: The two popular datasets ScanRefer [16] and ReferIt3D [3] connect natural
language to real-world 3D data. In this paper, we curate a large-scale and
complementary dataset extending both the aforementioned ones by associating all
objects mentioned in a referential sentence to their underlying instances
inside a 3D scene. Specifically, our Scan Entities in 3D (ScanEnts3D) dataset
provides explicit correspondences between 369k objects across 84k natural
referential sentences, covering 705 real-world scenes. Crucially, we show that
by incorporating intuitive losses that enable learning from this novel dataset,
we can significantly improve the performance of several recently introduced
neural listening architectures, including improving the SoTA in both the Nr3D
and ScanRefer benchmarks by 4.3% and 5.0%, respectively. Moreover, we
experiment with competitive baselines and recent methods for the task of
language generation and show that, as with neural listeners, 3D neural speakers
can also noticeably benefit by training with ScanEnts3D, including improving
the SoTA by 13.2 CIDEr points on the Nr3D benchmark. Overall, our carefully
conducted experimental studies strongly support the conclusion that, by
learning on ScanEnts3D, commonly used visio-linguistic 3D architectures can
become more efficient and interpretable in their generalization without needing
to provide these newly collected annotations at test time. The project's
webpage is https://scanents3d.github.io/ .
| new_dataset | 0.937268 |
2212.08067 | Yufan Ren | Yufan Ren, Fangjinhua Wang, Tong Zhang, Marc Pollefeys and Sabine
S\"usstrunk | VolRecon: Volume Rendering of Signed Ray Distance Functions for
Generalizable Multi-View Reconstruction | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The success of the Neural Radiance Fields (NeRF) in novel view synthesis has
inspired researchers to propose neural implicit scene reconstruction. However,
most existing neural implicit reconstruction methods optimize per-scene
parameters and therefore lack generalizability to new scenes. We introduce
VolRecon, a novel generalizable implicit reconstruction method with Signed Ray
Distance Function (SRDF). To reconstruct the scene with fine details and little
noise, VolRecon combines projection features aggregated from multi-view
features, and volume features interpolated from a coarse global feature volume.
Using a ray transformer, we compute SRDF values of sampled points on a ray and
then render color and depth. On DTU dataset, VolRecon outperforms SparseNeuS by
about 30% in sparse view reconstruction and achieves comparable accuracy as
MVSNet in full view reconstruction. Furthermore, our approach exhibits good
generalization performance on the large-scale ETH3D benchmark.
| [
{
"version": "v1",
"created": "Thu, 15 Dec 2022 18:59:54 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 06:54:50 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ren",
"Yufan",
""
],
[
"Wang",
"Fangjinhua",
""
],
[
"Zhang",
"Tong",
""
],
[
"Pollefeys",
"Marc",
""
],
[
"Süsstrunk",
"Sabine",
""
]
] | TITLE: VolRecon: Volume Rendering of Signed Ray Distance Functions for
Generalizable Multi-View Reconstruction
ABSTRACT: The success of the Neural Radiance Fields (NeRF) in novel view synthesis has
inspired researchers to propose neural implicit scene reconstruction. However,
most existing neural implicit reconstruction methods optimize per-scene
parameters and therefore lack generalizability to new scenes. We introduce
VolRecon, a novel generalizable implicit reconstruction method with Signed Ray
Distance Function (SRDF). To reconstruct the scene with fine details and little
noise, VolRecon combines projection features aggregated from multi-view
features, and volume features interpolated from a coarse global feature volume.
Using a ray transformer, we compute SRDF values of sampled points on a ray and
then render color and depth. On DTU dataset, VolRecon outperforms SparseNeuS by
about 30% in sparse view reconstruction and achieves comparable accuracy as
MVSNet in full view reconstruction. Furthermore, our approach exhibits good
generalization performance on the large-scale ETH3D benchmark.
| no_new_dataset | 0.950041 |
2301.02228 | Chaoyi Wu | Chaoyi Wu, Xiaoman Zhang, Ya Zhang, Yanfeng Wang, Weidi Xie | MedKLIP: Medical Knowledge Enhanced Language-Image Pre-Training in
Radiology | null | null | null | null | eess.IV cs.CL cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we consider enhancing medical visual-language pre-training
(VLP) with domain-specific knowledge, by exploiting the paired image-text
reports from the radiological daily practice. In particular, we make the
following contributions: First, unlike existing works that directly process the
raw reports, we adopt a novel triplet extraction module to extract the
medical-related information, avoiding unnecessary complexity from language
grammar and enhancing the supervision signals; Second, we propose a novel
triplet encoding module with entity translation by querying a knowledge base,
to exploit the rich domain knowledge in medical field, and implicitly build
relationships between medical entities in the language embedding space; Third,
we propose to use a Transformer-based fusion model for spatially aligning the
entity description with visual signals at the image patch level, enabling the
ability for medical diagnosis; Fourth, we conduct thorough experiments to
validate the effectiveness of our architecture, and benchmark on numerous
public benchmarks, e.g., ChestX-ray14, RSNA Pneumonia, SIIM-ACR Pneumothorax,
COVIDx CXR-2, COVID Rural, and EdemaSeverity. In both zero-shot and fine-tuning
settings, our model has demonstrated strong performance compared with the
former methods on disease classification and grounding.
| [
{
"version": "v1",
"created": "Thu, 5 Jan 2023 18:55:09 GMT"
},
{
"version": "v2",
"created": "Thu, 9 Mar 2023 12:45:10 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Apr 2023 09:57:51 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wu",
"Chaoyi",
""
],
[
"Zhang",
"Xiaoman",
""
],
[
"Zhang",
"Ya",
""
],
[
"Wang",
"Yanfeng",
""
],
[
"Xie",
"Weidi",
""
]
] | TITLE: MedKLIP: Medical Knowledge Enhanced Language-Image Pre-Training in
Radiology
ABSTRACT: In this paper, we consider enhancing medical visual-language pre-training
(VLP) with domain-specific knowledge, by exploiting the paired image-text
reports from the radiological daily practice. In particular, we make the
following contributions: First, unlike existing works that directly process the
raw reports, we adopt a novel triplet extraction module to extract the
medical-related information, avoiding unnecessary complexity from language
grammar and enhancing the supervision signals; Second, we propose a novel
triplet encoding module with entity translation by querying a knowledge base,
to exploit the rich domain knowledge in medical field, and implicitly build
relationships between medical entities in the language embedding space; Third,
we propose to use a Transformer-based fusion model for spatially aligning the
entity description with visual signals at the image patch level, enabling the
ability for medical diagnosis; Fourth, we conduct thorough experiments to
validate the effectiveness of our architecture, and benchmark on numerous
public benchmarks, e.g., ChestX-ray14, RSNA Pneumonia, SIIM-ACR Pneumothorax,
COVIDx CXR-2, COVID Rural, and EdemaSeverity. In both zero-shot and fine-tuning
settings, our model has demonstrated strong performance compared with the
former methods on disease classification and grounding.
| no_new_dataset | 0.947721 |
2301.02778 | Gongyang Li | Gongyang Li, Zhi Liu, Xinpeng Zhang, Weisi Lin | Lightweight Salient Object Detection in Optical Remote-Sensing Images
via Semantic Matching and Edge Alignment | 11 pages, 4 figures, Accepted by IEEE Transactions on Geoscience and
Remote Sensing 2023 | null | 10.1109/TGRS.2023.3235717 | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Recently, relying on convolutional neural networks (CNNs), many methods for
salient object detection in optical remote sensing images (ORSI-SOD) are
proposed. However, most methods ignore the huge parameters and computational
cost brought by CNNs, and only a few pay attention to the portability and
mobility. To facilitate practical applications, in this paper, we propose a
novel lightweight network for ORSI-SOD based on semantic matching and edge
alignment, termed SeaNet. Specifically, SeaNet includes a lightweight
MobileNet-V2 for feature extraction, a dynamic semantic matching module (DSMM)
for high-level features, an edge self-alignment module (ESAM) for low-level
features, and a portable decoder for inference. First, the high-level features
are compressed into semantic kernels. Then, semantic kernels are used to
activate salient object locations in two groups of high-level features through
dynamic convolution operations in DSMM. Meanwhile, in ESAM, cross-scale edge
information extracted from two groups of low-level features is self-aligned
through L2 loss and used for detail enhancement. Finally, starting from the
highest-level features, the decoder infers salient objects based on the
accurate locations and fine details contained in the outputs of the two
modules. Extensive experiments on two public datasets demonstrate that our
lightweight SeaNet not only outperforms most state-of-the-art lightweight
methods but also yields comparable accuracy with state-of-the-art conventional
methods, while having only 2.76M parameters and running with 1.7G FLOPs for
288x288 inputs. Our code and results are available at
https://github.com/MathLee/SeaNet.
| [
{
"version": "v1",
"created": "Sat, 7 Jan 2023 04:33:51 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 05:02:47 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Li",
"Gongyang",
""
],
[
"Liu",
"Zhi",
""
],
[
"Zhang",
"Xinpeng",
""
],
[
"Lin",
"Weisi",
""
]
] | TITLE: Lightweight Salient Object Detection in Optical Remote-Sensing Images
via Semantic Matching and Edge Alignment
ABSTRACT: Recently, relying on convolutional neural networks (CNNs), many methods for
salient object detection in optical remote sensing images (ORSI-SOD) are
proposed. However, most methods ignore the huge parameters and computational
cost brought by CNNs, and only a few pay attention to the portability and
mobility. To facilitate practical applications, in this paper, we propose a
novel lightweight network for ORSI-SOD based on semantic matching and edge
alignment, termed SeaNet. Specifically, SeaNet includes a lightweight
MobileNet-V2 for feature extraction, a dynamic semantic matching module (DSMM)
for high-level features, an edge self-alignment module (ESAM) for low-level
features, and a portable decoder for inference. First, the high-level features
are compressed into semantic kernels. Then, semantic kernels are used to
activate salient object locations in two groups of high-level features through
dynamic convolution operations in DSMM. Meanwhile, in ESAM, cross-scale edge
information extracted from two groups of low-level features is self-aligned
through L2 loss and used for detail enhancement. Finally, starting from the
highest-level features, the decoder infers salient objects based on the
accurate locations and fine details contained in the outputs of the two
modules. Extensive experiments on two public datasets demonstrate that our
lightweight SeaNet not only outperforms most state-of-the-art lightweight
methods but also yields comparable accuracy with state-of-the-art conventional
methods, while having only 2.76M parameters and running with 1.7G FLOPs for
288x288 inputs. Our code and results are available at
https://github.com/MathLee/SeaNet.
| no_new_dataset | 0.950041 |
2301.05169 | Yuejiang Liu | Yuejiang Liu, Alexandre Alahi, Chris Russell, Max Horn, Dominik
Zietlow, Bernhard Sch\"olkopf, Francesco Locatello | Causal Triplet: An Open Challenge for Intervention-centric Causal
Representation Learning | Conference on Causal Learning and Reasoning (CLeaR) 2023 | null | null | null | cs.LG cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent years have seen a surge of interest in learning high-level causal
representations from low-level image pairs under interventions. Yet, existing
efforts are largely limited to simple synthetic settings that are far away from
real-world problems. In this paper, we present Causal Triplet, a causal
representation learning benchmark featuring not only visually more complex
scenes, but also two crucial desiderata commonly overlooked in previous works:
(i) an actionable counterfactual setting, where only certain object-level
variables allow for counterfactual observations whereas others do not; (ii) an
interventional downstream task with an emphasis on out-of-distribution
robustness from the independent causal mechanisms principle. Through extensive
experiments, we find that models built with the knowledge of disentangled or
object-centric representations significantly outperform their distributed
counterparts. However, recent causal representation learning methods still
struggle to identify such latent structures, indicating substantial challenges
and opportunities for future work. Our code and datasets will be available at
https://sites.google.com/view/causaltriplet.
| [
{
"version": "v1",
"created": "Thu, 12 Jan 2023 17:43:38 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 17:19:51 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liu",
"Yuejiang",
""
],
[
"Alahi",
"Alexandre",
""
],
[
"Russell",
"Chris",
""
],
[
"Horn",
"Max",
""
],
[
"Zietlow",
"Dominik",
""
],
[
"Schölkopf",
"Bernhard",
""
],
[
"Locatello",
"Francesco",
""
]
] | TITLE: Causal Triplet: An Open Challenge for Intervention-centric Causal
Representation Learning
ABSTRACT: Recent years have seen a surge of interest in learning high-level causal
representations from low-level image pairs under interventions. Yet, existing
efforts are largely limited to simple synthetic settings that are far away from
real-world problems. In this paper, we present Causal Triplet, a causal
representation learning benchmark featuring not only visually more complex
scenes, but also two crucial desiderata commonly overlooked in previous works:
(i) an actionable counterfactual setting, where only certain object-level
variables allow for counterfactual observations whereas others do not; (ii) an
interventional downstream task with an emphasis on out-of-distribution
robustness from the independent causal mechanisms principle. Through extensive
experiments, we find that models built with the knowledge of disentangled or
object-centric representations significantly outperform their distributed
counterparts. However, recent causal representation learning methods still
struggle to identify such latent structures, indicating substantial challenges
and opportunities for future work. Our code and datasets will be available at
https://sites.google.com/view/causaltriplet.
| no_new_dataset | 0.901488 |
2301.09376 | Jing Huang | Hao Wen, Jing Huang, Huili Cui, Haozhe Lin, YuKun Lai, Lu Fang and Kun
Li | Crowd3D: Towards Hundreds of People Reconstruction from a Single Image | Accepted by CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Image-based multi-person reconstruction in wide-field large scenes is
critical for crowd analysis and security alert. However, existing methods
cannot deal with large scenes containing hundreds of people, which encounter
the challenges of large number of people, large variations in human scale, and
complex spatial distribution. In this paper, we propose Crowd3D, the first
framework to reconstruct the 3D poses, shapes and locations of hundreds of
people with global consistency from a single large-scene image. The core of our
approach is to convert the problem of complex crowd localization into pixel
localization with the help of our newly defined concept, Human-scene Virtual
Interaction Point (HVIP). To reconstruct the crowd with global consistency, we
propose a progressive reconstruction network based on HVIP by pre-estimating a
scene-level camera and a ground plane. To deal with a large number of persons
and various human sizes, we also design an adaptive human-centric cropping
scheme. Besides, we contribute a benchmark dataset, LargeCrowd, for crowd
reconstruction in a large scene. Experimental results demonstrate the
effectiveness of the proposed method. The code and datasets will be made
public.
| [
{
"version": "v1",
"created": "Mon, 23 Jan 2023 11:45:27 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 14:17:26 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wen",
"Hao",
""
],
[
"Huang",
"Jing",
""
],
[
"Cui",
"Huili",
""
],
[
"Lin",
"Haozhe",
""
],
[
"Lai",
"YuKun",
""
],
[
"Fang",
"Lu",
""
],
[
"Li",
"Kun",
""
]
] | TITLE: Crowd3D: Towards Hundreds of People Reconstruction from a Single Image
ABSTRACT: Image-based multi-person reconstruction in wide-field large scenes is
critical for crowd analysis and security alert. However, existing methods
cannot deal with large scenes containing hundreds of people, which encounter
the challenges of large number of people, large variations in human scale, and
complex spatial distribution. In this paper, we propose Crowd3D, the first
framework to reconstruct the 3D poses, shapes and locations of hundreds of
people with global consistency from a single large-scene image. The core of our
approach is to convert the problem of complex crowd localization into pixel
localization with the help of our newly defined concept, Human-scene Virtual
Interaction Point (HVIP). To reconstruct the crowd with global consistency, we
propose a progressive reconstruction network based on HVIP by pre-estimating a
scene-level camera and a ground plane. To deal with a large number of persons
and various human sizes, we also design an adaptive human-centric cropping
scheme. Besides, we contribute a benchmark dataset, LargeCrowd, for crowd
reconstruction in a large scene. Experimental results demonstrate the
effectiveness of the proposed method. The code and datasets will be made
public.
| new_dataset | 0.956104 |
2302.03802 | Ziqi Pang | Ziqi Pang, Jie Li, Pavel Tokmakov, Dian Chen, Sergey Zagoruyko,
Yu-Xiong Wang | Standing Between Past and Future: Spatio-Temporal Modeling for
Multi-Camera 3D Multi-Object Tracking | CVPR 2023 Camera Ready, 15 pages, 8 figures | null | null | null | cs.CV cs.AI cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work proposes an end-to-end multi-camera 3D multi-object tracking (MOT)
framework. It emphasizes spatio-temporal continuity and integrates both past
and future reasoning for tracked objects. Thus, we name it "Past-and-Future
reasoning for Tracking" (PF-Track). Specifically, our method adapts the
"tracking by attention" framework and represents tracked instances coherently
over time with object queries. To explicitly use historical cues, our "Past
Reasoning" module learns to refine the tracks and enhance the object features
by cross-attending to queries from previous frames and other objects. The
"Future Reasoning" module digests historical information and predicts robust
future trajectories. In the case of long-term occlusions, our method maintains
the object positions and enables re-association by integrating motion
predictions. On the nuScenes dataset, our method improves AMOTA by a large
margin and remarkably reduces ID-Switches by 90% compared to prior approaches,
which is an order of magnitude less. The code and models are made available at
https://github.com/TRI-ML/PF-Track.
| [
{
"version": "v1",
"created": "Tue, 7 Feb 2023 23:46:34 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 14:49:49 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Pang",
"Ziqi",
""
],
[
"Li",
"Jie",
""
],
[
"Tokmakov",
"Pavel",
""
],
[
"Chen",
"Dian",
""
],
[
"Zagoruyko",
"Sergey",
""
],
[
"Wang",
"Yu-Xiong",
""
]
] | TITLE: Standing Between Past and Future: Spatio-Temporal Modeling for
Multi-Camera 3D Multi-Object Tracking
ABSTRACT: This work proposes an end-to-end multi-camera 3D multi-object tracking (MOT)
framework. It emphasizes spatio-temporal continuity and integrates both past
and future reasoning for tracked objects. Thus, we name it "Past-and-Future
reasoning for Tracking" (PF-Track). Specifically, our method adapts the
"tracking by attention" framework and represents tracked instances coherently
over time with object queries. To explicitly use historical cues, our "Past
Reasoning" module learns to refine the tracks and enhance the object features
by cross-attending to queries from previous frames and other objects. The
"Future Reasoning" module digests historical information and predicts robust
future trajectories. In the case of long-term occlusions, our method maintains
the object positions and enables re-association by integrating motion
predictions. On the nuScenes dataset, our method improves AMOTA by a large
margin and remarkably reduces ID-Switches by 90% compared to prior approaches,
which is an order of magnitude less. The code and models are made available at
https://github.com/TRI-ML/PF-Track.
| no_new_dataset | 0.950824 |
2302.10296 | Xiangyu Wen | Xiangyu Wen, Yu Li, Wei Jiang, Qiang Xu | On Function-Coupled Watermarks for Deep Neural Networks | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Well-performed deep neural networks (DNNs) generally require massive labelled
data and computational resources for training. Various watermarking techniques
are proposed to protect such intellectual properties (IPs), wherein the DNN
providers implant secret information into the model so that they can later
claim IP ownership by retrieving their embedded watermarks with some dedicated
trigger inputs. While promising results are reported in the literature,
existing solutions suffer from watermark removal attacks, such as model
fine-tuning and model pruning.
In this paper, we propose a novel DNN watermarking solution that can
effectively defend against the above attacks. Our key insight is to enhance the
coupling of the watermark and model functionalities such that removing the
watermark would inevitably degrade the model's performance on normal inputs. To
this end, unlike previous methods relying on secret features learnt from
out-of-distribution data, our method only uses features learnt from
in-distribution data. Specifically, on the one hand, we propose to sample
inputs from the original training dataset and fuse them as watermark triggers.
On the other hand, we randomly mask model weights during training so that the
information of our embedded watermarks spreads in the network. By doing so,
model fine-tuning/pruning would not forget our function-coupled watermarks.
Evaluation results on various image classification tasks show a 100\% watermark
authentication success rate under aggressive watermark removal attacks,
significantly outperforming existing solutions. Code is available:
https://github.com/cure-lab/Function-Coupled-Watermark.
| [
{
"version": "v1",
"created": "Wed, 8 Feb 2023 05:55:16 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Mar 2023 17:30:34 GMT"
},
{
"version": "v3",
"created": "Sat, 1 Apr 2023 16:27:00 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wen",
"Xiangyu",
""
],
[
"Li",
"Yu",
""
],
[
"Jiang",
"Wei",
""
],
[
"Xu",
"Qiang",
""
]
] | TITLE: On Function-Coupled Watermarks for Deep Neural Networks
ABSTRACT: Well-performed deep neural networks (DNNs) generally require massive labelled
data and computational resources for training. Various watermarking techniques
are proposed to protect such intellectual properties (IPs), wherein the DNN
providers implant secret information into the model so that they can later
claim IP ownership by retrieving their embedded watermarks with some dedicated
trigger inputs. While promising results are reported in the literature,
existing solutions suffer from watermark removal attacks, such as model
fine-tuning and model pruning.
In this paper, we propose a novel DNN watermarking solution that can
effectively defend against the above attacks. Our key insight is to enhance the
coupling of the watermark and model functionalities such that removing the
watermark would inevitably degrade the model's performance on normal inputs. To
this end, unlike previous methods relying on secret features learnt from
out-of-distribution data, our method only uses features learnt from
in-distribution data. Specifically, on the one hand, we propose to sample
inputs from the original training dataset and fuse them as watermark triggers.
On the other hand, we randomly mask model weights during training so that the
information of our embedded watermarks spreads in the network. By doing so,
model fine-tuning/pruning would not forget our function-coupled watermarks.
Evaluation results on various image classification tasks show a 100\% watermark
authentication success rate under aggressive watermark removal attacks,
significantly outperforming existing solutions. Code is available:
https://github.com/cure-lab/Function-Coupled-Watermark.
| no_new_dataset | 0.944587 |
2303.00448 | Wenrui Li | Wenrui Li, Zhengyu Ma, Jinqiao Shi, Xiaopeng Fan | The style transformer with common knowledge optimization for image-text
retrieval | null | null | null | null | cs.CV cs.MM | http://creativecommons.org/licenses/by/4.0/ | Image-text retrieval which associates different modalities has drawn broad
attention due to its excellent research value and broad real-world application.
However, most of the existing methods haven't taken the high-level semantic
relationships ("style embedding") and common knowledge from multi-modalities
into full consideration. To this end, we introduce a novel style transformer
network with common knowledge optimization (CKSTN) for image-text retrieval.
The main module is the common knowledge adaptor (CKA) with both the style
embedding extractor (SEE) and the common knowledge optimization (CKO) modules.
Specifically, the SEE uses the sequential update strategy to effectively
connect the features of different stages in SEE. The CKO module is introduced
to dynamically capture the latent concepts of common knowledge from different
modalities. Besides, to get generalized temporal common knowledge, we propose a
sequential update strategy to effectively integrate the features of different
layers in SEE with previous common feature units. CKSTN demonstrates the
superiorities of the state-of-the-art methods in image-text retrieval on MSCOCO
and Flickr30K datasets. Moreover, CKSTN is constructed based on the lightweight
transformer which is more convenient and practical for the application of real
scenes, due to the better performance and lower parameters.
| [
{
"version": "v1",
"created": "Wed, 1 Mar 2023 12:17:33 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 11:17:11 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Li",
"Wenrui",
""
],
[
"Ma",
"Zhengyu",
""
],
[
"Shi",
"Jinqiao",
""
],
[
"Fan",
"Xiaopeng",
""
]
] | TITLE: The style transformer with common knowledge optimization for image-text
retrieval
ABSTRACT: Image-text retrieval which associates different modalities has drawn broad
attention due to its excellent research value and broad real-world application.
However, most of the existing methods haven't taken the high-level semantic
relationships ("style embedding") and common knowledge from multi-modalities
into full consideration. To this end, we introduce a novel style transformer
network with common knowledge optimization (CKSTN) for image-text retrieval.
The main module is the common knowledge adaptor (CKA) with both the style
embedding extractor (SEE) and the common knowledge optimization (CKO) modules.
Specifically, the SEE uses the sequential update strategy to effectively
connect the features of different stages in SEE. The CKO module is introduced
to dynamically capture the latent concepts of common knowledge from different
modalities. Besides, to get generalized temporal common knowledge, we propose a
sequential update strategy to effectively integrate the features of different
layers in SEE with previous common feature units. CKSTN demonstrates the
superiorities of the state-of-the-art methods in image-text retrieval on MSCOCO
and Flickr30K datasets. Moreover, CKSTN is constructed based on the lightweight
transformer which is more convenient and practical for the application of real
scenes, due to the better performance and lower parameters.
| no_new_dataset | 0.944177 |
2303.01991 | Kurt Stolle | Kurt Stolle and Gijs Dubbelman | Unified Perception: Efficient Depth-Aware Video Panoptic Segmentation
with Minimal Annotation Costs | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Depth-aware video panoptic segmentation is a promising approach to camera
based scene understanding. However, the current state-of-the-art methods
require costly video annotations and use a complex training pipeline compared
to their image-based equivalents. In this paper, we present a new approach
titled Unified Perception that achieves state-of-the-art performance without
requiring video-based training. Our method employs a simple two-stage cascaded
tracking algorithm that (re)uses object embeddings computed in an image-based
network. Experimental results on the Cityscapes-DVPS dataset demonstrate that
our method achieves an overall DVPQ of 57.1, surpassing state-of-the-art
methods. Furthermore, we show that our tracking strategies are effective for
long-term object association on KITTI-STEP, achieving an STQ of 59.1 which
exceeded the performance of state-of-the-art methods that employ the same
backbone network.
Code is available at: https://tue-mps.github.io/unipercept
| [
{
"version": "v1",
"created": "Fri, 3 Mar 2023 15:00:12 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Apr 2023 17:25:52 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Stolle",
"Kurt",
""
],
[
"Dubbelman",
"Gijs",
""
]
] | TITLE: Unified Perception: Efficient Depth-Aware Video Panoptic Segmentation
with Minimal Annotation Costs
ABSTRACT: Depth-aware video panoptic segmentation is a promising approach to camera
based scene understanding. However, the current state-of-the-art methods
require costly video annotations and use a complex training pipeline compared
to their image-based equivalents. In this paper, we present a new approach
titled Unified Perception that achieves state-of-the-art performance without
requiring video-based training. Our method employs a simple two-stage cascaded
tracking algorithm that (re)uses object embeddings computed in an image-based
network. Experimental results on the Cityscapes-DVPS dataset demonstrate that
our method achieves an overall DVPQ of 57.1, surpassing state-of-the-art
methods. Furthermore, we show that our tracking strategies are effective for
long-term object association on KITTI-STEP, achieving an STQ of 59.1 which
exceeded the performance of state-of-the-art methods that employ the same
backbone network.
Code is available at: https://tue-mps.github.io/unipercept
| no_new_dataset | 0.95253 |
2303.02454 | Yun Wang | Yun Wang, Cheng Chi, Xin Yang | Exploiting Implicit Rigidity Constraints via Weight-Sharing Aggregation
for Scene Flow Estimation from Point Clouds | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Scene flow estimation, which predicts the 3D motion of scene points from
point clouds, is a core task in autonomous driving and many other 3D vision
applications. Existing methods either suffer from structure distortion due to
ignorance of rigid motion consistency or require explicit pose estimation and
3D object segmentation. Errors of estimated poses and segmented objects would
yield inaccurate rigidity constraints and in turn mislead scene flow
estimation. In this paper, we propose a novel weight-sharing aggregation (WSA)
method for feature and scene flow up-sampling. WSA does not rely on estimated
poses and segmented objects, and can implicitly enforce rigidity constraints to
avoid structure distortion in scene flow estimation. To further exploit
geometric information and preserve local structure, we design a deformation
degree module aim to keep the local region invariance. We modify the
PointPWC-Net and integrate the proposed WSA and deformation degree module into
the enhanced PointPWC-Net to derive an end-to-end scene flow estimation
network, called WSAFlowNet. Extensive experimental results on the
FlyingThings3D and KITTI datasets demonstrate that our WSAFlowNet achieves the
state-of-the-art performance and outperforms previous methods by a large
margin. We will release the source code at
https://github.com/wangyunlhr/WSAFlowNet.git.
| [
{
"version": "v1",
"created": "Sat, 4 Mar 2023 16:55:57 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 09:01:55 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wang",
"Yun",
""
],
[
"Chi",
"Cheng",
""
],
[
"Yang",
"Xin",
""
]
] | TITLE: Exploiting Implicit Rigidity Constraints via Weight-Sharing Aggregation
for Scene Flow Estimation from Point Clouds
ABSTRACT: Scene flow estimation, which predicts the 3D motion of scene points from
point clouds, is a core task in autonomous driving and many other 3D vision
applications. Existing methods either suffer from structure distortion due to
ignorance of rigid motion consistency or require explicit pose estimation and
3D object segmentation. Errors of estimated poses and segmented objects would
yield inaccurate rigidity constraints and in turn mislead scene flow
estimation. In this paper, we propose a novel weight-sharing aggregation (WSA)
method for feature and scene flow up-sampling. WSA does not rely on estimated
poses and segmented objects, and can implicitly enforce rigidity constraints to
avoid structure distortion in scene flow estimation. To further exploit
geometric information and preserve local structure, we design a deformation
degree module aim to keep the local region invariance. We modify the
PointPWC-Net and integrate the proposed WSA and deformation degree module into
the enhanced PointPWC-Net to derive an end-to-end scene flow estimation
network, called WSAFlowNet. Extensive experimental results on the
FlyingThings3D and KITTI datasets demonstrate that our WSAFlowNet achieves the
state-of-the-art performance and outperforms previous methods by a large
margin. We will release the source code at
https://github.com/wangyunlhr/WSAFlowNet.git.
| no_new_dataset | 0.947624 |
2303.02508 | Zhenning Yang | Zhenning Yang, Luoxi Meng, Jae-Won Chung, Mosharaf Chowdhury | Chasing Low-Carbon Electricity for Practical and Sustainable DNN
Training | ICLR 23 Workshop | https://ml.energy/zeus | null | null | null | cs.LG cs.AI cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Deep learning has experienced significant growth in recent years, resulting
in increased energy consumption and carbon emission from the use of GPUs for
training deep neural networks (DNNs). Answering the call for sustainability,
conventional solutions have attempted to move training jobs to locations or
time frames with lower carbon intensity. However, moving jobs to other
locations may not always be feasible due to large dataset sizes or data
regulations. Moreover, postponing training can negatively impact application
service quality because the DNNs backing the service are not updated in a
timely fashion. In this work, we present a practical solution that reduces the
carbon footprint of DNN training without migrating or postponing jobs.
Specifically, our solution observes real-time carbon intensity shifts during
training and controls the energy consumption of GPUs, thereby reducing carbon
footprint while maintaining training performance. Furthermore, in order to
proactively adapt to shifting carbon intensity, we propose a lightweight
machine learning algorithm that predicts the carbon intensity of the upcoming
time frame. Our solution, Chase, reduces the total carbon footprint of training
ResNet-50 on ImageNet by 13.6% while only increasing training time by 2.5%.
| [
{
"version": "v1",
"created": "Sat, 4 Mar 2023 21:33:29 GMT"
},
{
"version": "v2",
"created": "Sun, 2 Apr 2023 02:11:08 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yang",
"Zhenning",
""
],
[
"Meng",
"Luoxi",
""
],
[
"Chung",
"Jae-Won",
""
],
[
"Chowdhury",
"Mosharaf",
""
]
] | TITLE: Chasing Low-Carbon Electricity for Practical and Sustainable DNN
Training
ABSTRACT: Deep learning has experienced significant growth in recent years, resulting
in increased energy consumption and carbon emission from the use of GPUs for
training deep neural networks (DNNs). Answering the call for sustainability,
conventional solutions have attempted to move training jobs to locations or
time frames with lower carbon intensity. However, moving jobs to other
locations may not always be feasible due to large dataset sizes or data
regulations. Moreover, postponing training can negatively impact application
service quality because the DNNs backing the service are not updated in a
timely fashion. In this work, we present a practical solution that reduces the
carbon footprint of DNN training without migrating or postponing jobs.
Specifically, our solution observes real-time carbon intensity shifts during
training and controls the energy consumption of GPUs, thereby reducing carbon
footprint while maintaining training performance. Furthermore, in order to
proactively adapt to shifting carbon intensity, we propose a lightweight
machine learning algorithm that predicts the carbon intensity of the upcoming
time frame. Our solution, Chase, reduces the total carbon footprint of training
ResNet-50 on ImageNet by 13.6% while only increasing training time by 2.5%.
| no_new_dataset | 0.952175 |
2303.03709 | Daoan Zhang | Shuai Wang, Daoan Zhang, Jianguo Zhang, Weiwei Zhang, and Rui Li | Bootstrap The Original Latent: Learning a Private Model from a Black-box
Model | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | In this paper, considering the balance of data/model privacy of model owners
and user needs, we propose a new setting called Back-Propagated Black-Box
Adaptation (BPBA) for users to better train their private models via the
guidance of the back-propagated results of a Black-box foundation/source model.
Our setting can ease the usage of foundation/source models as well as prevent
the leakage and misuse of foundation/source models. Moreover, we also propose a
new training strategy called Bootstrap The Original Latent (BTOL) to fully
utilize the foundation/source models. Our strategy consists of a domain adapter
and a freeze-and-thaw strategy. We apply our BTOL under BPBA and Black-box UDA
settings on three different datasets. Experiments show that our strategy is
efficient and robust in various settings without manual augmentations.
| [
{
"version": "v1",
"created": "Tue, 7 Mar 2023 07:47:41 GMT"
},
{
"version": "v2",
"created": "Thu, 9 Mar 2023 16:23:02 GMT"
},
{
"version": "v3",
"created": "Sat, 18 Mar 2023 18:13:28 GMT"
},
{
"version": "v4",
"created": "Mon, 3 Apr 2023 03:25:32 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wang",
"Shuai",
""
],
[
"Zhang",
"Daoan",
""
],
[
"Zhang",
"Jianguo",
""
],
[
"Zhang",
"Weiwei",
""
],
[
"Li",
"Rui",
""
]
] | TITLE: Bootstrap The Original Latent: Learning a Private Model from a Black-box
Model
ABSTRACT: In this paper, considering the balance of data/model privacy of model owners
and user needs, we propose a new setting called Back-Propagated Black-Box
Adaptation (BPBA) for users to better train their private models via the
guidance of the back-propagated results of a Black-box foundation/source model.
Our setting can ease the usage of foundation/source models as well as prevent
the leakage and misuse of foundation/source models. Moreover, we also propose a
new training strategy called Bootstrap The Original Latent (BTOL) to fully
utilize the foundation/source models. Our strategy consists of a domain adapter
and a freeze-and-thaw strategy. We apply our BTOL under BPBA and Black-box UDA
settings on three different datasets. Experiments show that our strategy is
efficient and robust in various settings without manual augmentations.
| no_new_dataset | 0.943034 |
2303.04743 | Daesoo Lee | Daesoo Lee, Sara Malacarne and Erlend Aune | Vector Quantized Time Series Generation with a Bidirectional Prior Model | accepted at AISTATS 2023 | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Time series generation (TSG) studies have mainly focused on the use of
Generative Adversarial Networks (GANs) combined with recurrent neural network
(RNN) variants. However, the fundamental limitations and challenges of training
GANs still remain. In addition, the RNN-family typically has difficulties with
temporal consistency between distant timesteps. Motivated by the successes in
the image generation (IMG) domain, we propose TimeVQVAE, the first work, to our
knowledge, that uses vector quantization (VQ) techniques to address the TSG
problem. Moreover, the priors of the discrete latent spaces are learned with
bidirectional transformer models that can better capture global temporal
consistency. We also propose VQ modeling in a time-frequency domain, separated
into low-frequency (LF) and high-frequency (HF). This allows us to retain
important characteristics of the time series and, in turn, generate new
synthetic signals that are of better quality, with sharper changes in
modularity, than its competing TSG methods. Our experimental evaluation is
conducted on all datasets from the UCR archive, using well-established metrics
in the IMG literature, such as Fr\'echet inception distance and inception
scores. Our implementation on GitHub:
\url{https://github.com/ML4ITS/TimeVQVAE}.
| [
{
"version": "v1",
"created": "Wed, 8 Mar 2023 17:27:39 GMT"
},
{
"version": "v2",
"created": "Wed, 15 Mar 2023 11:50:41 GMT"
},
{
"version": "v3",
"created": "Sun, 2 Apr 2023 02:25:06 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Lee",
"Daesoo",
""
],
[
"Malacarne",
"Sara",
""
],
[
"Aune",
"Erlend",
""
]
] | TITLE: Vector Quantized Time Series Generation with a Bidirectional Prior Model
ABSTRACT: Time series generation (TSG) studies have mainly focused on the use of
Generative Adversarial Networks (GANs) combined with recurrent neural network
(RNN) variants. However, the fundamental limitations and challenges of training
GANs still remain. In addition, the RNN-family typically has difficulties with
temporal consistency between distant timesteps. Motivated by the successes in
the image generation (IMG) domain, we propose TimeVQVAE, the first work, to our
knowledge, that uses vector quantization (VQ) techniques to address the TSG
problem. Moreover, the priors of the discrete latent spaces are learned with
bidirectional transformer models that can better capture global temporal
consistency. We also propose VQ modeling in a time-frequency domain, separated
into low-frequency (LF) and high-frequency (HF). This allows us to retain
important characteristics of the time series and, in turn, generate new
synthetic signals that are of better quality, with sharper changes in
modularity, than its competing TSG methods. Our experimental evaluation is
conducted on all datasets from the UCR archive, using well-established metrics
in the IMG literature, such as Fr\'echet inception distance and inception
scores. Our implementation on GitHub:
\url{https://github.com/ML4ITS/TimeVQVAE}.
| no_new_dataset | 0.946941 |
2303.05062 | Chattu Bhargavi | Chattu Bhargavi and Vikash Kumar Singh | Influence before Hiring: A Two-tired Incentive Compatible Mechanism for
IoT-based Crowdsourcing in Strategic Setting | 40 pages, 21 figures, Changed Figure 1, Simulation results are added,
Conclusion and future works stretched a bit | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In crowdsourcing, a group of common people is asked to execute the tasks and
in return will receive some incentives. In this article, one of the
crowdsourcing scenarios with multiple heterogeneous tasks and multiple IoT
devices (as task executors) is studied as a two-tiered process. In the first
tier of the proposed model, it is assumed that a substantial number of IoT
devices are not aware of the hiring process and are made aware by utilizing
their social connections. Each of the IoT devices reports a cost (private
value) that it will charge in return for its services. The participating IoT
devices are rational and strategic. The goal of the first tier is to select the
subset of IoT devices as initial notifiers so as to maximize the number of IoT
devices notified with the constraint that the total payment made to the
notifiers is within the budget. For this purpose, an incentive compatible
mechanism is proposed. In the second tier, a set of quality IoT devices is
determined by utilizing the idea of single-peaked preferences. The next
objective of the second tier is to hire quality IoT devices for the floated
tasks. For this purpose, each quality IoT device reports private valuation
along with its favorite bundle of tasks. In the second tier, it is assumed that
the valuation of the IoT devices satisfies gross substitute criteria and is
private. For the second tier, the truthful mechanisms are designed
independently for determining the quality IoT devices and for hiring them and
deciding their payment respectively. Theoretical analysis shows that the
proposed mechanisms are computationally efficient, truthful, correct, budget
feasible, and individually rational. The simulation is done to measure the
efficacy of the proposed mechanisms with the benchmark mechanisms based on
truthfulness, budget feasibility, and running time.
| [
{
"version": "v1",
"created": "Thu, 9 Mar 2023 06:24:10 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 05:13:35 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Bhargavi",
"Chattu",
""
],
[
"Singh",
"Vikash Kumar",
""
]
] | TITLE: Influence before Hiring: A Two-tired Incentive Compatible Mechanism for
IoT-based Crowdsourcing in Strategic Setting
ABSTRACT: In crowdsourcing, a group of common people is asked to execute the tasks and
in return will receive some incentives. In this article, one of the
crowdsourcing scenarios with multiple heterogeneous tasks and multiple IoT
devices (as task executors) is studied as a two-tiered process. In the first
tier of the proposed model, it is assumed that a substantial number of IoT
devices are not aware of the hiring process and are made aware by utilizing
their social connections. Each of the IoT devices reports a cost (private
value) that it will charge in return for its services. The participating IoT
devices are rational and strategic. The goal of the first tier is to select the
subset of IoT devices as initial notifiers so as to maximize the number of IoT
devices notified with the constraint that the total payment made to the
notifiers is within the budget. For this purpose, an incentive compatible
mechanism is proposed. In the second tier, a set of quality IoT devices is
determined by utilizing the idea of single-peaked preferences. The next
objective of the second tier is to hire quality IoT devices for the floated
tasks. For this purpose, each quality IoT device reports private valuation
along with its favorite bundle of tasks. In the second tier, it is assumed that
the valuation of the IoT devices satisfies gross substitute criteria and is
private. For the second tier, the truthful mechanisms are designed
independently for determining the quality IoT devices and for hiring them and
deciding their payment respectively. Theoretical analysis shows that the
proposed mechanisms are computationally efficient, truthful, correct, budget
feasible, and individually rational. The simulation is done to measure the
efficacy of the proposed mechanisms with the benchmark mechanisms based on
truthfulness, budget feasibility, and running time.
| no_new_dataset | 0.947866 |
2303.08594 | Junjie He | Junjie He, Pengyu Li, Yifeng Geng, Xuansong Xie | FastInst: A Simple Query-Based Model for Real-Time Instance Segmentation | CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent attention in instance segmentation has focused on query-based models.
Despite being non-maximum suppression (NMS)-free and end-to-end, the
superiority of these models on high-accuracy real-time benchmarks has not been
well demonstrated. In this paper, we show the strong potential of query-based
models on efficient instance segmentation algorithm designs. We present
FastInst, a simple, effective query-based framework for real-time instance
segmentation. FastInst can execute at a real-time speed (i.e., 32.5 FPS) while
yielding an AP of more than 40 (i.e., 40.5 AP) on COCO test-dev without bells
and whistles. Specifically, FastInst follows the meta-architecture of recently
introduced Mask2Former. Its key designs include instance activation-guided
queries, dual-path update strategy, and ground truth mask-guided learning,
which enable us to use lighter pixel decoders, fewer Transformer decoder
layers, while achieving better performance. The experiments show that FastInst
outperforms most state-of-the-art real-time counterparts, including strong
fully convolutional baselines, in both speed and accuracy. Code can be found at
https://github.com/junjiehe96/FastInst .
| [
{
"version": "v1",
"created": "Wed, 15 Mar 2023 13:06:30 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 17:55:21 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"He",
"Junjie",
""
],
[
"Li",
"Pengyu",
""
],
[
"Geng",
"Yifeng",
""
],
[
"Xie",
"Xuansong",
""
]
] | TITLE: FastInst: A Simple Query-Based Model for Real-Time Instance Segmentation
ABSTRACT: Recent attention in instance segmentation has focused on query-based models.
Despite being non-maximum suppression (NMS)-free and end-to-end, the
superiority of these models on high-accuracy real-time benchmarks has not been
well demonstrated. In this paper, we show the strong potential of query-based
models on efficient instance segmentation algorithm designs. We present
FastInst, a simple, effective query-based framework for real-time instance
segmentation. FastInst can execute at a real-time speed (i.e., 32.5 FPS) while
yielding an AP of more than 40 (i.e., 40.5 AP) on COCO test-dev without bells
and whistles. Specifically, FastInst follows the meta-architecture of recently
introduced Mask2Former. Its key designs include instance activation-guided
queries, dual-path update strategy, and ground truth mask-guided learning,
which enable us to use lighter pixel decoders, fewer Transformer decoder
layers, while achieving better performance. The experiments show that FastInst
outperforms most state-of-the-art real-time counterparts, including strong
fully convolutional baselines, in both speed and accuracy. Code can be found at
https://github.com/junjiehe96/FastInst .
| no_new_dataset | 0.94743 |
2303.09184 | Gaochen Dong | Gaochen Dong, Wei Chen | Block-wise Bit-Compression of Transformer-based Models | Need to add figures and adjust languages to improve readability | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the popularity of the recent Transformer-based models represented by
BERT, GPT-3 and ChatGPT, there has been state-of-the-art performance in a range
of natural language processing tasks. However, the massive computations, huge
memory footprint, and thus high latency of Transformer-based models is an
inevitable challenge for the cloud with high real-time requirement. To tackle
the issue, we propose BBCT, a method of block-wise bit-compression for
transformer without retraining. Our method achieves more fine-grained
compression of the whole transformer, including embedding, matrix
multiplication, GELU, softmax, layer normalization, and all the intermediate
results. As a case, we compress an efficient BERT with the method of BBCT. Our
benchmark test results on General Language Understanding Evaluation (GLUE) show
that BBCT can achieve less than 1% accuracy drop in most tasks.
| [
{
"version": "v1",
"created": "Thu, 16 Mar 2023 09:53:57 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 12:50:29 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Dong",
"Gaochen",
""
],
[
"Chen",
"Wei",
""
]
] | TITLE: Block-wise Bit-Compression of Transformer-based Models
ABSTRACT: With the popularity of the recent Transformer-based models represented by
BERT, GPT-3 and ChatGPT, there has been state-of-the-art performance in a range
of natural language processing tasks. However, the massive computations, huge
memory footprint, and thus high latency of Transformer-based models is an
inevitable challenge for the cloud with high real-time requirement. To tackle
the issue, we propose BBCT, a method of block-wise bit-compression for
transformer without retraining. Our method achieves more fine-grained
compression of the whole transformer, including embedding, matrix
multiplication, GELU, softmax, layer normalization, and all the intermediate
results. As a case, we compress an efficient BERT with the method of BBCT. Our
benchmark test results on General Language Understanding Evaluation (GLUE) show
that BBCT can achieve less than 1% accuracy drop in most tasks.
| no_new_dataset | 0.940408 |
2303.11240 | Patrick Gerard | Patrick Gerard, Nicholas Botzer, Tim Weninger | Truth Social Dataset | 7 pages, 5 figures, ICWSM 2023 | null | null | null | cs.SI | http://creativecommons.org/licenses/by/4.0/ | Formally announced to the public following former President Donald Trump's
bans and suspensions from mainstream social networks in early 2022 after his
role in the January 6 Capitol Riots, Truth Social was launched as an
"alternative" social media platform that claims to be a refuge for free speech,
offering a platform for those disaffected by the content moderation policies of
the existing, mainstream social networks. The subsequent rise of Truth Social
has been driven largely by hard-line supporters of the former president as well
as those affected by the content moderation of other social networks. These
distinct qualities combined with its status as the main mouthpiece of the
former president positions Truth Social as a particularly influential social
media platform and give rise to several research questions. However, outside of
a handful of news reports, little is known about the new social media platform
partially due to a lack of well-curated data. In the current work, we describe
a dataset of over 823,000 posts to Truth Social and and social network with
over 454,000 distinct users. In addition to the dataset itself, we also present
some basic analysis of its content, certain temporal features, and its network.
| [
{
"version": "v1",
"created": "Mon, 20 Mar 2023 16:26:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Gerard",
"Patrick",
""
],
[
"Botzer",
"Nicholas",
""
],
[
"Weninger",
"Tim",
""
]
] | TITLE: Truth Social Dataset
ABSTRACT: Formally announced to the public following former President Donald Trump's
bans and suspensions from mainstream social networks in early 2022 after his
role in the January 6 Capitol Riots, Truth Social was launched as an
"alternative" social media platform that claims to be a refuge for free speech,
offering a platform for those disaffected by the content moderation policies of
the existing, mainstream social networks. The subsequent rise of Truth Social
has been driven largely by hard-line supporters of the former president as well
as those affected by the content moderation of other social networks. These
distinct qualities combined with its status as the main mouthpiece of the
former president positions Truth Social as a particularly influential social
media platform and give rise to several research questions. However, outside of
a handful of news reports, little is known about the new social media platform
partially due to a lack of well-curated data. In the current work, we describe
a dataset of over 823,000 posts to Truth Social and and social network with
over 454,000 distinct users. In addition to the dataset itself, we also present
some basic analysis of its content, certain temporal features, and its network.
| new_dataset | 0.958382 |
2303.12848 | Yun-Yun Tsai | Yun-Yun Tsai, Ju-Chin Chao, Albert Wen, Zhaoyuan Yang, Chengzhi Mao,
Tapan Shah, Junfeng Yang | Test-time Detection and Repair of Adversarial Samples via Masked
Autoencoder | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Training-time defenses, known as adversarial training, incur high training
costs and do not generalize to unseen attacks. Test-time defenses solve these
issues but most existing test-time defenses require adapting the model weights,
therefore they do not work on frozen models and complicate model memory
management. The only test-time defense that does not adapt model weights aims
to adapt the input with self-supervision tasks. However, we empirically found
these self-supervision tasks are not sensitive enough to detect adversarial
attacks accurately. In this paper, we propose DRAM, a novel defense method to
detect and repair adversarial samples at test time via Masked autoencoder
(MAE). We demonstrate how to use MAE losses to build a Kolmogorov-Smirnov test
to detect adversarial samples. Moreover, we use the MAE losses to calculate
input reversal vectors that repair adversarial samples resulting from
previously unseen attacks. Results on large-scale ImageNet dataset show that,
compared to all detection baselines evaluated, DRAM achieves the best detection
rate (82% on average) on all eight adversarial attacks evaluated. For attack
repair, DRAM improves the robust accuracy by 6% ~ 41% for standard ResNet50 and
3% ~ 8% for robust ResNet50 compared with the baselines that use contrastive
learning and rotation prediction.
| [
{
"version": "v1",
"created": "Wed, 22 Mar 2023 18:14:02 GMT"
},
{
"version": "v2",
"created": "Sat, 25 Mar 2023 03:53:43 GMT"
},
{
"version": "v3",
"created": "Sun, 2 Apr 2023 21:27:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Tsai",
"Yun-Yun",
""
],
[
"Chao",
"Ju-Chin",
""
],
[
"Wen",
"Albert",
""
],
[
"Yang",
"Zhaoyuan",
""
],
[
"Mao",
"Chengzhi",
""
],
[
"Shah",
"Tapan",
""
],
[
"Yang",
"Junfeng",
""
]
] | TITLE: Test-time Detection and Repair of Adversarial Samples via Masked
Autoencoder
ABSTRACT: Training-time defenses, known as adversarial training, incur high training
costs and do not generalize to unseen attacks. Test-time defenses solve these
issues but most existing test-time defenses require adapting the model weights,
therefore they do not work on frozen models and complicate model memory
management. The only test-time defense that does not adapt model weights aims
to adapt the input with self-supervision tasks. However, we empirically found
these self-supervision tasks are not sensitive enough to detect adversarial
attacks accurately. In this paper, we propose DRAM, a novel defense method to
detect and repair adversarial samples at test time via Masked autoencoder
(MAE). We demonstrate how to use MAE losses to build a Kolmogorov-Smirnov test
to detect adversarial samples. Moreover, we use the MAE losses to calculate
input reversal vectors that repair adversarial samples resulting from
previously unseen attacks. Results on large-scale ImageNet dataset show that,
compared to all detection baselines evaluated, DRAM achieves the best detection
rate (82% on average) on all eight adversarial attacks evaluated. For attack
repair, DRAM improves the robust accuracy by 6% ~ 41% for standard ResNet50 and
3% ~ 8% for robust ResNet50 compared with the baselines that use contrastive
learning and rotation prediction.
| no_new_dataset | 0.949809 |
2303.14186 | Andrew Ilyas | Sung Min Park, Kristian Georgiev, Andrew Ilyas, Guillaume Leclerc,
Aleksander Madry | TRAK: Attributing Model Behavior at Scale | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The goal of data attribution is to trace model predictions back to training
data. Despite a long line of work towards this goal, existing approaches to
data attribution tend to force users to choose between computational
tractability and efficacy. That is, computationally tractable methods can
struggle with accurately attributing model predictions in non-convex settings
(e.g., in the context of deep neural networks), while methods that are
effective in such regimes require training thousands of models, which makes
them impractical for large models or datasets.
In this work, we introduce TRAK (Tracing with the Randomly-projected After
Kernel), a data attribution method that is both effective and computationally
tractable for large-scale, differentiable models. In particular, by leveraging
only a handful of trained models, TRAK can match the performance of attribution
methods that require training thousands of models. We demonstrate the utility
of TRAK across various modalities and scales: image classifiers trained on
ImageNet, vision-language models (CLIP), and language models (BERT and mT5). We
provide code for using TRAK (and reproducing our work) at
https://github.com/MadryLab/trak .
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 17:56:22 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 17:37:50 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Park",
"Sung Min",
""
],
[
"Georgiev",
"Kristian",
""
],
[
"Ilyas",
"Andrew",
""
],
[
"Leclerc",
"Guillaume",
""
],
[
"Madry",
"Aleksander",
""
]
] | TITLE: TRAK: Attributing Model Behavior at Scale
ABSTRACT: The goal of data attribution is to trace model predictions back to training
data. Despite a long line of work towards this goal, existing approaches to
data attribution tend to force users to choose between computational
tractability and efficacy. That is, computationally tractable methods can
struggle with accurately attributing model predictions in non-convex settings
(e.g., in the context of deep neural networks), while methods that are
effective in such regimes require training thousands of models, which makes
them impractical for large models or datasets.
In this work, we introduce TRAK (Tracing with the Randomly-projected After
Kernel), a data attribution method that is both effective and computationally
tractable for large-scale, differentiable models. In particular, by leveraging
only a handful of trained models, TRAK can match the performance of attribution
methods that require training thousands of models. We demonstrate the utility
of TRAK across various modalities and scales: image classifiers trained on
ImageNet, vision-language models (CLIP), and language models (BERT and mT5). We
provide code for using TRAK (and reproducing our work) at
https://github.com/MadryLab/trak .
| no_new_dataset | 0.944434 |
2303.16589 | Mahum Naseer | Mahum Naseer and Muhammad Shafique | Poster: Link between Bias, Node Sensitivity and Long-Tail Distribution
in trained DNNs | To appear at the 16th IEEE International Conference on Software
Testing, Verification and Validation (ICST 2023), Dublin, Ireland | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Owing to their remarkable learning (and relearning) capabilities, deep neural
networks (DNNs) find use in numerous real-world applications. However, the
learning of these data-driven machine learning models is generally as good as
the data available to them for training. Hence, training datasets with
long-tail distribution pose a challenge for DNNs, since the DNNs trained on
them may provide a varying degree of classification performance across
different output classes. While the overall bias of such networks is already
highlighted in existing works, this work identifies the node bias that leads to
a varying sensitivity of the nodes for different output classes. To the best of
our knowledge, this is the first work highlighting this unique challenge in
DNNs, discussing its probable causes, and providing open challenges for this
new research direction. We support our reasoning using an empirical case study
of the networks trained on a real-world dataset.
| [
{
"version": "v1",
"created": "Wed, 29 Mar 2023 10:49:31 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 10:42:20 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Naseer",
"Mahum",
""
],
[
"Shafique",
"Muhammad",
""
]
] | TITLE: Poster: Link between Bias, Node Sensitivity and Long-Tail Distribution
in trained DNNs
ABSTRACT: Owing to their remarkable learning (and relearning) capabilities, deep neural
networks (DNNs) find use in numerous real-world applications. However, the
learning of these data-driven machine learning models is generally as good as
the data available to them for training. Hence, training datasets with
long-tail distribution pose a challenge for DNNs, since the DNNs trained on
them may provide a varying degree of classification performance across
different output classes. While the overall bias of such networks is already
highlighted in existing works, this work identifies the node bias that leads to
a varying sensitivity of the nodes for different output classes. To the best of
our knowledge, this is the first work highlighting this unique challenge in
DNNs, discussing its probable causes, and providing open challenges for this
new research direction. We support our reasoning using an empirical case study
of the networks trained on a real-world dataset.
| no_new_dataset | 0.948251 |
2303.17200 | Xubo Liu | Xubo Liu, Egor Lakomkin, Konstantinos Vougioukas, Pingchuan Ma,
Honglie Chen, Ruiming Xie, Morrie Doulaty, Niko Moritz, J\'achym Kol\'a\v{r},
Stavros Petridis, Maja Pantic, Christian Fuegen | SynthVSR: Scaling Up Visual Speech Recognition With Synthetic
Supervision | IEEE/CVF CVPR 2023 | null | null | null | cs.CV cs.AI cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently reported state-of-the-art results in visual speech recognition (VSR)
often rely on increasingly large amounts of video data, while the publicly
available transcribed video datasets are limited in size. In this paper, for
the first time, we study the potential of leveraging synthetic visual data for
VSR. Our method, termed SynthVSR, substantially improves the performance of VSR
systems with synthetic lip movements. The key idea behind SynthVSR is to
leverage a speech-driven lip animation model that generates lip movements
conditioned on the input speech. The speech-driven lip animation model is
trained on an unlabeled audio-visual dataset and could be further optimized
towards a pre-trained VSR model when labeled videos are available. As plenty of
transcribed acoustic data and face images are available, we are able to
generate large-scale synthetic data using the proposed lip animation model for
semi-supervised VSR training. We evaluate the performance of our approach on
the largest public VSR benchmark - Lip Reading Sentences 3 (LRS3). SynthVSR
achieves a WER of 43.3% with only 30 hours of real labeled data, outperforming
off-the-shelf approaches using thousands of hours of video. The WER is further
reduced to 27.9% when using all 438 hours of labeled data from LRS3, which is
on par with the state-of-the-art self-supervised AV-HuBERT method. Furthermore,
when combined with large-scale pseudo-labeled audio-visual data SynthVSR yields
a new state-of-the-art VSR WER of 16.9% using publicly available data only,
surpassing the recent state-of-the-art approaches trained with 29 times more
non-public machine-transcribed video data (90,000 hours). Finally, we perform
extensive ablation studies to understand the effect of each component in our
proposed method.
| [
{
"version": "v1",
"created": "Thu, 30 Mar 2023 07:43:27 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 06:30:19 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liu",
"Xubo",
""
],
[
"Lakomkin",
"Egor",
""
],
[
"Vougioukas",
"Konstantinos",
""
],
[
"Ma",
"Pingchuan",
""
],
[
"Chen",
"Honglie",
""
],
[
"Xie",
"Ruiming",
""
],
[
"Doulaty",
"Morrie",
""
],
[
"Moritz",
"Niko",
""
],
[
"Kolář",
"Jáchym",
""
],
[
"Petridis",
"Stavros",
""
],
[
"Pantic",
"Maja",
""
],
[
"Fuegen",
"Christian",
""
]
] | TITLE: SynthVSR: Scaling Up Visual Speech Recognition With Synthetic
Supervision
ABSTRACT: Recently reported state-of-the-art results in visual speech recognition (VSR)
often rely on increasingly large amounts of video data, while the publicly
available transcribed video datasets are limited in size. In this paper, for
the first time, we study the potential of leveraging synthetic visual data for
VSR. Our method, termed SynthVSR, substantially improves the performance of VSR
systems with synthetic lip movements. The key idea behind SynthVSR is to
leverage a speech-driven lip animation model that generates lip movements
conditioned on the input speech. The speech-driven lip animation model is
trained on an unlabeled audio-visual dataset and could be further optimized
towards a pre-trained VSR model when labeled videos are available. As plenty of
transcribed acoustic data and face images are available, we are able to
generate large-scale synthetic data using the proposed lip animation model for
semi-supervised VSR training. We evaluate the performance of our approach on
the largest public VSR benchmark - Lip Reading Sentences 3 (LRS3). SynthVSR
achieves a WER of 43.3% with only 30 hours of real labeled data, outperforming
off-the-shelf approaches using thousands of hours of video. The WER is further
reduced to 27.9% when using all 438 hours of labeled data from LRS3, which is
on par with the state-of-the-art self-supervised AV-HuBERT method. Furthermore,
when combined with large-scale pseudo-labeled audio-visual data SynthVSR yields
a new state-of-the-art VSR WER of 16.9% using publicly available data only,
surpassing the recent state-of-the-art approaches trained with 29 times more
non-public machine-transcribed video data (90,000 hours). Finally, we perform
extensive ablation studies to understand the effect of each component in our
proposed method.
| no_new_dataset | 0.952264 |
2303.17418 | Qisen Shang | Qisen Shang, Han Hu, Haojia Yu, Bo Xu, Libin Wang, Qing Zhu | Semantic Image Translation for Repairing the Texture Defects of Building
Models | 31 pages, 16 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The accurate representation of 3D building models in urban environments is
significantly hindered by challenges such as texture occlusion, blurring, and
missing details, which are difficult to mitigate through standard
photogrammetric texture mapping pipelines. Current image completion methods
often struggle to produce structured results and effectively handle the
intricate nature of highly-structured fa\c{c}ade textures with diverse
architectural styles. Furthermore, existing image synthesis methods encounter
difficulties in preserving high-frequency details and artificial regular
structures, which are essential for achieving realistic fa\c{c}ade texture
synthesis. To address these challenges, we introduce a novel approach for
synthesizing fa\c{c}ade texture images that authentically reflect the
architectural style from a structured label map, guided by a ground-truth
fa\c{c}ade image. In order to preserve fine details and regular structures, we
propose a regularity-aware multi-domain method that capitalizes on frequency
information and corner maps. We also incorporate SEAN blocks into our generator
to enable versatile style transfer. To generate plausible structured images
without undesirable regions, we employ image completion techniques to remove
occlusions according to semantics prior to image inference. Our proposed method
is also capable of synthesizing texture images with specific styles for
fa\c{c}ades that lack pre-existing textures, using manually annotated labels.
Experimental results on publicly available fa\c{c}ade image and 3D model
datasets demonstrate that our method yields superior results and effectively
addresses issues associated with flawed textures. The code and datasets will be
made publicly available for further research and development.
| [
{
"version": "v1",
"created": "Thu, 30 Mar 2023 14:38:53 GMT"
},
{
"version": "v2",
"created": "Sat, 1 Apr 2023 07:55:06 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Shang",
"Qisen",
""
],
[
"Hu",
"Han",
""
],
[
"Yu",
"Haojia",
""
],
[
"Xu",
"Bo",
""
],
[
"Wang",
"Libin",
""
],
[
"Zhu",
"Qing",
""
]
] | TITLE: Semantic Image Translation for Repairing the Texture Defects of Building
Models
ABSTRACT: The accurate representation of 3D building models in urban environments is
significantly hindered by challenges such as texture occlusion, blurring, and
missing details, which are difficult to mitigate through standard
photogrammetric texture mapping pipelines. Current image completion methods
often struggle to produce structured results and effectively handle the
intricate nature of highly-structured fa\c{c}ade textures with diverse
architectural styles. Furthermore, existing image synthesis methods encounter
difficulties in preserving high-frequency details and artificial regular
structures, which are essential for achieving realistic fa\c{c}ade texture
synthesis. To address these challenges, we introduce a novel approach for
synthesizing fa\c{c}ade texture images that authentically reflect the
architectural style from a structured label map, guided by a ground-truth
fa\c{c}ade image. In order to preserve fine details and regular structures, we
propose a regularity-aware multi-domain method that capitalizes on frequency
information and corner maps. We also incorporate SEAN blocks into our generator
to enable versatile style transfer. To generate plausible structured images
without undesirable regions, we employ image completion techniques to remove
occlusions according to semantics prior to image inference. Our proposed method
is also capable of synthesizing texture images with specific styles for
fa\c{c}ades that lack pre-existing textures, using manually annotated labels.
Experimental results on publicly available fa\c{c}ade image and 3D model
datasets demonstrate that our method yields superior results and effectively
addresses issues associated with flawed textures. The code and datasets will be
made publicly available for further research and development.
| no_new_dataset | 0.917451 |
2303.17774 | Gengxin Liu | Gengxin Liu, Qian Sun, Haibin Huang, Chongyang Ma, Yulan Guo, Li Yi,
Hui Huang, Ruizhen Hu | Semi-Weakly Supervised Object Kinematic Motion Prediction | CVPR 2023 | null | null | null | cs.CV cs.AI cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given a 3D object, kinematic motion prediction aims to identify the mobile
parts as well as the corresponding motion parameters. Due to the large
variations in both topological structure and geometric details of 3D objects,
this remains a challenging task and the lack of large scale labeled data also
constrain the performance of deep learning based approaches. In this paper, we
tackle the task of object kinematic motion prediction problem in a semi-weakly
supervised manner. Our key observations are two-fold. First, although 3D
dataset with fully annotated motion labels is limited, there are existing
datasets and methods for object part semantic segmentation at large scale.
Second, semantic part segmentation and mobile part segmentation is not always
consistent but it is possible to detect the mobile parts from the underlying 3D
structure. Towards this end, we propose a graph neural network to learn the map
between hierarchical part-level segmentation and mobile parts parameters, which
are further refined based on geometric alignment. This network can be first
trained on PartNet-Mobility dataset with fully labeled mobility information and
then applied on PartNet dataset with fine-grained and hierarchical part-level
segmentation. The network predictions yield a large scale of 3D objects with
pseudo labeled mobility information and can further be used for
weakly-supervised learning with pre-existing segmentation. Our experiments show
there are significant performance boosts with the augmented data for previous
method designed for kinematic motion prediction on 3D partial scans.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 02:37:36 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 02:36:17 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liu",
"Gengxin",
""
],
[
"Sun",
"Qian",
""
],
[
"Huang",
"Haibin",
""
],
[
"Ma",
"Chongyang",
""
],
[
"Guo",
"Yulan",
""
],
[
"Yi",
"Li",
""
],
[
"Huang",
"Hui",
""
],
[
"Hu",
"Ruizhen",
""
]
] | TITLE: Semi-Weakly Supervised Object Kinematic Motion Prediction
ABSTRACT: Given a 3D object, kinematic motion prediction aims to identify the mobile
parts as well as the corresponding motion parameters. Due to the large
variations in both topological structure and geometric details of 3D objects,
this remains a challenging task and the lack of large scale labeled data also
constrain the performance of deep learning based approaches. In this paper, we
tackle the task of object kinematic motion prediction problem in a semi-weakly
supervised manner. Our key observations are two-fold. First, although 3D
dataset with fully annotated motion labels is limited, there are existing
datasets and methods for object part semantic segmentation at large scale.
Second, semantic part segmentation and mobile part segmentation is not always
consistent but it is possible to detect the mobile parts from the underlying 3D
structure. Towards this end, we propose a graph neural network to learn the map
between hierarchical part-level segmentation and mobile parts parameters, which
are further refined based on geometric alignment. This network can be first
trained on PartNet-Mobility dataset with fully labeled mobility information and
then applied on PartNet dataset with fine-grained and hierarchical part-level
segmentation. The network predictions yield a large scale of 3D objects with
pseudo labeled mobility information and can further be used for
weakly-supervised learning with pre-existing segmentation. Our experiments show
there are significant performance boosts with the augmented data for previous
method designed for kinematic motion prediction on 3D partial scans.
| no_new_dataset | 0.948632 |
2303.17811 | Seonghoon Yu | Seonghoon Yu, Paul Hongsuck Seo, Jeany Son | Zero-shot Referring Image Segmentation with Global-Local Context
Features | CVPR 2023 | null | null | null | cs.CV cs.AI cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Referring image segmentation (RIS) aims to find a segmentation mask given a
referring expression grounded to a region of the input image. Collecting
labelled datasets for this task, however, is notoriously costly and
labor-intensive. To overcome this issue, we propose a simple yet effective
zero-shot referring image segmentation method by leveraging the pre-trained
cross-modal knowledge from CLIP. In order to obtain segmentation masks grounded
to the input text, we propose a mask-guided visual encoder that captures global
and local contextual information of an input image. By utilizing instance masks
obtained from off-the-shelf mask proposal techniques, our method is able to
segment fine-detailed Istance-level groundings. We also introduce a
global-local text encoder where the global feature captures complex
sentence-level semantics of the entire input expression while the local feature
focuses on the target noun phrase extracted by a dependency parser. In our
experiments, the proposed method outperforms several zero-shot baselines of the
task and even the weakly supervised referring expression segmentation method
with substantial margins. Our code is available at
https://github.com/Seonghoon-Yu/Zero-shot-RIS.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 06:00:50 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 08:58:36 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yu",
"Seonghoon",
""
],
[
"Seo",
"Paul Hongsuck",
""
],
[
"Son",
"Jeany",
""
]
] | TITLE: Zero-shot Referring Image Segmentation with Global-Local Context
Features
ABSTRACT: Referring image segmentation (RIS) aims to find a segmentation mask given a
referring expression grounded to a region of the input image. Collecting
labelled datasets for this task, however, is notoriously costly and
labor-intensive. To overcome this issue, we propose a simple yet effective
zero-shot referring image segmentation method by leveraging the pre-trained
cross-modal knowledge from CLIP. In order to obtain segmentation masks grounded
to the input text, we propose a mask-guided visual encoder that captures global
and local contextual information of an input image. By utilizing instance masks
obtained from off-the-shelf mask proposal techniques, our method is able to
segment fine-detailed Istance-level groundings. We also introduce a
global-local text encoder where the global feature captures complex
sentence-level semantics of the entire input expression while the local feature
focuses on the target noun phrase extracted by a dependency parser. In our
experiments, the proposed method outperforms several zero-shot baselines of the
task and even the weakly supervised referring expression segmentation method
with substantial margins. Our code is available at
https://github.com/Seonghoon-Yu/Zero-shot-RIS.
| no_new_dataset | 0.947186 |
2303.18171 | Md Yousuf Harun | Md Yousuf Harun, Jhair Gallardo, Tyler L. Hayes, Christopher Kanan | How Efficient Are Today's Continual Learning Algorithms? | To appear in the IEEE Conference on Computer Vision and Pattern
Recognition Workshop (CVPR-W) on Continual Learning in Computer Vision
(CLVision) 2023 | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Supervised Continual learning involves updating a deep neural network (DNN)
from an ever-growing stream of labeled data. While most work has focused on
overcoming catastrophic forgetting, one of the major motivations behind
continual learning is being able to efficiently update a network with new
information, rather than retraining from scratch on the training dataset as it
grows over time. Despite recent continual learning methods largely solving the
catastrophic forgetting problem, there has been little attention paid to the
efficiency of these algorithms. Here, we study recent methods for incremental
class learning and illustrate that many are highly inefficient in terms of
compute, memory, and storage. Some methods even require more compute than
training from scratch! We argue that for continual learning to have real-world
applicability, the research community cannot ignore the resources used by these
algorithms. There is more to continual learning than mitigating catastrophic
forgetting.
| [
{
"version": "v1",
"created": "Wed, 29 Mar 2023 18:52:10 GMT"
},
{
"version": "v2",
"created": "Mon, 3 Apr 2023 13:53:33 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Harun",
"Md Yousuf",
""
],
[
"Gallardo",
"Jhair",
""
],
[
"Hayes",
"Tyler L.",
""
],
[
"Kanan",
"Christopher",
""
]
] | TITLE: How Efficient Are Today's Continual Learning Algorithms?
ABSTRACT: Supervised Continual learning involves updating a deep neural network (DNN)
from an ever-growing stream of labeled data. While most work has focused on
overcoming catastrophic forgetting, one of the major motivations behind
continual learning is being able to efficiently update a network with new
information, rather than retraining from scratch on the training dataset as it
grows over time. Despite recent continual learning methods largely solving the
catastrophic forgetting problem, there has been little attention paid to the
efficiency of these algorithms. Here, we study recent methods for incremental
class learning and illustrate that many are highly inefficient in terms of
compute, memory, and storage. Some methods even require more compute than
training from scratch! We argue that for continual learning to have real-world
applicability, the research community cannot ignore the resources used by these
algorithms. There is more to continual learning than mitigating catastrophic
forgetting.
| no_new_dataset | 0.942082 |
2304.00018 | Liu Hongen | Hongen Liu | Video text tracking for dense and small text based on pp-yoloe-r and
sort algorithm | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although end-to-end video text spotting methods based on Transformer can
model long-range dependencies and simplify the train process, it will lead to
large computation cost with the increase of the frame size in the input video.
Therefore, considering the resolution of ICDAR 2023 DSText is 1080 * 1920 and
slicing the video frame into several areas will destroy the spatial correlation
of text, we divided the small and dense text spotting into two tasks, text
detection and tracking. For text detection, we adopt the PP-YOLOE-R which is
proven effective in small object detection as our detection model. For text
detection, we use the sort algorithm for high inference speed. Experiments on
DSText dataset demonstrate that our method is competitive on small and dense
text spotting.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 05:40:39 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liu",
"Hongen",
""
]
] | TITLE: Video text tracking for dense and small text based on pp-yoloe-r and
sort algorithm
ABSTRACT: Although end-to-end video text spotting methods based on Transformer can
model long-range dependencies and simplify the train process, it will lead to
large computation cost with the increase of the frame size in the input video.
Therefore, considering the resolution of ICDAR 2023 DSText is 1080 * 1920 and
slicing the video frame into several areas will destroy the spatial correlation
of text, we divided the small and dense text spotting into two tasks, text
detection and tracking. For text detection, we adopt the PP-YOLOE-R which is
proven effective in small object detection as our detection model. For text
detection, we use the sort algorithm for high inference speed. Experiments on
DSText dataset demonstrate that our method is competitive on small and dense
text spotting.
| no_new_dataset | 0.950319 |
2304.00019 | Rafael Ferreira da Silva | Rafael Ferreira da Silva, Rosa M. Badia, Venkat Bala, Debbie Bard,
Peer-Timo Bremer, Ian Buckley, Silvina Caino-Lores, Kyle Chard, Carole Goble,
Shantenu Jha, Daniel S. Katz, Daniel Laney, Manish Parashar, Frederic Suter,
Nick Tyler, Thomas Uram, Ilkay Altintas, Stefan Andersson, William Arndt,
Juan Aznar, Jonathan Bader, Bartosz Balis, Chris Blanton, Kelly Rosa
Braghetto, Aharon Brodutch, Paul Brunk, Henri Casanova, Alba Cervera Lierta,
Justin Chigu, Taina Coleman, Nick Collier, Iacopo Colonnelli, Frederik
Coppens, Michael Crusoe, Will Cunningham, Bruno de Paula Kinoshita, Paolo Di
Tommaso, Charles Doutriaux, Matthew Downton, Wael Elwasif, Bjoern Enders,
Chris Erdmann, Thomas Fahringer, Ludmilla Figueiredo, Rosa Filgueira, Martin
Foltin, Anne Fouilloux, Luiz Gadelha, Andy Gallo, Artur Garcia Saez, Daniel
Garijo, Roman Gerlach, Ryan Grant, Samuel Grayson, Patricia Grubel, Johan
Gustafsson, Valerie Hayot-Sasson, Oscar Hernandez, Marcus Hilbrich, AnnMary
Justine, Ian Laflotte, Fabian Lehmann, Andre Luckow, Jakob Luettgau, Ketan
Maheshwari, Motohiko Matsuda, Doriana Medic, Pete Mendygral, Marek
Michalewicz, Jorji Nonaka, Maciej Pawlik, Loic Pottier, Line Pouchard,
Mathias Putz, Santosh Kumar Radha, Lavanya Ramakrishnan, Sashko Ristov, Paul
Romano, Daniel Rosendo, Martin Ruefenacht, Katarzyna Rycerz, Nishant Saurabh,
Volodymyr Savchenko, Martin Schulz, Christine Simpson, Raul Sirvent, Tyler
Skluzacek, Stian Soiland-Reyes, Renan Souza, Sreenivas Rangan Sukumar, Ziheng
Sun, Alan Sussman, Douglas Thain, Mikhail Titov, Benjamin Tovar, Aalap
Tripathy, Matteo Turilli, Bartosz Tuznik, Hubertus van Dam, Aurelio Vivas,
Logan Ward, Patrick Widener, Sean Wilkinson, Justyna Zawalska, Mahnoor
Zulfiqar | Workflows Community Summit 2022: A Roadmap Revolution | null | null | 10.5281/zenodo.7750670 | ORNL/TM-2023/2885 | cs.DC | http://creativecommons.org/licenses/by/4.0/ | Scientific workflows have become integral tools in broad scientific computing
use cases. Science discovery is increasingly dependent on workflows to
orchestrate large and complex scientific experiments that range from execution
of a cloud-based data preprocessing pipeline to multi-facility
instrument-to-edge-to-HPC computational workflows. Given the changing landscape
of scientific computing and the evolving needs of emerging scientific
applications, it is paramount that the development of novel scientific
workflows and system functionalities seek to increase the efficiency,
resilience, and pervasiveness of existing systems and applications.
Specifically, the proliferation of machine learning/artificial intelligence
(ML/AI) workflows, need for processing large scale datasets produced by
instruments at the edge, intensification of near real-time data processing,
support for long-term experiment campaigns, and emergence of quantum computing
as an adjunct to HPC, have significantly changed the functional and operational
requirements of workflow systems. Workflow systems now need to, for example,
support data streams from the edge-to-cloud-to-HPC enable the management of
many small-sized files, allow data reduction while ensuring high accuracy,
orchestrate distributed services (workflows, instruments, data movement,
provenance, publication, etc.) across computing and user facilities, among
others. Further, to accelerate science, it is also necessary that these systems
implement specifications/standards and APIs for seamless (horizontal and
vertical) integration between systems and applications, as well as enabling the
publication of workflows and their associated products according to the FAIR
principles. This document reports on discussions and findings from the 2022
international edition of the Workflows Community Summit that took place on
November 29 and 30, 2022.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 07:33:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"da Silva",
"Rafael Ferreira",
""
],
[
"Badia",
"Rosa M.",
""
],
[
"Bala",
"Venkat",
""
],
[
"Bard",
"Debbie",
""
],
[
"Bremer",
"Peer-Timo",
""
],
[
"Buckley",
"Ian",
""
],
[
"Caino-Lores",
"Silvina",
""
],
[
"Chard",
"Kyle",
""
],
[
"Goble",
"Carole",
""
],
[
"Jha",
"Shantenu",
""
],
[
"Katz",
"Daniel S.",
""
],
[
"Laney",
"Daniel",
""
],
[
"Parashar",
"Manish",
""
],
[
"Suter",
"Frederic",
""
],
[
"Tyler",
"Nick",
""
],
[
"Uram",
"Thomas",
""
],
[
"Altintas",
"Ilkay",
""
],
[
"Andersson",
"Stefan",
""
],
[
"Arndt",
"William",
""
],
[
"Aznar",
"Juan",
""
],
[
"Bader",
"Jonathan",
""
],
[
"Balis",
"Bartosz",
""
],
[
"Blanton",
"Chris",
""
],
[
"Braghetto",
"Kelly Rosa",
""
],
[
"Brodutch",
"Aharon",
""
],
[
"Brunk",
"Paul",
""
],
[
"Casanova",
"Henri",
""
],
[
"Lierta",
"Alba Cervera",
""
],
[
"Chigu",
"Justin",
""
],
[
"Coleman",
"Taina",
""
],
[
"Collier",
"Nick",
""
],
[
"Colonnelli",
"Iacopo",
""
],
[
"Coppens",
"Frederik",
""
],
[
"Crusoe",
"Michael",
""
],
[
"Cunningham",
"Will",
""
],
[
"Kinoshita",
"Bruno de Paula",
""
],
[
"Di Tommaso",
"Paolo",
""
],
[
"Doutriaux",
"Charles",
""
],
[
"Downton",
"Matthew",
""
],
[
"Elwasif",
"Wael",
""
],
[
"Enders",
"Bjoern",
""
],
[
"Erdmann",
"Chris",
""
],
[
"Fahringer",
"Thomas",
""
],
[
"Figueiredo",
"Ludmilla",
""
],
[
"Filgueira",
"Rosa",
""
],
[
"Foltin",
"Martin",
""
],
[
"Fouilloux",
"Anne",
""
],
[
"Gadelha",
"Luiz",
""
],
[
"Gallo",
"Andy",
""
],
[
"Saez",
"Artur Garcia",
""
],
[
"Garijo",
"Daniel",
""
],
[
"Gerlach",
"Roman",
""
],
[
"Grant",
"Ryan",
""
],
[
"Grayson",
"Samuel",
""
],
[
"Grubel",
"Patricia",
""
],
[
"Gustafsson",
"Johan",
""
],
[
"Hayot-Sasson",
"Valerie",
""
],
[
"Hernandez",
"Oscar",
""
],
[
"Hilbrich",
"Marcus",
""
],
[
"Justine",
"AnnMary",
""
],
[
"Laflotte",
"Ian",
""
],
[
"Lehmann",
"Fabian",
""
],
[
"Luckow",
"Andre",
""
],
[
"Luettgau",
"Jakob",
""
],
[
"Maheshwari",
"Ketan",
""
],
[
"Matsuda",
"Motohiko",
""
],
[
"Medic",
"Doriana",
""
],
[
"Mendygral",
"Pete",
""
],
[
"Michalewicz",
"Marek",
""
],
[
"Nonaka",
"Jorji",
""
],
[
"Pawlik",
"Maciej",
""
],
[
"Pottier",
"Loic",
""
],
[
"Pouchard",
"Line",
""
],
[
"Putz",
"Mathias",
""
],
[
"Radha",
"Santosh Kumar",
""
],
[
"Ramakrishnan",
"Lavanya",
""
],
[
"Ristov",
"Sashko",
""
],
[
"Romano",
"Paul",
""
],
[
"Rosendo",
"Daniel",
""
],
[
"Ruefenacht",
"Martin",
""
],
[
"Rycerz",
"Katarzyna",
""
],
[
"Saurabh",
"Nishant",
""
],
[
"Savchenko",
"Volodymyr",
""
],
[
"Schulz",
"Martin",
""
],
[
"Simpson",
"Christine",
""
],
[
"Sirvent",
"Raul",
""
],
[
"Skluzacek",
"Tyler",
""
],
[
"Soiland-Reyes",
"Stian",
""
],
[
"Souza",
"Renan",
""
],
[
"Sukumar",
"Sreenivas Rangan",
""
],
[
"Sun",
"Ziheng",
""
],
[
"Sussman",
"Alan",
""
],
[
"Thain",
"Douglas",
""
],
[
"Titov",
"Mikhail",
""
],
[
"Tovar",
"Benjamin",
""
],
[
"Tripathy",
"Aalap",
""
],
[
"Turilli",
"Matteo",
""
],
[
"Tuznik",
"Bartosz",
""
],
[
"van Dam",
"Hubertus",
""
],
[
"Vivas",
"Aurelio",
""
],
[
"Ward",
"Logan",
""
],
[
"Widener",
"Patrick",
""
],
[
"Wilkinson",
"Sean",
""
],
[
"Zawalska",
"Justyna",
""
],
[
"Zulfiqar",
"Mahnoor",
""
]
] | TITLE: Workflows Community Summit 2022: A Roadmap Revolution
ABSTRACT: Scientific workflows have become integral tools in broad scientific computing
use cases. Science discovery is increasingly dependent on workflows to
orchestrate large and complex scientific experiments that range from execution
of a cloud-based data preprocessing pipeline to multi-facility
instrument-to-edge-to-HPC computational workflows. Given the changing landscape
of scientific computing and the evolving needs of emerging scientific
applications, it is paramount that the development of novel scientific
workflows and system functionalities seek to increase the efficiency,
resilience, and pervasiveness of existing systems and applications.
Specifically, the proliferation of machine learning/artificial intelligence
(ML/AI) workflows, need for processing large scale datasets produced by
instruments at the edge, intensification of near real-time data processing,
support for long-term experiment campaigns, and emergence of quantum computing
as an adjunct to HPC, have significantly changed the functional and operational
requirements of workflow systems. Workflow systems now need to, for example,
support data streams from the edge-to-cloud-to-HPC enable the management of
many small-sized files, allow data reduction while ensuring high accuracy,
orchestrate distributed services (workflows, instruments, data movement,
provenance, publication, etc.) across computing and user facilities, among
others. Further, to accelerate science, it is also necessary that these systems
implement specifications/standards and APIs for seamless (horizontal and
vertical) integration between systems and applications, as well as enabling the
publication of workflows and their associated products according to the FAIR
principles. This document reports on discussions and findings from the 2022
international edition of the Workflows Community Summit that took place on
November 29 and 30, 2022.
| no_new_dataset | 0.945701 |
2304.00022 | Chuangguan Ye | Chuangguan Ye, Hongyuan Zhu, Yongbin Liao, Yanggang Zhang, Tao Chen,
Jiayuan Fan | What Makes for Effective Few-shot Point Cloud Classification? | Accepted by WACV 2022. arXiv admin note: text overlap with
arXiv:2303.18210 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the emergence of powerful computing resources and large-scale
annotated datasets, deep learning has seen wide applications in our daily life.
However, most current methods require extensive data collection and retraining
when dealing with novel classes never seen before. On the other hand, we humans
can quickly recognize new classes by looking at a few samples, which motivates
the recent popularity of few-shot learning (FSL) in machine learning
communities. Most current FSL approaches work on 2D image domain, however, its
implication in 3D perception is relatively under-explored. Not only needs to
recognize the unseen examples as in 2D domain, 3D few-shot learning is more
challenging with unordered structures, high intra-class variances, and subtle
inter-class differences. Moreover, different architectures and learning
algorithms make it difficult to study the effectiveness of existing 2D methods
when migrating to the 3D domain. In this work, for the first time, we perform
systematic and extensive studies of recent 2D FSL and 3D backbone networks for
benchmarking few-shot point cloud classification, and we suggest a strong
baseline and learning architectures for 3D FSL. Then, we propose a novel
plug-and-play component called Cross-Instance Adaptation (CIA) module, to
address the high intra-class variances and subtle inter-class differences
issues, which can be easily inserted into current baselines with significant
performance improvement. Extensive experiments on two newly introduced
benchmark datasets, ModelNet40-FS and ShapeNet70-FS, demonstrate the
superiority of our proposed network for 3D FSL.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 15:55:06 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ye",
"Chuangguan",
""
],
[
"Zhu",
"Hongyuan",
""
],
[
"Liao",
"Yongbin",
""
],
[
"Zhang",
"Yanggang",
""
],
[
"Chen",
"Tao",
""
],
[
"Fan",
"Jiayuan",
""
]
] | TITLE: What Makes for Effective Few-shot Point Cloud Classification?
ABSTRACT: Due to the emergence of powerful computing resources and large-scale
annotated datasets, deep learning has seen wide applications in our daily life.
However, most current methods require extensive data collection and retraining
when dealing with novel classes never seen before. On the other hand, we humans
can quickly recognize new classes by looking at a few samples, which motivates
the recent popularity of few-shot learning (FSL) in machine learning
communities. Most current FSL approaches work on 2D image domain, however, its
implication in 3D perception is relatively under-explored. Not only needs to
recognize the unseen examples as in 2D domain, 3D few-shot learning is more
challenging with unordered structures, high intra-class variances, and subtle
inter-class differences. Moreover, different architectures and learning
algorithms make it difficult to study the effectiveness of existing 2D methods
when migrating to the 3D domain. In this work, for the first time, we perform
systematic and extensive studies of recent 2D FSL and 3D backbone networks for
benchmarking few-shot point cloud classification, and we suggest a strong
baseline and learning architectures for 3D FSL. Then, we propose a novel
plug-and-play component called Cross-Instance Adaptation (CIA) module, to
address the high intra-class variances and subtle inter-class differences
issues, which can be easily inserted into current baselines with significant
performance improvement. Extensive experiments on two newly introduced
benchmark datasets, ModelNet40-FS and ShapeNet70-FS, demonstrate the
superiority of our proposed network for 3D FSL.
| no_new_dataset | 0.939248 |
2304.00046 | Bogdan Mazoure | Bogdan Mazoure, Jake Bruce, Doina Precup, Rob Fergus, Ankit Anand | Accelerating exploration and representation learning with offline
pre-training | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | Sequential decision-making agents struggle with long horizon tasks, since
solving them requires multi-step reasoning. Most reinforcement learning (RL)
algorithms address this challenge by improved credit assignment, introducing
memory capability, altering the agent's intrinsic motivation (i.e. exploration)
or its worldview (i.e. knowledge representation). Many of these components
could be learned from offline data. In this work, we follow the hypothesis that
exploration and representation learning can be improved by separately learning
two different models from a single offline dataset. We show that learning a
state representation using noise-contrastive estimation and a model of
auxiliary reward separately from a single collection of human demonstrations
can significantly improve the sample efficiency on the challenging NetHack
benchmark. We also ablate various components of our experimental setting and
highlight crucial insights.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 18:03:30 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Mazoure",
"Bogdan",
""
],
[
"Bruce",
"Jake",
""
],
[
"Precup",
"Doina",
""
],
[
"Fergus",
"Rob",
""
],
[
"Anand",
"Ankit",
""
]
] | TITLE: Accelerating exploration and representation learning with offline
pre-training
ABSTRACT: Sequential decision-making agents struggle with long horizon tasks, since
solving them requires multi-step reasoning. Most reinforcement learning (RL)
algorithms address this challenge by improved credit assignment, introducing
memory capability, altering the agent's intrinsic motivation (i.e. exploration)
or its worldview (i.e. knowledge representation). Many of these components
could be learned from offline data. In this work, we follow the hypothesis that
exploration and representation learning can be improved by separately learning
two different models from a single offline dataset. We show that learning a
state representation using noise-contrastive estimation and a model of
auxiliary reward separately from a single collection of human demonstrations
can significantly improve the sample efficiency on the challenging NetHack
benchmark. We also ablate various components of our experimental setting and
highlight crucial insights.
| no_new_dataset | 0.945551 |
2304.00049 | He Zhao | Mohammadi Kiarash and Zhao He and Mengyao Zhai and Frederick Tung | Ranking Regularization for Critical Rare Classes: Minimizing False
Positives at a High True Positive Rate | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In many real-world settings, the critical class is rare and a missed
detection carries a disproportionately high cost. For example, tumors are rare
and a false negative diagnosis could have severe consequences on treatment
outcomes; fraudulent banking transactions are rare and an undetected occurrence
could result in significant losses or legal penalties. In such contexts,
systems are often operated at a high true positive rate, which may require
tolerating high false positives. In this paper, we present a novel approach to
address the challenge of minimizing false positives for systems that need to
operate at a high true positive rate. We propose a ranking-based regularization
(RankReg) approach that is easy to implement, and show empirically that it not
only effectively reduces false positives, but also complements conventional
imbalanced learning losses. With this novel technique in hand, we conduct a
series of experiments on three broadly explored datasets (CIFAR-10&100 and
Melanoma) and show that our approach lifts the previous state-of-the-art
performance by notable margins.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 18:05:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Kiarash",
"Mohammadi",
""
],
[
"He",
"Zhao",
""
],
[
"Zhai",
"Mengyao",
""
],
[
"Tung",
"Frederick",
""
]
] | TITLE: Ranking Regularization for Critical Rare Classes: Minimizing False
Positives at a High True Positive Rate
ABSTRACT: In many real-world settings, the critical class is rare and a missed
detection carries a disproportionately high cost. For example, tumors are rare
and a false negative diagnosis could have severe consequences on treatment
outcomes; fraudulent banking transactions are rare and an undetected occurrence
could result in significant losses or legal penalties. In such contexts,
systems are often operated at a high true positive rate, which may require
tolerating high false positives. In this paper, we present a novel approach to
address the challenge of minimizing false positives for systems that need to
operate at a high true positive rate. We propose a ranking-based regularization
(RankReg) approach that is easy to implement, and show empirically that it not
only effectively reduces false positives, but also complements conventional
imbalanced learning losses. With this novel technique in hand, we conduct a
series of experiments on three broadly explored datasets (CIFAR-10&100 and
Melanoma) and show that our approach lifts the previous state-of-the-art
performance by notable margins.
| no_new_dataset | 0.952794 |
2304.00095 | Daniele Boffi | Daniele Boffi, Ramon Codina, and \"Onder T\"urk | Finite element formulations for Maxwell's eigenvalue problem using
continuous Lagrangian interpolations | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider nodal-based Lagrangian interpolations for the finite element
approximation of the Maxwell eigenvalue problem. The first approach introduced
is a standard Galerkin method on Powell-Sabin meshes, which has recently been
shown to yield convergent approximations in two dimensions, whereas the other
two are stabilized formulations that can be motivated by a variational
multiscale approach. For the latter, a mixed formulation equivalent to the
original problem is used, in which the operator has a saddle point structure.
The Lagrange multiplier introduced to enforce the divergence constraint
vanishes in an appropriate functional setting. The first stabilized method we
consider consists of an augmented formulation with the introduction of a mesh
dependent term that can be regarded as the Laplacian of the multiplier of the
divergence constraint. The second formulation is based on orthogonal
projections, which can be recast as a residual based stabilization technique.
We rely on the classical spectral theory to analyze the approximating methods
for the eigenproblem. The stability and convergence aspects are inherited from
the associated source problems. We investigate the numerical performance of the
proposed formulations and provide some convergence results validating the
theoretical ones for several benchmark tests, including ones with smooth and
singular solutions.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 19:43:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Boffi",
"Daniele",
""
],
[
"Codina",
"Ramon",
""
],
[
"Türk",
"Önder",
""
]
] | TITLE: Finite element formulations for Maxwell's eigenvalue problem using
continuous Lagrangian interpolations
ABSTRACT: We consider nodal-based Lagrangian interpolations for the finite element
approximation of the Maxwell eigenvalue problem. The first approach introduced
is a standard Galerkin method on Powell-Sabin meshes, which has recently been
shown to yield convergent approximations in two dimensions, whereas the other
two are stabilized formulations that can be motivated by a variational
multiscale approach. For the latter, a mixed formulation equivalent to the
original problem is used, in which the operator has a saddle point structure.
The Lagrange multiplier introduced to enforce the divergence constraint
vanishes in an appropriate functional setting. The first stabilized method we
consider consists of an augmented formulation with the introduction of a mesh
dependent term that can be regarded as the Laplacian of the multiplier of the
divergence constraint. The second formulation is based on orthogonal
projections, which can be recast as a residual based stabilization technique.
We rely on the classical spectral theory to analyze the approximating methods
for the eigenproblem. The stability and convergence aspects are inherited from
the associated source problems. We investigate the numerical performance of the
proposed formulations and provide some convergence results validating the
theoretical ones for several benchmark tests, including ones with smooth and
singular solutions.
| no_new_dataset | 0.946349 |
2304.00114 | Daniel Campos | Daniel Campos, ChengXiang Zhai | Dense Sparse Retrieval: Using Sparse Language Models for Inference
Efficient Dense Retrieval | null | null | null | null | cs.IR cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Vector-based retrieval systems have become a common staple for academic and
industrial search applications because they provide a simple and scalable way
of extending the search to leverage contextual representations for documents
and queries. As these vector-based systems rely on contextual language models,
their usage commonly requires GPUs, which can be expensive and difficult to
manage. Given recent advances in introducing sparsity into language models for
improved inference efficiency, in this paper, we study how sparse language
models can be used for dense retrieval to improve inference efficiency. Using
the popular retrieval library Tevatron and the MSMARCO, NQ, and TriviaQA
datasets, we find that sparse language models can be used as direct
replacements with little to no drop in accuracy and up to 4.3x improved
inference speeds
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 20:21:32 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Campos",
"Daniel",
""
],
[
"Zhai",
"ChengXiang",
""
]
] | TITLE: Dense Sparse Retrieval: Using Sparse Language Models for Inference
Efficient Dense Retrieval
ABSTRACT: Vector-based retrieval systems have become a common staple for academic and
industrial search applications because they provide a simple and scalable way
of extending the search to leverage contextual representations for documents
and queries. As these vector-based systems rely on contextual language models,
their usage commonly requires GPUs, which can be expensive and difficult to
manage. Given recent advances in introducing sparsity into language models for
improved inference efficiency, in this paper, we study how sparse language
models can be used for dense retrieval to improve inference efficiency. Using
the popular retrieval library Tevatron and the MSMARCO, NQ, and TriviaQA
datasets, we find that sparse language models can be used as direct
replacements with little to no drop in accuracy and up to 4.3x improved
inference speeds
| no_new_dataset | 0.942507 |
2304.00121 | Ryan Koo | Ryan Koo, Anna Martin, Linghe Wang, Dongyeop Kang | Decoding the End-to-end Writing Trajectory in Scholarly Manuscripts | null | null | null | null | cs.CL cs.HC | http://creativecommons.org/licenses/by/4.0/ | Scholarly writing presents a complex space that generally follows a
methodical procedure to plan and produce both rationally sound and creative
compositions. Recent works involving large language models (LLM) demonstrate
considerable success in text generation and revision tasks; however, LLMs still
struggle to provide structural and creative feedback on the document level that
is crucial to academic writing. In this paper, we introduce a novel taxonomy
that categorizes scholarly writing behaviors according to intention, writer
actions, and the information types of the written data. We also provide
ManuScript, an original dataset annotated with a simplified version of our
taxonomy to show writer actions and the intentions behind them. Motivated by
cognitive writing theory, our taxonomy for scientific papers includes three
levels of categorization in order to trace the general writing flow and
identify the distinct writer activities embedded within each higher-level
process. ManuScript intends to provide a complete picture of the scholarly
writing process by capturing the linearity and non-linearity of writing
trajectory, such that writing assistants can provide stronger feedback and
suggestions on an end-to-end level. The collected writing trajectories are
viewed at https://minnesotanlp.github.io/REWARD_demo/
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 20:33:03 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Koo",
"Ryan",
""
],
[
"Martin",
"Anna",
""
],
[
"Wang",
"Linghe",
""
],
[
"Kang",
"Dongyeop",
""
]
] | TITLE: Decoding the End-to-end Writing Trajectory in Scholarly Manuscripts
ABSTRACT: Scholarly writing presents a complex space that generally follows a
methodical procedure to plan and produce both rationally sound and creative
compositions. Recent works involving large language models (LLM) demonstrate
considerable success in text generation and revision tasks; however, LLMs still
struggle to provide structural and creative feedback on the document level that
is crucial to academic writing. In this paper, we introduce a novel taxonomy
that categorizes scholarly writing behaviors according to intention, writer
actions, and the information types of the written data. We also provide
ManuScript, an original dataset annotated with a simplified version of our
taxonomy to show writer actions and the intentions behind them. Motivated by
cognitive writing theory, our taxonomy for scientific papers includes three
levels of categorization in order to trace the general writing flow and
identify the distinct writer activities embedded within each higher-level
process. ManuScript intends to provide a complete picture of the scholarly
writing process by capturing the linearity and non-linearity of writing
trajectory, such that writing assistants can provide stronger feedback and
suggestions on an end-to-end level. The collected writing trajectories are
viewed at https://minnesotanlp.github.io/REWARD_demo/
| new_dataset | 0.956145 |
2304.00129 | David Froelicher | David Froelicher, Hyunghoon Cho, Manaswitha Edupalli, Joao Sa Sousa,
Jean-Philippe Bossuat, Apostolos Pyrgelis, Juan R. Troncoso-Pastoriza, Bonnie
Berger and Jean-Pierre Hubaux | Scalable and Privacy-Preserving Federated Principal Component Analysis | Published elsewhere. IEEE Symposium on Security and Privacy 2023 | null | 10.1109/SP46215.2023.00051 | null | cs.CR | http://creativecommons.org/licenses/by/4.0/ | Principal component analysis (PCA) is an essential algorithm for
dimensionality reduction in many data science domains. We address the problem
of performing a federated PCA on private data distributed among multiple data
providers while ensuring data confidentiality. Our solution, SF-PCA, is an
end-to-end secure system that preserves the confidentiality of both the
original data and all intermediate results in a passive-adversary model with up
to all-but-one colluding parties. SF-PCA jointly leverages multiparty
homomorphic encryption, interactive protocols, and edge computing to
efficiently interleave computations on local cleartext data with operations on
collectively encrypted data. SF-PCA obtains results as accurate as non-secure
centralized solutions, independently of the data distribution among the
parties. It scales linearly or better with the dataset dimensions and with the
number of data providers. SF-PCA is more precise than existing approaches that
approximate the solution by combining local analysis results, and between 3x
and 250x faster than privacy-preserving alternatives based solely on secure
multiparty computation or homomorphic encryption. Our work demonstrates the
practical applicability of secure and federated PCA on private distributed
datasets.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 21:06:57 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Froelicher",
"David",
""
],
[
"Cho",
"Hyunghoon",
""
],
[
"Edupalli",
"Manaswitha",
""
],
[
"Sousa",
"Joao Sa",
""
],
[
"Bossuat",
"Jean-Philippe",
""
],
[
"Pyrgelis",
"Apostolos",
""
],
[
"Troncoso-Pastoriza",
"Juan R.",
""
],
[
"Berger",
"Bonnie",
""
],
[
"Hubaux",
"Jean-Pierre",
""
]
] | TITLE: Scalable and Privacy-Preserving Federated Principal Component Analysis
ABSTRACT: Principal component analysis (PCA) is an essential algorithm for
dimensionality reduction in many data science domains. We address the problem
of performing a federated PCA on private data distributed among multiple data
providers while ensuring data confidentiality. Our solution, SF-PCA, is an
end-to-end secure system that preserves the confidentiality of both the
original data and all intermediate results in a passive-adversary model with up
to all-but-one colluding parties. SF-PCA jointly leverages multiparty
homomorphic encryption, interactive protocols, and edge computing to
efficiently interleave computations on local cleartext data with operations on
collectively encrypted data. SF-PCA obtains results as accurate as non-secure
centralized solutions, independently of the data distribution among the
parties. It scales linearly or better with the dataset dimensions and with the
number of data providers. SF-PCA is more precise than existing approaches that
approximate the solution by combining local analysis results, and between 3x
and 250x faster than privacy-preserving alternatives based solely on secure
multiparty computation or homomorphic encryption. Our work demonstrates the
practical applicability of secure and federated PCA on private distributed
datasets.
| no_new_dataset | 0.945601 |
2304.00145 | Ziyun Yang | Ziyun Yang and Sina Farsiu | Directional Connectivity-based Segmentation of Medical Images | Accepted by CVPR 2023 | null | null | null | eess.IV cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anatomical consistency in biomarker segmentation is crucial for many medical
image analysis tasks. A promising paradigm for achieving anatomically
consistent segmentation via deep networks is incorporating pixel connectivity,
a basic concept in digital topology, to model inter-pixel relationships.
However, previous works on connectivity modeling have ignored the rich
channel-wise directional information in the latent space. In this work, we
demonstrate that effective disentanglement of directional sub-space from the
shared latent space can significantly enhance the feature representation in the
connectivity-based network. To this end, we propose a directional connectivity
modeling scheme for segmentation that decouples, tracks, and utilizes the
directional information across the network. Experiments on various public
medical image segmentation benchmarks show the effectiveness of our model as
compared to the state-of-the-art methods. Code is available at
https://github.com/Zyun-Y/DconnNet.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 21:48:41 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yang",
"Ziyun",
""
],
[
"Farsiu",
"Sina",
""
]
] | TITLE: Directional Connectivity-based Segmentation of Medical Images
ABSTRACT: Anatomical consistency in biomarker segmentation is crucial for many medical
image analysis tasks. A promising paradigm for achieving anatomically
consistent segmentation via deep networks is incorporating pixel connectivity,
a basic concept in digital topology, to model inter-pixel relationships.
However, previous works on connectivity modeling have ignored the rich
channel-wise directional information in the latent space. In this work, we
demonstrate that effective disentanglement of directional sub-space from the
shared latent space can significantly enhance the feature representation in the
connectivity-based network. To this end, we propose a directional connectivity
modeling scheme for segmentation that decouples, tracks, and utilizes the
directional information across the network. Experiments on various public
medical image segmentation benchmarks show the effectiveness of our model as
compared to the state-of-the-art methods. Code is available at
https://github.com/Zyun-Y/DconnNet.
| no_new_dataset | 0.949669 |
2304.00150 | Artur Petrov Toshev | Artur P. Toshev, Gianluca Galletti, Johannes Brandstetter, Stefan
Adami and Nikolaus A. Adams | E($3$) Equivariant Graph Neural Networks for Particle-Based Fluid
Mechanics | ICLR 2023 Workshop on Physics for Machine Learning | null | null | null | cs.LG physics.flu-dyn | http://creativecommons.org/licenses/by/4.0/ | We contribute to the vastly growing field of machine learning for engineering
systems by demonstrating that equivariant graph neural networks have the
potential to learn more accurate dynamic-interaction models than their
non-equivariant counterparts. We benchmark two well-studied fluid flow systems,
namely the 3D decaying Taylor-Green vortex and the 3D reverse Poiseuille flow,
and compare equivariant graph neural networks to their non-equivariant
counterparts on different performance measures, such as kinetic energy or
Sinkhorn distance. Such measures are typically used in engineering to validate
numerical solvers. Our main findings are that while being rather slow to train
and evaluate, equivariant models learn more physically accurate interactions.
This indicates opportunities for future work towards coarse-grained models for
turbulent flows, and generalization across system dynamics and parameters.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 21:56:35 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Toshev",
"Artur P.",
""
],
[
"Galletti",
"Gianluca",
""
],
[
"Brandstetter",
"Johannes",
""
],
[
"Adami",
"Stefan",
""
],
[
"Adams",
"Nikolaus A.",
""
]
] | TITLE: E($3$) Equivariant Graph Neural Networks for Particle-Based Fluid
Mechanics
ABSTRACT: We contribute to the vastly growing field of machine learning for engineering
systems by demonstrating that equivariant graph neural networks have the
potential to learn more accurate dynamic-interaction models than their
non-equivariant counterparts. We benchmark two well-studied fluid flow systems,
namely the 3D decaying Taylor-Green vortex and the 3D reverse Poiseuille flow,
and compare equivariant graph neural networks to their non-equivariant
counterparts on different performance measures, such as kinetic energy or
Sinkhorn distance. Such measures are typically used in engineering to validate
numerical solvers. Our main findings are that while being rather slow to train
and evaluate, equivariant models learn more physically accurate interactions.
This indicates opportunities for future work towards coarse-grained models for
turbulent flows, and generalization across system dynamics and parameters.
| no_new_dataset | 0.949153 |
2304.00152 | Liyan Chen | Liyan Chen, Weihan Wang, Philippos Mordohai | Learning the Distribution of Errors in Stereo Matching for Joint
Disparity and Uncertainty Estimation | CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a new loss function for joint disparity and uncertainty estimation
in deep stereo matching. Our work is motivated by the need for precise
uncertainty estimates and the observation that multi-task learning often leads
to improved performance in all tasks. We show that this can be achieved by
requiring the distribution of uncertainty to match the distribution of
disparity errors via a KL divergence term in the network's loss function. A
differentiable soft-histogramming technique is used to approximate the
distributions so that they can be used in the loss. We experimentally assess
the effectiveness of our approach and observe significant improvements in both
disparity and uncertainty prediction on large datasets.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 21:58:19 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Chen",
"Liyan",
""
],
[
"Wang",
"Weihan",
""
],
[
"Mordohai",
"Philippos",
""
]
] | TITLE: Learning the Distribution of Errors in Stereo Matching for Joint
Disparity and Uncertainty Estimation
ABSTRACT: We present a new loss function for joint disparity and uncertainty estimation
in deep stereo matching. Our work is motivated by the need for precise
uncertainty estimates and the observation that multi-task learning often leads
to improved performance in all tasks. We show that this can be achieved by
requiring the distribution of uncertainty to match the distribution of
disparity errors via a KL divergence term in the network's loss function. A
differentiable soft-histogramming technique is used to approximate the
distributions so that they can be used in the loss. We experimentally assess
the effectiveness of our approach and observe significant improvements in both
disparity and uncertainty prediction on large datasets.
| no_new_dataset | 0.947575 |
2304.00177 | Lhuqita Fazry | Lhuqita Fazry, Asep Haryono, Nuzulul Khairu Nissa, Sunarno, Naufal
Muhammad Hirzi, Muhammad Febrian Rachmadi, Wisnu Jatmiko | Hierarchical Vision Transformers for Cardiac Ejection Fraction
Estimation | null | IWBIS 2022 International Workshop on Big Data and Information
Security (IWBIS) 2022 | 10.1109/IWBIS56557.2022.9924664 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The left ventricular of ejection fraction is one of the most important metric
of cardiac function. It is used by cardiologist to identify patients who are
eligible for lifeprolonging therapies. However, the assessment of ejection
fraction suffers from inter-observer variability. To overcome this challenge,
we propose a deep learning approach, based on hierarchical vision Transformers,
to estimate the ejection fraction from echocardiogram videos. The proposed
method can estimate ejection fraction without the need for left ventrice
segmentation first, make it more efficient than other methods. We evaluated our
method on EchoNet-Dynamic dataset resulting 5.59, 7.59 and 0.59 for MAE, RMSE
and R2 respectivelly. This results are better compared to the state-of-the-art
method, Ultrasound Video Transformer (UVT). The source code is available on
https://github.com/lhfazry/UltraSwin.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 23:42:17 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Fazry",
"Lhuqita",
""
],
[
"Haryono",
"Asep",
""
],
[
"Nissa",
"Nuzulul Khairu",
""
],
[
"Sunarno",
"",
""
],
[
"Hirzi",
"Naufal Muhammad",
""
],
[
"Rachmadi",
"Muhammad Febrian",
""
],
[
"Jatmiko",
"Wisnu",
""
]
] | TITLE: Hierarchical Vision Transformers for Cardiac Ejection Fraction
Estimation
ABSTRACT: The left ventricular of ejection fraction is one of the most important metric
of cardiac function. It is used by cardiologist to identify patients who are
eligible for lifeprolonging therapies. However, the assessment of ejection
fraction suffers from inter-observer variability. To overcome this challenge,
we propose a deep learning approach, based on hierarchical vision Transformers,
to estimate the ejection fraction from echocardiogram videos. The proposed
method can estimate ejection fraction without the need for left ventrice
segmentation first, make it more efficient than other methods. We evaluated our
method on EchoNet-Dynamic dataset resulting 5.59, 7.59 and 0.59 for MAE, RMSE
and R2 respectivelly. This results are better compared to the state-of-the-art
method, Ultrasound Video Transformer (UVT). The source code is available on
https://github.com/lhfazry/UltraSwin.
| no_new_dataset | 0.947624 |
2304.00180 | Zihao Wang | Zihao Wang, Eugene Agichtein and Jinho Choi | FCC: Fusing Conversation History and Candidate Provenance for Contextual
Response Ranking in Dialogue Systems | The 13th International Workshop on Spoken Dialogue Systems Technology | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Response ranking in dialogues plays a crucial role in retrieval-based
conversational systems. In a multi-turn dialogue, to capture the gist of a
conversation, contextual information serves as essential knowledge to achieve
this goal. In this paper, we present a flexible neural framework that can
integrate contextual information from multiple channels. Specifically for the
current task, our approach is to provide two information channels in parallel,
Fusing Conversation history and domain knowledge extracted from Candidate
provenance (FCC), where candidate responses are curated, as contextual
information to improve the performance of multi-turn dialogue response ranking.
The proposed approach can be generalized as a module to incorporate
miscellaneous contextual features for other context-oriented tasks. We evaluate
our model on the MSDialog dataset widely used for evaluating conversational
response ranking tasks. Our experimental results show that our framework
significantly outperforms the previous state-of-the-art models, improving
Recall@1 by 7% and MAP by 4%. Furthermore, we conduct ablation studies to
evaluate the contributions of each information channel, and of the framework
components, to the overall ranking performance, providing additional insights
and directions for further improvements.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 23:58:28 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wang",
"Zihao",
""
],
[
"Agichtein",
"Eugene",
""
],
[
"Choi",
"Jinho",
""
]
] | TITLE: FCC: Fusing Conversation History and Candidate Provenance for Contextual
Response Ranking in Dialogue Systems
ABSTRACT: Response ranking in dialogues plays a crucial role in retrieval-based
conversational systems. In a multi-turn dialogue, to capture the gist of a
conversation, contextual information serves as essential knowledge to achieve
this goal. In this paper, we present a flexible neural framework that can
integrate contextual information from multiple channels. Specifically for the
current task, our approach is to provide two information channels in parallel,
Fusing Conversation history and domain knowledge extracted from Candidate
provenance (FCC), where candidate responses are curated, as contextual
information to improve the performance of multi-turn dialogue response ranking.
The proposed approach can be generalized as a module to incorporate
miscellaneous contextual features for other context-oriented tasks. We evaluate
our model on the MSDialog dataset widely used for evaluating conversational
response ranking tasks. Our experimental results show that our framework
significantly outperforms the previous state-of-the-art models, improving
Recall@1 by 7% and MAP by 4%. Furthermore, we conduct ablation studies to
evaluate the contributions of each information channel, and of the framework
components, to the overall ranking performance, providing additional insights
and directions for further improvements.
| no_new_dataset | 0.950088 |
2304.00198 | Debdipta Goswami | Debdipta Goswami | Sequential Learning from Noisy Data: Data-Assimilation Meets Echo-State
Network | 7 pages, 9 figures, 1 table. arXiv admin note: text overlap with
arXiv:2211.05992 | null | null | null | eess.SY cs.LG cs.SY | http://creativecommons.org/licenses/by/4.0/ | This paper explores the problem of training a recurrent neural network from
noisy data. While neural network based dynamic predictors perform well with
noise-free training data, prediction with noisy inputs during training phase
poses a significant challenge. Here a sequential training algorithm is
developed for an echo-state network (ESN) by incorporating noisy observations
using an ensemble Kalman filter. The resultant Kalman-trained echo-state
network (KalT-ESN) outperforms the traditionally trained ESN with least square
algorithm while still being computationally cheap. The proposed method is
demonstrated on noisy observations from three systems: two synthetic datasets
from chaotic dynamical systems and a set of real-time traffic data.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 02:03:08 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Goswami",
"Debdipta",
""
]
] | TITLE: Sequential Learning from Noisy Data: Data-Assimilation Meets Echo-State
Network
ABSTRACT: This paper explores the problem of training a recurrent neural network from
noisy data. While neural network based dynamic predictors perform well with
noise-free training data, prediction with noisy inputs during training phase
poses a significant challenge. Here a sequential training algorithm is
developed for an echo-state network (ESN) by incorporating noisy observations
using an ensemble Kalman filter. The resultant Kalman-trained echo-state
network (KalT-ESN) outperforms the traditionally trained ESN with least square
algorithm while still being computationally cheap. The proposed method is
demonstrated on noisy observations from three systems: two synthetic datasets
from chaotic dynamical systems and a set of real-time traffic data.
| no_new_dataset | 0.951051 |
2304.00221 | Mang Tik Chiu | Mang Tik Chiu, Xuaner Zhang, Zijun Wei, Yuqian Zhou, Eli Shechtman,
Connelly Barnes, Zhe Lin, Florian Kainz, Sohrab Amirghodsi, Humphrey Shi | Automatic High Resolution Wire Segmentation and Removal | https://github.com/adobe-research/auto-wire-removal | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Wires and powerlines are common visual distractions that often undermine the
aesthetics of photographs. The manual process of precisely segmenting and
removing them is extremely tedious and may take up hours, especially on
high-resolution photos where wires may span the entire space. In this paper, we
present an automatic wire clean-up system that eases the process of wire
segmentation and removal/inpainting to within a few seconds. We observe several
unique challenges: wires are thin, lengthy, and sparse. These are rare
properties of subjects that common segmentation tasks cannot handle, especially
in high-resolution images. We thus propose a two-stage method that leverages
both global and local contexts to accurately segment wires in high-resolution
images efficiently, and a tile-based inpainting strategy to remove the wires
given our predicted segmentation masks. We also introduce the first wire
segmentation benchmark dataset, WireSegHR. Finally, we demonstrate
quantitatively and qualitatively that our wire clean-up system enables fully
automated wire removal with great generalization to various wire appearances.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 04:24:12 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Chiu",
"Mang Tik",
""
],
[
"Zhang",
"Xuaner",
""
],
[
"Wei",
"Zijun",
""
],
[
"Zhou",
"Yuqian",
""
],
[
"Shechtman",
"Eli",
""
],
[
"Barnes",
"Connelly",
""
],
[
"Lin",
"Zhe",
""
],
[
"Kainz",
"Florian",
""
],
[
"Amirghodsi",
"Sohrab",
""
],
[
"Shi",
"Humphrey",
""
]
] | TITLE: Automatic High Resolution Wire Segmentation and Removal
ABSTRACT: Wires and powerlines are common visual distractions that often undermine the
aesthetics of photographs. The manual process of precisely segmenting and
removing them is extremely tedious and may take up hours, especially on
high-resolution photos where wires may span the entire space. In this paper, we
present an automatic wire clean-up system that eases the process of wire
segmentation and removal/inpainting to within a few seconds. We observe several
unique challenges: wires are thin, lengthy, and sparse. These are rare
properties of subjects that common segmentation tasks cannot handle, especially
in high-resolution images. We thus propose a two-stage method that leverages
both global and local contexts to accurately segment wires in high-resolution
images efficiently, and a tile-based inpainting strategy to remove the wires
given our predicted segmentation masks. We also introduce the first wire
segmentation benchmark dataset, WireSegHR. Finally, we demonstrate
quantitatively and qualitatively that our wire clean-up system enables fully
automated wire removal with great generalization to various wire appearances.
| new_dataset | 0.960435 |
2304.00235 | Suman Adhya | Suman Adhya, Debarshi Kumar Sanyal | What Does the Indian Parliament Discuss? An Exploratory Analysis of the
Question Hour in the Lok Sabha | Accepted at the workshop PoliticalNLP co-located with the conference
LREC 2022 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The TCPD-IPD dataset is a collection of questions and answers discussed in
the Lower House of the Parliament of India during the Question Hour between
1999 and 2019. Although it is difficult to analyze such a huge collection
manually, modern text analysis tools can provide a powerful means to navigate
it. In this paper, we perform an exploratory analysis of the dataset. In
particular, we present insightful corpus-level statistics and a detailed
analysis of three subsets of the dataset. In the latter analysis, the focus is
on understanding the temporal evolution of topics using a dynamic topic model.
We observe that the parliamentary conversation indeed mirrors the political and
socio-economic tensions of each period.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 05:43:22 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Adhya",
"Suman",
""
],
[
"Sanyal",
"Debarshi Kumar",
""
]
] | TITLE: What Does the Indian Parliament Discuss? An Exploratory Analysis of the
Question Hour in the Lok Sabha
ABSTRACT: The TCPD-IPD dataset is a collection of questions and answers discussed in
the Lower House of the Parliament of India during the Question Hour between
1999 and 2019. Although it is difficult to analyze such a huge collection
manually, modern text analysis tools can provide a powerful means to navigate
it. In this paper, we perform an exploratory analysis of the dataset. In
particular, we present insightful corpus-level statistics and a detailed
analysis of three subsets of the dataset. In the latter analysis, the focus is
on understanding the temporal evolution of topics using a dynamic topic model.
We observe that the parliamentary conversation indeed mirrors the political and
socio-economic tensions of each period.
| new_dataset | 0.741019 |
2304.00241 | Yankai Chen | Yankai Chen, Yixiang Fang, Yifei Zhang, Irwin King | Bipartite Graph Convolutional Hashing for Effective and Efficient Top-N
Search in Hamming Space | Accepted by WWW 2023 | null | 10.1145/3543507.3583219 | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Searching on bipartite graphs is basal and versatile to many real-world Web
applications, e.g., online recommendation, database retrieval, and
query-document searching. Given a query node, the conventional approaches rely
on the similarity matching with the vectorized node embeddings in the
continuous Euclidean space. To efficiently manage intensive similarity
computation, developing hashing techniques for graph structured data has
recently become an emerging research direction. Despite the retrieval
efficiency in Hamming space, prior work is however confronted with catastrophic
performance decay. In this work, we investigate the problem of hashing with
Graph Convolutional Network on bipartite graphs for effective Top-N search. We
propose an end-to-end Bipartite Graph Convolutional Hashing approach, namely
BGCH, which consists of three novel and effective modules: (1) adaptive graph
convolutional hashing, (2) latent feature dispersion, and (3) Fourier
serialized gradient estimation. Specifically, the former two modules achieve
the substantial retention of the structural information against the inevitable
information loss in hash encoding; the last module develops Fourier Series
decomposition to the hashing function in the frequency domain mainly for more
accurate gradient estimation. The extensive experiments on six real-world
datasets not only show the performance superiority over the competing
hashing-based counterparts, but also demonstrate the effectiveness of all
proposed model components contained therein.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 06:39:33 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Chen",
"Yankai",
""
],
[
"Fang",
"Yixiang",
""
],
[
"Zhang",
"Yifei",
""
],
[
"King",
"Irwin",
""
]
] | TITLE: Bipartite Graph Convolutional Hashing for Effective and Efficient Top-N
Search in Hamming Space
ABSTRACT: Searching on bipartite graphs is basal and versatile to many real-world Web
applications, e.g., online recommendation, database retrieval, and
query-document searching. Given a query node, the conventional approaches rely
on the similarity matching with the vectorized node embeddings in the
continuous Euclidean space. To efficiently manage intensive similarity
computation, developing hashing techniques for graph structured data has
recently become an emerging research direction. Despite the retrieval
efficiency in Hamming space, prior work is however confronted with catastrophic
performance decay. In this work, we investigate the problem of hashing with
Graph Convolutional Network on bipartite graphs for effective Top-N search. We
propose an end-to-end Bipartite Graph Convolutional Hashing approach, namely
BGCH, which consists of three novel and effective modules: (1) adaptive graph
convolutional hashing, (2) latent feature dispersion, and (3) Fourier
serialized gradient estimation. Specifically, the former two modules achieve
the substantial retention of the structural information against the inevitable
information loss in hash encoding; the last module develops Fourier Series
decomposition to the hashing function in the frequency domain mainly for more
accurate gradient estimation. The extensive experiments on six real-world
datasets not only show the performance superiority over the competing
hashing-based counterparts, but also demonstrate the effectiveness of all
proposed model components contained therein.
| no_new_dataset | 0.946843 |
2304.00249 | Leila Ismail Prof. | Leila Ismail, Huned Materwala | From Conception to Deployment: Intelligent Stroke Prediction Framework
using Machine Learning and Performance Evaluation | Proceedings of IEEE International Conference on Omni-layer
Intelligent Systems, 2022 | null | 10.1109/COINS54846.2022.9854961 | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Stroke is the second leading cause of death worldwide. Machine learning
classification algorithms have been widely adopted for stroke prediction.
However, these algorithms were evaluated using different datasets and
evaluation metrics. Moreover, there is no comprehensive framework for stroke
data analytics. This paper proposes an intelligent stroke prediction framework
based on a critical examination of machine learning prediction algorithms in
the literature. The five most used machine learning algorithms for stroke
prediction are evaluated using a unified setup for objective comparison.
Comparative analysis and numerical results reveal that the Random Forest
algorithm is best suited for stroke prediction.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 07:11:31 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ismail",
"Leila",
""
],
[
"Materwala",
"Huned",
""
]
] | TITLE: From Conception to Deployment: Intelligent Stroke Prediction Framework
using Machine Learning and Performance Evaluation
ABSTRACT: Stroke is the second leading cause of death worldwide. Machine learning
classification algorithms have been widely adopted for stroke prediction.
However, these algorithms were evaluated using different datasets and
evaluation metrics. Moreover, there is no comprehensive framework for stroke
data analytics. This paper proposes an intelligent stroke prediction framework
based on a critical examination of machine learning prediction algorithms in
the literature. The five most used machine learning algorithms for stroke
prediction are evaluated using a unified setup for objective comparison.
Comparative analysis and numerical results reveal that the Random Forest
algorithm is best suited for stroke prediction.
| no_new_dataset | 0.950365 |
2304.00253 | Sheng Xu | Sheng Xu, Yanjing Li, Mingbao Lin, Peng Gao, Guodong Guo, Jinhu Lu,
Baochang Zhang | Q-DETR: An Efficient Low-Bit Quantized Detection Transformer | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The recent detection transformer (DETR) has advanced object detection, but
its application on resource-constrained devices requires massive computation
and memory resources. Quantization stands out as a solution by representing the
network in low-bit parameters and operations. However, there is a significant
performance drop when performing low-bit quantized DETR (Q-DETR) with existing
quantization methods. We find that the bottlenecks of Q-DETR come from the
query information distortion through our empirical analyses. This paper
addresses this problem based on a distribution rectification distillation
(DRD). We formulate our DRD as a bi-level optimization problem, which can be
derived by generalizing the information bottleneck (IB) principle to the
learning of Q-DETR. At the inner level, we conduct a distribution alignment for
the queries to maximize the self-information entropy. At the upper level, we
introduce a new foreground-aware query matching scheme to effectively transfer
the teacher information to distillation-desired features to minimize the
conditional information entropy. Extensive experimental results show that our
method performs much better than prior arts. For example, the 4-bit Q-DETR can
theoretically accelerate DETR with ResNet-50 backbone by 6.6x and achieve 39.4%
AP, with only 2.6% performance gaps than its real-valued counterpart on the
COCO dataset.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 08:05:14 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Xu",
"Sheng",
""
],
[
"Li",
"Yanjing",
""
],
[
"Lin",
"Mingbao",
""
],
[
"Gao",
"Peng",
""
],
[
"Guo",
"Guodong",
""
],
[
"Lu",
"Jinhu",
""
],
[
"Zhang",
"Baochang",
""
]
] | TITLE: Q-DETR: An Efficient Low-Bit Quantized Detection Transformer
ABSTRACT: The recent detection transformer (DETR) has advanced object detection, but
its application on resource-constrained devices requires massive computation
and memory resources. Quantization stands out as a solution by representing the
network in low-bit parameters and operations. However, there is a significant
performance drop when performing low-bit quantized DETR (Q-DETR) with existing
quantization methods. We find that the bottlenecks of Q-DETR come from the
query information distortion through our empirical analyses. This paper
addresses this problem based on a distribution rectification distillation
(DRD). We formulate our DRD as a bi-level optimization problem, which can be
derived by generalizing the information bottleneck (IB) principle to the
learning of Q-DETR. At the inner level, we conduct a distribution alignment for
the queries to maximize the self-information entropy. At the upper level, we
introduce a new foreground-aware query matching scheme to effectively transfer
the teacher information to distillation-desired features to minimize the
conditional information entropy. Extensive experimental results show that our
method performs much better than prior arts. For example, the 4-bit Q-DETR can
theoretically accelerate DETR with ResNet-50 backbone by 6.6x and achieve 39.4%
AP, with only 2.6% performance gaps than its real-valued counterpart on the
COCO dataset.
| no_new_dataset | 0.941761 |
2304.00280 | Lin Geng Foo | Jianhong Pan, Siyuan Yang, Lin Geng Foo, Qiuhong Ke, Hossein Rahmani,
Zhipeng Fan, Jun Liu | Progressive Channel-Shrinking Network | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Currently, salience-based channel pruning makes continuous breakthroughs in
network compression. In the realization, the salience mechanism is used as a
metric of channel salience to guide pruning. Therefore, salience-based channel
pruning can dynamically adjust the channel width at run-time, which provides a
flexible pruning scheme. However, there are two problems emerging: a gating
function is often needed to truncate the specific salience entries to zero,
which destabilizes the forward propagation; dynamic architecture brings more
cost for indexing in inference which bottlenecks the inference speed. In this
paper, we propose a Progressive Channel-Shrinking (PCS) method to compress the
selected salience entries at run-time instead of roughly approximating them to
zero. We also propose a Running Shrinking Policy to provide a testing-static
pruning scheme that can reduce the memory access cost for filter indexing. We
evaluate our method on ImageNet and CIFAR10 datasets over two prevalent
networks: ResNet and VGG, and demonstrate that our PCS outperforms all
baselines and achieves state-of-the-art in terms of compression-performance
tradeoff. Moreover, we observe a significant and practical acceleration of
inference.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 10:01:14 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Pan",
"Jianhong",
""
],
[
"Yang",
"Siyuan",
""
],
[
"Foo",
"Lin Geng",
""
],
[
"Ke",
"Qiuhong",
""
],
[
"Rahmani",
"Hossein",
""
],
[
"Fan",
"Zhipeng",
""
],
[
"Liu",
"Jun",
""
]
] | TITLE: Progressive Channel-Shrinking Network
ABSTRACT: Currently, salience-based channel pruning makes continuous breakthroughs in
network compression. In the realization, the salience mechanism is used as a
metric of channel salience to guide pruning. Therefore, salience-based channel
pruning can dynamically adjust the channel width at run-time, which provides a
flexible pruning scheme. However, there are two problems emerging: a gating
function is often needed to truncate the specific salience entries to zero,
which destabilizes the forward propagation; dynamic architecture brings more
cost for indexing in inference which bottlenecks the inference speed. In this
paper, we propose a Progressive Channel-Shrinking (PCS) method to compress the
selected salience entries at run-time instead of roughly approximating them to
zero. We also propose a Running Shrinking Policy to provide a testing-static
pruning scheme that can reduce the memory access cost for filter indexing. We
evaluate our method on ImageNet and CIFAR10 datasets over two prevalent
networks: ResNet and VGG, and demonstrate that our PCS outperforms all
baselines and achieves state-of-the-art in terms of compression-performance
tradeoff. Moreover, we observe a significant and practical acceleration of
inference.
| no_new_dataset | 0.946843 |
2304.00289 | Xi Li | Xi Li, Youcai Xu and Minfu Feng | A Pressure-Stabilized Continuous Data Assimilation Reduced Order Model | null | null | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a novel reduced-order pressure stabilization strategy based on
continuous data assimilation(CDA) for two-dimensional incompressible
Navier-Stokes equations. A feedback control term is incorporated into
pressure-correction projection method to derive the Galerkin projection-based
CDA proper orthogonal decomposition reduced order model(POD-ROM) that uses
pressure modes as well as velocity's simultaneously to compute the
reduced-order solutions. The greatest advantage over this ROM is circumventing
the standard discrete inf-sup condition for the mixed POD velocity-pressure
spaces with the help of CDA which also guarantees the high accuracy of
reduced-order solutions; moreover, the classical projection method decouples
reduced-order velocity and pressure, which further enhances computational
efficiency. Unconditional stability and convergence over POD modes(up to
discretization error) are presented, and a benchmark test is performed to
validate the theoretical results.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 10:54:25 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Li",
"Xi",
""
],
[
"Xu",
"Youcai",
""
],
[
"Feng",
"Minfu",
""
]
] | TITLE: A Pressure-Stabilized Continuous Data Assimilation Reduced Order Model
ABSTRACT: We present a novel reduced-order pressure stabilization strategy based on
continuous data assimilation(CDA) for two-dimensional incompressible
Navier-Stokes equations. A feedback control term is incorporated into
pressure-correction projection method to derive the Galerkin projection-based
CDA proper orthogonal decomposition reduced order model(POD-ROM) that uses
pressure modes as well as velocity's simultaneously to compute the
reduced-order solutions. The greatest advantage over this ROM is circumventing
the standard discrete inf-sup condition for the mixed POD velocity-pressure
spaces with the help of CDA which also guarantees the high accuracy of
reduced-order solutions; moreover, the classical projection method decouples
reduced-order velocity and pressure, which further enhances computational
efficiency. Unconditional stability and convergence over POD modes(up to
discretization error) are presented, and a benchmark test is performed to
validate the theoretical results.
| no_new_dataset | 0.948058 |
2304.00295 | Rui Sun | Rui Sun, Fengwei Zhou, Zhenhua Dong, Chuanlong Xie, Lanqing Hong,
Jiawei Li, Rui Zhang, Zhen Li, Zhenguo Li | Fair-CDA: Continuous and Directional Augmentation for Group Fairness | Accepted by AAAI-23 | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In this work, we propose {\it Fair-CDA}, a fine-grained data augmentation
strategy for imposing fairness constraints. We use a feature disentanglement
method to extract the features highly related to the sensitive attributes. Then
we show that group fairness can be achieved by regularizing the models on
transition paths of sensitive features between groups. By adjusting the
perturbation strength in the direction of the paths, our proposed augmentation
is controllable and auditable. To alleviate the accuracy degradation caused by
fairness constraints, we further introduce a calibrated model to impute labels
for the augmented data. Our proposed method does not assume any data generative
model and ensures good generalization for both accuracy and fairness.
Experimental results show that Fair-CDA consistently outperforms
state-of-the-art methods on widely-used benchmarks, e.g., Adult, CelebA and
MovieLens. Especially, Fair-CDA obtains an 86.3\% relative improvement for
fairness while maintaining the accuracy on the Adult dataset. Moreover, we
evaluate Fair-CDA in an online recommendation system to demonstrate the
effectiveness of our method in terms of accuracy and fairness.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 11:23:00 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Sun",
"Rui",
""
],
[
"Zhou",
"Fengwei",
""
],
[
"Dong",
"Zhenhua",
""
],
[
"Xie",
"Chuanlong",
""
],
[
"Hong",
"Lanqing",
""
],
[
"Li",
"Jiawei",
""
],
[
"Zhang",
"Rui",
""
],
[
"Li",
"Zhen",
""
],
[
"Li",
"Zhenguo",
""
]
] | TITLE: Fair-CDA: Continuous and Directional Augmentation for Group Fairness
ABSTRACT: In this work, we propose {\it Fair-CDA}, a fine-grained data augmentation
strategy for imposing fairness constraints. We use a feature disentanglement
method to extract the features highly related to the sensitive attributes. Then
we show that group fairness can be achieved by regularizing the models on
transition paths of sensitive features between groups. By adjusting the
perturbation strength in the direction of the paths, our proposed augmentation
is controllable and auditable. To alleviate the accuracy degradation caused by
fairness constraints, we further introduce a calibrated model to impute labels
for the augmented data. Our proposed method does not assume any data generative
model and ensures good generalization for both accuracy and fairness.
Experimental results show that Fair-CDA consistently outperforms
state-of-the-art methods on widely-used benchmarks, e.g., Adult, CelebA and
MovieLens. Especially, Fair-CDA obtains an 86.3\% relative improvement for
fairness while maintaining the accuracy on the Adult dataset. Moreover, we
evaluate Fair-CDA in an online recommendation system to demonstrate the
effectiveness of our method in terms of accuracy and fairness.
| no_new_dataset | 0.941868 |
2304.00338 | Phan Nguyen | Brian R. Bartoldson, Yeping Hu, Amar Saini, Jose Cadena, Yucheng Fu,
Jie Bao, Zhijie Xu, Brenda Ng, Phan Nguyen | Scientific Computing Algorithms to Learn Enhanced Scalable Surrogates
for Mesh Physics | ICLR 2023 Workshop on Physics for Machine Learning | null | null | null | cs.LG cs.NA math.NA | http://creativecommons.org/licenses/by/4.0/ | Data-driven modeling approaches can produce fast surrogates to study
large-scale physics problems. Among them, graph neural networks (GNNs) that
operate on mesh-based data are desirable because they possess inductive biases
that promote physical faithfulness, but hardware limitations have precluded
their application to large computational domains. We show that it is
\textit{possible} to train a class of GNN surrogates on 3D meshes. We scale
MeshGraphNets (MGN), a subclass of GNNs for mesh-based physics modeling, via
our domain decomposition approach to facilitate training that is mathematically
equivalent to training on the whole domain under certain conditions. With this,
we were able to train MGN on meshes with \textit{millions} of nodes to generate
computational fluid dynamics (CFD) simulations. Furthermore, we show how to
enhance MGN via higher-order numerical integration, which can reduce MGN's
error and training time. We validated our methods on an accompanying dataset of
3D $\text{CO}_2$-capture CFD simulations on a 3.1M-node mesh. This work
presents a practical path to scaling MGN for real-world applications.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 15:42:18 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Bartoldson",
"Brian R.",
""
],
[
"Hu",
"Yeping",
""
],
[
"Saini",
"Amar",
""
],
[
"Cadena",
"Jose",
""
],
[
"Fu",
"Yucheng",
""
],
[
"Bao",
"Jie",
""
],
[
"Xu",
"Zhijie",
""
],
[
"Ng",
"Brenda",
""
],
[
"Nguyen",
"Phan",
""
]
] | TITLE: Scientific Computing Algorithms to Learn Enhanced Scalable Surrogates
for Mesh Physics
ABSTRACT: Data-driven modeling approaches can produce fast surrogates to study
large-scale physics problems. Among them, graph neural networks (GNNs) that
operate on mesh-based data are desirable because they possess inductive biases
that promote physical faithfulness, but hardware limitations have precluded
their application to large computational domains. We show that it is
\textit{possible} to train a class of GNN surrogates on 3D meshes. We scale
MeshGraphNets (MGN), a subclass of GNNs for mesh-based physics modeling, via
our domain decomposition approach to facilitate training that is mathematically
equivalent to training on the whole domain under certain conditions. With this,
we were able to train MGN on meshes with \textit{millions} of nodes to generate
computational fluid dynamics (CFD) simulations. Furthermore, we show how to
enhance MGN via higher-order numerical integration, which can reduce MGN's
error and training time. We validated our methods on an accompanying dataset of
3D $\text{CO}_2$-capture CFD simulations on a 3.1M-node mesh. This work
presents a practical path to scaling MGN for real-world applications.
| no_new_dataset | 0.943086 |
2304.00350 | Won Ik Cho | Won Ik Cho, Yoon Kyung Lee, Seoyeon Bae, Jihwan Kim, Sangah Park,
Moosung Kim, Sowon Hahn, Nam Soo Kim | When Crowd Meets Persona: Creating a Large-Scale Open-Domain Persona
Dialogue Corpus | Presented at HCOMP 2022 as Works-in-Progress | null | null | null | cs.CL | http://creativecommons.org/licenses/by-sa/4.0/ | Building a natural language dataset requires caution since word semantics is
vulnerable to subtle text change or the definition of the annotated concept.
Such a tendency can be seen in generative tasks like question-answering and
dialogue generation and also in tasks that create a categorization-based
corpus, like topic classification or sentiment analysis. Open-domain
conversations involve two or more crowdworkers freely conversing about any
topic, and collecting such data is particularly difficult for two reasons: 1)
the dataset should be ``crafted" rather than ``obtained" due to privacy
concerns, and 2) paid creation of such dialogues may differ from how
crowdworkers behave in real-world settings. In this study, we tackle these
issues when creating a large-scale open-domain persona dialogue corpus, where
persona implies that the conversation is performed by several actors with a
fixed persona and user-side workers from an unspecified crowd.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 16:10:36 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Cho",
"Won Ik",
""
],
[
"Lee",
"Yoon Kyung",
""
],
[
"Bae",
"Seoyeon",
""
],
[
"Kim",
"Jihwan",
""
],
[
"Park",
"Sangah",
""
],
[
"Kim",
"Moosung",
""
],
[
"Hahn",
"Sowon",
""
],
[
"Kim",
"Nam Soo",
""
]
] | TITLE: When Crowd Meets Persona: Creating a Large-Scale Open-Domain Persona
Dialogue Corpus
ABSTRACT: Building a natural language dataset requires caution since word semantics is
vulnerable to subtle text change or the definition of the annotated concept.
Such a tendency can be seen in generative tasks like question-answering and
dialogue generation and also in tasks that create a categorization-based
corpus, like topic classification or sentiment analysis. Open-domain
conversations involve two or more crowdworkers freely conversing about any
topic, and collecting such data is particularly difficult for two reasons: 1)
the dataset should be ``crafted" rather than ``obtained" due to privacy
concerns, and 2) paid creation of such dialogues may differ from how
crowdworkers behave in real-world settings. In this study, we tackle these
issues when creating a large-scale open-domain persona dialogue corpus, where
persona implies that the conversation is performed by several actors with a
fixed persona and user-side workers from an unspecified crowd.
| no_new_dataset | 0.612678 |
2304.00359 | Yukang Cao | Yukang Cao, Kai Han, Kwan-Yee K. Wong | SeSDF: Self-evolved Signed Distance Field for Implicit 3D Clothed Human
Reconstruction | 25 pages, 21 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We address the problem of clothed human reconstruction from a single image or
uncalibrated multi-view images. Existing methods struggle with reconstructing
detailed geometry of a clothed human and often require a calibrated setting for
multi-view reconstruction. We propose a flexible framework which, by leveraging
the parametric SMPL-X model, can take an arbitrary number of input images to
reconstruct a clothed human model under an uncalibrated setting. At the core of
our framework is our novel self-evolved signed distance field (SeSDF) module
which allows the framework to learn to deform the signed distance field (SDF)
derived from the fitted SMPL-X model, such that detailed geometry reflecting
the actual clothed human can be encoded for better reconstruction. Besides, we
propose a simple method for self-calibration of multi-view images via the
fitted SMPL-X parameters. This lifts the requirement of tedious manual
calibration and largely increases the flexibility of our method. Further, we
introduce an effective occlusion-aware feature fusion strategy to account for
the most useful features to reconstruct the human model. We thoroughly evaluate
our framework on public benchmarks, demonstrating significant superiority over
the state-of-the-arts both qualitatively and quantitatively.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 16:58:19 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Cao",
"Yukang",
""
],
[
"Han",
"Kai",
""
],
[
"Wong",
"Kwan-Yee K.",
""
]
] | TITLE: SeSDF: Self-evolved Signed Distance Field for Implicit 3D Clothed Human
Reconstruction
ABSTRACT: We address the problem of clothed human reconstruction from a single image or
uncalibrated multi-view images. Existing methods struggle with reconstructing
detailed geometry of a clothed human and often require a calibrated setting for
multi-view reconstruction. We propose a flexible framework which, by leveraging
the parametric SMPL-X model, can take an arbitrary number of input images to
reconstruct a clothed human model under an uncalibrated setting. At the core of
our framework is our novel self-evolved signed distance field (SeSDF) module
which allows the framework to learn to deform the signed distance field (SDF)
derived from the fitted SMPL-X model, such that detailed geometry reflecting
the actual clothed human can be encoded for better reconstruction. Besides, we
propose a simple method for self-calibration of multi-view images via the
fitted SMPL-X parameters. This lifts the requirement of tedious manual
calibration and largely increases the flexibility of our method. Further, we
introduce an effective occlusion-aware feature fusion strategy to account for
the most useful features to reconstruct the human model. We thoroughly evaluate
our framework on public benchmarks, demonstrating significant superiority over
the state-of-the-arts both qualitatively and quantitatively.
| no_new_dataset | 0.943919 |
2304.00378 | Xiou Ge | Xiou Ge, Yun-Cheng Wang, Bin Wang, C.-C. Jay Kuo | Knowledge Graph Embedding with 3D Compound Geometric Transformations | null | null | null | null | cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | The cascade of 2D geometric transformations were exploited to model relations
between entities in a knowledge graph (KG), leading to an effective KG
embedding (KGE) model, CompoundE. Furthermore, the rotation in the 3D space was
proposed as a new KGE model, Rotate3D, by leveraging its non-commutative
property. Inspired by CompoundE and Rotate3D, we leverage 3D compound geometric
transformations, including translation, rotation, scaling, reflection, and
shear and propose a family of KGE models, named CompoundE3D, in this work.
CompoundE3D allows multiple design variants to match rich underlying
characteristics of a KG. Since each variant has its own advantages on a subset
of relations, an ensemble of multiple variants can yield superior performance.
The effectiveness and flexibility of CompoundE3D are experimentally verified on
four popular link prediction datasets.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 19:56:51 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ge",
"Xiou",
""
],
[
"Wang",
"Yun-Cheng",
""
],
[
"Wang",
"Bin",
""
],
[
"Kuo",
"C. -C. Jay",
""
]
] | TITLE: Knowledge Graph Embedding with 3D Compound Geometric Transformations
ABSTRACT: The cascade of 2D geometric transformations were exploited to model relations
between entities in a knowledge graph (KG), leading to an effective KG
embedding (KGE) model, CompoundE. Furthermore, the rotation in the 3D space was
proposed as a new KGE model, Rotate3D, by leveraging its non-commutative
property. Inspired by CompoundE and Rotate3D, we leverage 3D compound geometric
transformations, including translation, rotation, scaling, reflection, and
shear and propose a family of KGE models, named CompoundE3D, in this work.
CompoundE3D allows multiple design variants to match rich underlying
characteristics of a KG. Since each variant has its own advantages on a subset
of relations, an ensemble of multiple variants can yield superior performance.
The effectiveness and flexibility of CompoundE3D are experimentally verified on
four popular link prediction datasets.
| no_new_dataset | 0.948822 |
2304.00387 | Anshul Shah | Anshul Shah, Aniket Roy, Ketul Shah, Shlok Kumar Mishra, David Jacobs,
Anoop Cherian, Rama Chellappa | HaLP: Hallucinating Latent Positives for Skeleton-based Self-Supervised
Learning of Actions | To be presented at CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Supervised learning of skeleton sequence encoders for action recognition has
received significant attention in recent times. However, learning such encoders
without labels continues to be a challenging problem. While prior works have
shown promising results by applying contrastive learning to pose sequences, the
quality of the learned representations is often observed to be closely tied to
data augmentations that are used to craft the positives. However, augmenting
pose sequences is a difficult task as the geometric constraints among the
skeleton joints need to be enforced to make the augmentations realistic for
that action. In this work, we propose a new contrastive learning approach to
train models for skeleton-based action recognition without labels. Our key
contribution is a simple module, HaLP - to Hallucinate Latent Positives for
contrastive learning. Specifically, HaLP explores the latent space of poses in
suitable directions to generate new positives. To this end, we present a novel
optimization formulation to solve for the synthetic positives with an explicit
control on their hardness. We propose approximations to the objective, making
them solvable in closed form with minimal overhead. We show via experiments
that using these generated positives within a standard contrastive learning
framework leads to consistent improvements across benchmarks such as NTU-60,
NTU-120, and PKU-II on tasks like linear evaluation, transfer learning, and kNN
evaluation. Our code will be made available at
https://github.com/anshulbshah/HaLP.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 21:09:43 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Shah",
"Anshul",
""
],
[
"Roy",
"Aniket",
""
],
[
"Shah",
"Ketul",
""
],
[
"Mishra",
"Shlok Kumar",
""
],
[
"Jacobs",
"David",
""
],
[
"Cherian",
"Anoop",
""
],
[
"Chellappa",
"Rama",
""
]
] | TITLE: HaLP: Hallucinating Latent Positives for Skeleton-based Self-Supervised
Learning of Actions
ABSTRACT: Supervised learning of skeleton sequence encoders for action recognition has
received significant attention in recent times. However, learning such encoders
without labels continues to be a challenging problem. While prior works have
shown promising results by applying contrastive learning to pose sequences, the
quality of the learned representations is often observed to be closely tied to
data augmentations that are used to craft the positives. However, augmenting
pose sequences is a difficult task as the geometric constraints among the
skeleton joints need to be enforced to make the augmentations realistic for
that action. In this work, we propose a new contrastive learning approach to
train models for skeleton-based action recognition without labels. Our key
contribution is a simple module, HaLP - to Hallucinate Latent Positives for
contrastive learning. Specifically, HaLP explores the latent space of poses in
suitable directions to generate new positives. To this end, we present a novel
optimization formulation to solve for the synthetic positives with an explicit
control on their hardness. We propose approximations to the objective, making
them solvable in closed form with minimal overhead. We show via experiments
that using these generated positives within a standard contrastive learning
framework leads to consistent improvements across benchmarks such as NTU-60,
NTU-120, and PKU-II on tasks like linear evaluation, transfer learning, and kNN
evaluation. Our code will be made available at
https://github.com/anshulbshah/HaLP.
| no_new_dataset | 0.950915 |
2304.00394 | Donald Pinckney | Donald Pinckney, Federico Cassano, Arjun Guha, Jonathan Bell | A Large Scale Analysis of Semantic Versioning in NPM | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The NPM package repository contains over two million packages and serves tens
of billions of downloads per-week. Nearly every single JavaScript application
uses the NPM package manager to install packages from the NPM repository. NPM
relies on a "semantic versioning" ('semver') scheme to maintain a healthy
ecosystem, where bug-fixes are reliably delivered to downstream packages as
quickly as possible, while breaking changes require manual intervention by
downstream package maintainers. In order to understand how developers use
semver, we build a dataset containing every version of every package on NPM and
analyze the flow of updates throughout the ecosystem. We build a
time-travelling dependency resolver for NPM, which allows us to determine
precisely which versions of each dependency would have been resolved at
different times. We segment our analysis to allow for a direct analysis of
security-relevant updates (those that introduce or patch vulnerabilities) in
comparison to the rest of the ecosystem. We find that when developers use
semver correctly, critical updates such as security patches can flow quite
rapidly to downstream dependencies in the majority of cases (90.09%), but this
does not always occur, due to developers' imperfect use of both semver version
constraints and semver version number increments. Our findings have
implications for developers and researchers alike. We make our infrastructure
and dataset publicly available under an open source license.
| [
{
"version": "v1",
"created": "Sat, 1 Apr 2023 21:36:12 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Pinckney",
"Donald",
""
],
[
"Cassano",
"Federico",
""
],
[
"Guha",
"Arjun",
""
],
[
"Bell",
"Jonathan",
""
]
] | TITLE: A Large Scale Analysis of Semantic Versioning in NPM
ABSTRACT: The NPM package repository contains over two million packages and serves tens
of billions of downloads per-week. Nearly every single JavaScript application
uses the NPM package manager to install packages from the NPM repository. NPM
relies on a "semantic versioning" ('semver') scheme to maintain a healthy
ecosystem, where bug-fixes are reliably delivered to downstream packages as
quickly as possible, while breaking changes require manual intervention by
downstream package maintainers. In order to understand how developers use
semver, we build a dataset containing every version of every package on NPM and
analyze the flow of updates throughout the ecosystem. We build a
time-travelling dependency resolver for NPM, which allows us to determine
precisely which versions of each dependency would have been resolved at
different times. We segment our analysis to allow for a direct analysis of
security-relevant updates (those that introduce or patch vulnerabilities) in
comparison to the rest of the ecosystem. We find that when developers use
semver correctly, critical updates such as security patches can flow quite
rapidly to downstream dependencies in the majority of cases (90.09%), but this
does not always occur, due to developers' imperfect use of both semver version
constraints and semver version number increments. Our findings have
implications for developers and researchers alike. We make our infrastructure
and dataset publicly available under an open source license.
| new_dataset | 0.968471 |
2304.00419 | Gregory Schwartzman | Gregory Schwartzman | Mini-batch $k$-means terminates within $O(d/\epsilon)$ iterations | ICLR 2023 | null | null | null | cs.LG cs.AI cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We answer the question: "Does local progress (on batches) imply global
progress (on the entire dataset) for mini-batch $k$-means?". Specifically, we
consider mini-batch $k$-means which terminates only when the improvement in the
quality of the clustering on the sampled batch is below some threshold.
Although at first glance it appears that this algorithm might execute
forever, we answer the above question in the affirmative and show that if the
batch is of size $\tilde{\Omega}((d/\epsilon)^2)$, it must terminate within
$O(d/\epsilon)$ iterations with high probability, where $d$ is the dimension of
the input, and $\epsilon$ is a threshold parameter for termination. This is
true regardless of how the centers are initialized. When the algorithm is
initialized with the $k$-means++ initialization scheme, it achieves an
approximation ratio of $O(\log k)$ (the same as the full-batch version).
Finally, we show the applicability of our results to the mini-batch $k$-means
algorithm implemented in the scikit-learn (sklearn) python library.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 00:58:29 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Schwartzman",
"Gregory",
""
]
] | TITLE: Mini-batch $k$-means terminates within $O(d/\epsilon)$ iterations
ABSTRACT: We answer the question: "Does local progress (on batches) imply global
progress (on the entire dataset) for mini-batch $k$-means?". Specifically, we
consider mini-batch $k$-means which terminates only when the improvement in the
quality of the clustering on the sampled batch is below some threshold.
Although at first glance it appears that this algorithm might execute
forever, we answer the above question in the affirmative and show that if the
batch is of size $\tilde{\Omega}((d/\epsilon)^2)$, it must terminate within
$O(d/\epsilon)$ iterations with high probability, where $d$ is the dimension of
the input, and $\epsilon$ is a threshold parameter for termination. This is
true regardless of how the centers are initialized. When the algorithm is
initialized with the $k$-means++ initialization scheme, it achieves an
approximation ratio of $O(\log k)$ (the same as the full-batch version).
Finally, we show the applicability of our results to the mini-batch $k$-means
algorithm implemented in the scikit-learn (sklearn) python library.
| no_new_dataset | 0.943295 |
2304.00424 | Seokeon Choi | Seokeon Choi, Debasmit Das, Sungha Choi, Seunghan Yang, Hyunsin Park,
Sungrack Yun | Progressive Random Convolutions for Single Domain Generalization | CVPR 2023 (The combined version of the main paper and supplementary
materials) | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Single domain generalization aims to train a generalizable model with only
one source domain to perform well on arbitrary unseen target domains. Image
augmentation based on Random Convolutions (RandConv), consisting of one
convolution layer randomly initialized for each mini-batch, enables the model
to learn generalizable visual representations by distorting local textures
despite its simple and lightweight structure. However, RandConv has structural
limitations in that the generated image easily loses semantics as the kernel
size increases, and lacks the inherent diversity of a single convolution
operation. To solve the problem, we propose a Progressive Random Convolution
(Pro-RandConv) method that recursively stacks random convolution layers with a
small kernel size instead of increasing the kernel size. This progressive
approach can not only mitigate semantic distortions by reducing the influence
of pixels away from the center in the theoretical receptive field, but also
create more effective virtual domains by gradually increasing the style
diversity. In addition, we develop a basic random convolution layer into a
random convolution block including deformable offsets and affine transformation
to support texture and contrast diversification, both of which are also
randomly initialized. Without complex generators or adversarial learning, we
demonstrate that our simple yet effective augmentation strategy outperforms
state-of-the-art methods on single domain generalization benchmarks.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 01:42:51 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Choi",
"Seokeon",
""
],
[
"Das",
"Debasmit",
""
],
[
"Choi",
"Sungha",
""
],
[
"Yang",
"Seunghan",
""
],
[
"Park",
"Hyunsin",
""
],
[
"Yun",
"Sungrack",
""
]
] | TITLE: Progressive Random Convolutions for Single Domain Generalization
ABSTRACT: Single domain generalization aims to train a generalizable model with only
one source domain to perform well on arbitrary unseen target domains. Image
augmentation based on Random Convolutions (RandConv), consisting of one
convolution layer randomly initialized for each mini-batch, enables the model
to learn generalizable visual representations by distorting local textures
despite its simple and lightweight structure. However, RandConv has structural
limitations in that the generated image easily loses semantics as the kernel
size increases, and lacks the inherent diversity of a single convolution
operation. To solve the problem, we propose a Progressive Random Convolution
(Pro-RandConv) method that recursively stacks random convolution layers with a
small kernel size instead of increasing the kernel size. This progressive
approach can not only mitigate semantic distortions by reducing the influence
of pixels away from the center in the theoretical receptive field, but also
create more effective virtual domains by gradually increasing the style
diversity. In addition, we develop a basic random convolution layer into a
random convolution block including deformable offsets and affine transformation
to support texture and contrast diversification, both of which are also
randomly initialized. Without complex generators or adversarial learning, we
demonstrate that our simple yet effective augmentation strategy outperforms
state-of-the-art methods on single domain generalization benchmarks.
| no_new_dataset | 0.949389 |
2304.00426 | Yifan Zhao | Zeyin Song, Yifan Zhao, Yujun Shi, Peixi Peng, Li Yuan, Yonghong Tian | Learning with Fantasy: Semantic-Aware Virtual Contrastive Constraint for
Few-Shot Class-Incremental Learning | Accepted by CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Few-shot class-incremental learning (FSCIL) aims at learning to classify new
classes continually from limited samples without forgetting the old classes.
The mainstream framework tackling FSCIL is first to adopt the cross-entropy
(CE) loss for training at the base session, then freeze the feature extractor
to adapt to new classes. However, in this work, we find that the CE loss is not
ideal for the base session training as it suffers poor class separation in
terms of representations, which further degrades generalization to novel
classes. One tempting method to mitigate this problem is to apply an additional
naive supervised contrastive learning (SCL) in the base session. Unfortunately,
we find that although SCL can create a slightly better representation
separation among different base classes, it still struggles to separate base
classes and new classes. Inspired by the observations made, we propose
Semantic-Aware Virtual Contrastive model (SAVC), a novel method that
facilitates separation between new classes and base classes by introducing
virtual classes to SCL. These virtual classes, which are generated via
pre-defined transformations, not only act as placeholders for unseen classes in
the representation space, but also provide diverse semantic information. By
learning to recognize and contrast in the fantasy space fostered by virtual
classes, our SAVC significantly boosts base class separation and novel class
generalization, achieving new state-of-the-art performance on the three
widely-used FSCIL benchmark datasets. Code is available at:
https://github.com/zysong0113/SAVC.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 01:51:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Song",
"Zeyin",
""
],
[
"Zhao",
"Yifan",
""
],
[
"Shi",
"Yujun",
""
],
[
"Peng",
"Peixi",
""
],
[
"Yuan",
"Li",
""
],
[
"Tian",
"Yonghong",
""
]
] | TITLE: Learning with Fantasy: Semantic-Aware Virtual Contrastive Constraint for
Few-Shot Class-Incremental Learning
ABSTRACT: Few-shot class-incremental learning (FSCIL) aims at learning to classify new
classes continually from limited samples without forgetting the old classes.
The mainstream framework tackling FSCIL is first to adopt the cross-entropy
(CE) loss for training at the base session, then freeze the feature extractor
to adapt to new classes. However, in this work, we find that the CE loss is not
ideal for the base session training as it suffers poor class separation in
terms of representations, which further degrades generalization to novel
classes. One tempting method to mitigate this problem is to apply an additional
naive supervised contrastive learning (SCL) in the base session. Unfortunately,
we find that although SCL can create a slightly better representation
separation among different base classes, it still struggles to separate base
classes and new classes. Inspired by the observations made, we propose
Semantic-Aware Virtual Contrastive model (SAVC), a novel method that
facilitates separation between new classes and base classes by introducing
virtual classes to SCL. These virtual classes, which are generated via
pre-defined transformations, not only act as placeholders for unseen classes in
the representation space, but also provide diverse semantic information. By
learning to recognize and contrast in the fantasy space fostered by virtual
classes, our SAVC significantly boosts base class separation and novel class
generalization, achieving new state-of-the-art performance on the three
widely-used FSCIL benchmark datasets. Code is available at:
https://github.com/zysong0113/SAVC.
| no_new_dataset | 0.951051 |
2304.00440 | Songjie Yang | Songjie Yang, Chenfei Xie, Wanting Lyu, Boyu Ning, Zhongpei Zhang, and
Chau Yuen | Near-Field Channel Estimation for Extremely Large-Scale Reconfigurable
Intelligent Surface (XL-RIS)-Aided Wideband mmWave Systems | null | null | null | null | cs.IT eess.SP math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Near-field communications present new opportunities over near-field channels,
however, the spherical wavefront propagation makes near-field signal processing
challenging. In this context, this paper proposes efficient near-field channel
estimation methods for wideband MIMO mmWave systems with the aid of extremely
large-scale reconfigurable intelligent surfaces (XL-RIS). For the wideband
signals reflected by the analog RIS, we characterize their near-field beam
squint effect in both angle and distance domains. Based on the mathematical
analysis of the near-field beam patterns over all frequencies, a wideband
spherical-domain dictionary is constructed by minimizing the coherence of two
arbitrary beams. In light of this, we formulate a two-dimensional compressive
sensing problem to recover the channel parameter based on the spherical-domain
sparsity of mmWave channels. To this end, we present a correlation
coefficient-based atom matching method within our proposed multi-frequency
parallelizable subspace recovery framework for efficient solutions.
Additionally, we propose a two-dimensional oracle estimator as a benchmark and
derive its lower bound across all subcarriers. Our findings emphasize the
significance of system hyperparameters and the sensing matrix of each
subcarrier in determining the accuracy of the estimation. Finally, numerical
results show that our proposed method achieves considerable performance
compared with the lower bound and has a time complexity linear to the number of
RIS elements.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 03:27:51 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yang",
"Songjie",
""
],
[
"Xie",
"Chenfei",
""
],
[
"Lyu",
"Wanting",
""
],
[
"Ning",
"Boyu",
""
],
[
"Zhang",
"Zhongpei",
""
],
[
"Yuen",
"Chau",
""
]
] | TITLE: Near-Field Channel Estimation for Extremely Large-Scale Reconfigurable
Intelligent Surface (XL-RIS)-Aided Wideband mmWave Systems
ABSTRACT: Near-field communications present new opportunities over near-field channels,
however, the spherical wavefront propagation makes near-field signal processing
challenging. In this context, this paper proposes efficient near-field channel
estimation methods for wideband MIMO mmWave systems with the aid of extremely
large-scale reconfigurable intelligent surfaces (XL-RIS). For the wideband
signals reflected by the analog RIS, we characterize their near-field beam
squint effect in both angle and distance domains. Based on the mathematical
analysis of the near-field beam patterns over all frequencies, a wideband
spherical-domain dictionary is constructed by minimizing the coherence of two
arbitrary beams. In light of this, we formulate a two-dimensional compressive
sensing problem to recover the channel parameter based on the spherical-domain
sparsity of mmWave channels. To this end, we present a correlation
coefficient-based atom matching method within our proposed multi-frequency
parallelizable subspace recovery framework for efficient solutions.
Additionally, we propose a two-dimensional oracle estimator as a benchmark and
derive its lower bound across all subcarriers. Our findings emphasize the
significance of system hyperparameters and the sensing matrix of each
subcarrier in determining the accuracy of the estimation. Finally, numerical
results show that our proposed method achieves considerable performance
compared with the lower bound and has a time complexity linear to the number of
RIS elements.
| no_new_dataset | 0.949106 |
2304.00445 | Jiawei Zhang | Jiawei Zhang, Tiantian Wang, Zhixi Feng, Shuyuan Yang | AMC-Net: An Effective Network for Automatic Modulation Classification | Accepted to ICASSP 2023 (5 pages, 2 figures, 3 table) | null | null | null | eess.SP cs.LG cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatic modulation classification (AMC) is a crucial stage in the spectrum
management, signal monitoring, and control of wireless communication systems.
The accurate classification of the modulation format plays a vital role in the
subsequent decoding of the transmitted data. End-to-end deep learning methods
have been recently applied to AMC, outperforming traditional feature
engineering techniques. However, AMC still has limitations in low
signal-to-noise ratio (SNR) environments. To address the drawback, we propose a
novel AMC-Net that improves recognition by denoising the input signal in the
frequency domain while performing multi-scale and effective feature extraction.
Experiments on two representative datasets demonstrate that our model performs
better in efficiency and effectiveness than the most current methods.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 04:26:30 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Zhang",
"Jiawei",
""
],
[
"Wang",
"Tiantian",
""
],
[
"Feng",
"Zhixi",
""
],
[
"Yang",
"Shuyuan",
""
]
] | TITLE: AMC-Net: An Effective Network for Automatic Modulation Classification
ABSTRACT: Automatic modulation classification (AMC) is a crucial stage in the spectrum
management, signal monitoring, and control of wireless communication systems.
The accurate classification of the modulation format plays a vital role in the
subsequent decoding of the transmitted data. End-to-end deep learning methods
have been recently applied to AMC, outperforming traditional feature
engineering techniques. However, AMC still has limitations in low
signal-to-noise ratio (SNR) environments. To address the drawback, we propose a
novel AMC-Net that improves recognition by denoising the input signal in the
frequency domain while performing multi-scale and effective feature extraction.
Experiments on two representative datasets demonstrate that our model performs
better in efficiency and effectiveness than the most current methods.
| no_new_dataset | 0.945298 |
2304.00460 | Yibo Yan | Yibo Yan, Seth Frey, Amy Zhang, Vladimir Filkov, Likang Yin | GitHub OSS Governance File Dataset | 5 pages, 1 figure, 1 table, to be published in MSR 2023 Data and Tool
Showcase Track | null | null | null | cs.SE | http://creativecommons.org/licenses/by/4.0/ | Open-source Software (OSS) has become a valuable resource in both industry
and academia over the last few decades. Despite the innovative structures they
develop to support the projects, OSS projects and their communities have
complex needs and face risks such as getting abandoned. To manage the internal
social dynamics and community evolution, OSS developer communities have started
relying on written governance documents that assign roles and responsibilities
to different community actors. To facilitate the study of the impact and
effectiveness of formal governance documents on OSS projects and communities,
we present a longitudinal dataset of 710 GitHub-hosted OSS projects with
\path{GOVERNANCE.MD} governance files. This dataset includes all commits made
to the repository, all issues and comments created on GitHub, and all revisions
made to the governance file. We hope its availability will foster more research
interest in studying how OSS communities govern their projects and the impact
of governance files on communities.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 06:07:00 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yan",
"Yibo",
""
],
[
"Frey",
"Seth",
""
],
[
"Zhang",
"Amy",
""
],
[
"Filkov",
"Vladimir",
""
],
[
"Yin",
"Likang",
""
]
] | TITLE: GitHub OSS Governance File Dataset
ABSTRACT: Open-source Software (OSS) has become a valuable resource in both industry
and academia over the last few decades. Despite the innovative structures they
develop to support the projects, OSS projects and their communities have
complex needs and face risks such as getting abandoned. To manage the internal
social dynamics and community evolution, OSS developer communities have started
relying on written governance documents that assign roles and responsibilities
to different community actors. To facilitate the study of the impact and
effectiveness of formal governance documents on OSS projects and communities,
we present a longitudinal dataset of 710 GitHub-hosted OSS projects with
\path{GOVERNANCE.MD} governance files. This dataset includes all commits made
to the repository, all issues and comments created on GitHub, and all revisions
made to the governance file. We hope its availability will foster more research
interest in studying how OSS communities govern their projects and the impact
of governance files on communities.
| new_dataset | 0.957833 |
2304.00467 | Haiping Wang | Haiping Wang, Yuan Liu, Zhen Dong, Yulan Guo, Yu-Shen Liu, Wenping
Wang, Bisheng Yang | Robust Multiview Point Cloud Registration with Reliable Pose Graph
Initialization and History Reweighting | Accepted by CVPR 2023; Code at https://github.com/WHU-USI3DV/SGHR | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In this paper, we present a new method for the multiview registration of
point cloud. Previous multiview registration methods rely on exhaustive
pairwise registration to construct a densely-connected pose graph and apply
Iteratively Reweighted Least Square (IRLS) on the pose graph to compute the
scan poses. However, constructing a densely-connected graph is time-consuming
and contains lots of outlier edges, which makes the subsequent IRLS struggle to
find correct poses. To address the above problems, we first propose to use a
neural network to estimate the overlap between scan pairs, which enables us to
construct a sparse but reliable pose graph. Then, we design a novel history
reweighting function in the IRLS scheme, which has strong robustness to outlier
edges on the graph. In comparison with existing multiview registration methods,
our method achieves 11% higher registration recall on the 3DMatch dataset and
~13% lower registration errors on the ScanNet dataset while reducing ~70%
required pairwise registrations. Comprehensive ablation studies are conducted
to demonstrate the effectiveness of our designs.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 06:43:40 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wang",
"Haiping",
""
],
[
"Liu",
"Yuan",
""
],
[
"Dong",
"Zhen",
""
],
[
"Guo",
"Yulan",
""
],
[
"Liu",
"Yu-Shen",
""
],
[
"Wang",
"Wenping",
""
],
[
"Yang",
"Bisheng",
""
]
] | TITLE: Robust Multiview Point Cloud Registration with Reliable Pose Graph
Initialization and History Reweighting
ABSTRACT: In this paper, we present a new method for the multiview registration of
point cloud. Previous multiview registration methods rely on exhaustive
pairwise registration to construct a densely-connected pose graph and apply
Iteratively Reweighted Least Square (IRLS) on the pose graph to compute the
scan poses. However, constructing a densely-connected graph is time-consuming
and contains lots of outlier edges, which makes the subsequent IRLS struggle to
find correct poses. To address the above problems, we first propose to use a
neural network to estimate the overlap between scan pairs, which enables us to
construct a sparse but reliable pose graph. Then, we design a novel history
reweighting function in the IRLS scheme, which has strong robustness to outlier
edges on the graph. In comparison with existing multiview registration methods,
our method achieves 11% higher registration recall on the 3DMatch dataset and
~13% lower registration errors on the ScanNet dataset while reducing ~70%
required pairwise registrations. Comprehensive ablation studies are conducted
to demonstrate the effectiveness of our designs.
| no_new_dataset | 0.95511 |
2304.00486 | Ramtin Hosseini | Ramtin Hosseini, Li Zhang, Bhanu Garg, Pengtao Xie | Learning by Grouping: A Multilevel Optimization Framework for Improving
Fairness in Classification without Losing Accuracy | null | null | null | null | cs.LG cs.CY | http://creativecommons.org/licenses/by/4.0/ | The integration of machine learning models in various real-world applications
is becoming more prevalent to assist humans in their daily decision-making
tasks as a result of recent advancements in this field. However, it has been
discovered that there is a tradeoff between the accuracy and fairness of these
decision-making tasks. In some cases, these AI systems can be unfair by
exhibiting bias or discrimination against certain social groups, which can have
severe consequences in real life. Inspired by one of the most well-known human
learning skills called grouping, we address this issue by proposing a novel
machine learning framework where the ML model learns to group a diverse set of
problems into distinct subgroups to solve each subgroup using its specific
sub-model. Our proposed framework involves three stages of learning, which are
formulated as a three-level optimization problem: (i) learning to group
problems into different subgroups; (ii) learning group-specific sub-models for
problem-solving; and (iii) updating group assignments of training examples by
minimizing the validation loss. These three learning stages are performed
end-to-end in a joint manner using gradient descent. To improve fairness and
accuracy, we develop an efficient optimization algorithm to solve this
three-level optimization problem. To further reduce the risk of overfitting in
small datasets, we incorporate domain adaptation techniques in the second stage
of training. We further apply our method to neural architecture search.
Extensive experiments on various datasets demonstrate our method's
effectiveness and performance improvements in both fairness and accuracy. Our
proposed Learning by Grouping can reduce overfitting and achieve
state-of-the-art performances with fixed human-designed network architectures
and searchable network architectures on various datasets.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 08:45:08 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Hosseini",
"Ramtin",
""
],
[
"Zhang",
"Li",
""
],
[
"Garg",
"Bhanu",
""
],
[
"Xie",
"Pengtao",
""
]
] | TITLE: Learning by Grouping: A Multilevel Optimization Framework for Improving
Fairness in Classification without Losing Accuracy
ABSTRACT: The integration of machine learning models in various real-world applications
is becoming more prevalent to assist humans in their daily decision-making
tasks as a result of recent advancements in this field. However, it has been
discovered that there is a tradeoff between the accuracy and fairness of these
decision-making tasks. In some cases, these AI systems can be unfair by
exhibiting bias or discrimination against certain social groups, which can have
severe consequences in real life. Inspired by one of the most well-known human
learning skills called grouping, we address this issue by proposing a novel
machine learning framework where the ML model learns to group a diverse set of
problems into distinct subgroups to solve each subgroup using its specific
sub-model. Our proposed framework involves three stages of learning, which are
formulated as a three-level optimization problem: (i) learning to group
problems into different subgroups; (ii) learning group-specific sub-models for
problem-solving; and (iii) updating group assignments of training examples by
minimizing the validation loss. These three learning stages are performed
end-to-end in a joint manner using gradient descent. To improve fairness and
accuracy, we develop an efficient optimization algorithm to solve this
three-level optimization problem. To further reduce the risk of overfitting in
small datasets, we incorporate domain adaptation techniques in the second stage
of training. We further apply our method to neural architecture search.
Extensive experiments on various datasets demonstrate our method's
effectiveness and performance improvements in both fairness and accuracy. Our
proposed Learning by Grouping can reduce overfitting and achieve
state-of-the-art performances with fixed human-designed network architectures
and searchable network architectures on various datasets.
| no_new_dataset | 0.945651 |
2304.00495 | Lu Huo | Lu Huo, Jiahao Xia, Leijie Zhang, Haimin Zhang, Min Xu | Multimodal Hyperspectral Image Classification via Interconnected Fusion | 11 pages, five figures | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing multiple modality fusion methods, such as concatenation, summation,
and encoder-decoder-based fusion, have recently been employed to combine
modality characteristics of Hyperspectral Image (HSI) and Light Detection And
Ranging (LiDAR). However, these methods consider the relationship of HSI-LiDAR
signals from limited perspectives. More specifically, they overlook the
contextual information across modalities of HSI and LiDAR and the
intra-modality characteristics of LiDAR. In this paper, we provide a new
insight into feature fusion to explore the relationships across HSI and LiDAR
modalities comprehensively. An Interconnected Fusion (IF) framework is
proposed. Firstly, the center patch of the HSI input is extracted and
replicated to the size of the HSI input. Then, nine different perspectives in
the fusion matrix are generated by calculating self-attention and
cross-attention among the replicated center patch, HSI input, and corresponding
LiDAR input. In this way, the intra- and inter-modality characteristics can be
fully exploited, and contextual information is considered in both
intra-modality and inter-modality manner. These nine interrelated elements in
the fusion matrix can complement each other and eliminate biases, which can
generate a multi-modality representation for classification accurately.
Extensive experiments have been conducted on three widely used datasets:
Trento, MUUFL, and Houston. The IF framework achieves state-of-the-art results
on these datasets compared to existing approaches.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 09:46:13 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Huo",
"Lu",
""
],
[
"Xia",
"Jiahao",
""
],
[
"Zhang",
"Leijie",
""
],
[
"Zhang",
"Haimin",
""
],
[
"Xu",
"Min",
""
]
] | TITLE: Multimodal Hyperspectral Image Classification via Interconnected Fusion
ABSTRACT: Existing multiple modality fusion methods, such as concatenation, summation,
and encoder-decoder-based fusion, have recently been employed to combine
modality characteristics of Hyperspectral Image (HSI) and Light Detection And
Ranging (LiDAR). However, these methods consider the relationship of HSI-LiDAR
signals from limited perspectives. More specifically, they overlook the
contextual information across modalities of HSI and LiDAR and the
intra-modality characteristics of LiDAR. In this paper, we provide a new
insight into feature fusion to explore the relationships across HSI and LiDAR
modalities comprehensively. An Interconnected Fusion (IF) framework is
proposed. Firstly, the center patch of the HSI input is extracted and
replicated to the size of the HSI input. Then, nine different perspectives in
the fusion matrix are generated by calculating self-attention and
cross-attention among the replicated center patch, HSI input, and corresponding
LiDAR input. In this way, the intra- and inter-modality characteristics can be
fully exploited, and contextual information is considered in both
intra-modality and inter-modality manner. These nine interrelated elements in
the fusion matrix can complement each other and eliminate biases, which can
generate a multi-modality representation for classification accurately.
Extensive experiments have been conducted on three widely used datasets:
Trento, MUUFL, and Houston. The IF framework achieves state-of-the-art results
on these datasets compared to existing approaches.
| no_new_dataset | 0.946843 |
2304.00498 | Cheng Chen | Cheng Chen, Yueming Lyu, Ivor W.Tsang | Adversary-Aware Partial label learning with Label distillation | null | null | null | null | cs.LG cs.CV | http://creativecommons.org/licenses/by/4.0/ | To ensure that the data collected from human subjects is entrusted with a
secret, rival labels are introduced to conceal the information provided by the
participants on purpose. The corresponding learning task can be formulated as a
noisy partial-label learning problem. However, conventional partial-label
learning (PLL) methods are still vulnerable to the high ratio of noisy partial
labels, especially in a large labelling space. To learn a more robust model, we
present Adversary-Aware Partial Label Learning and introduce the
$\textit{rival}$, a set of noisy labels, to the collection of candidate labels
for each instance. By introducing the rival label, the predictive distribution
of PLL is factorised such that a handy predictive label is achieved with less
uncertainty coming from the transition matrix, assuming the rival generation
process is known. Nonetheless, the predictive accuracy is still insufficient to
produce an sufficiently accurate positive sample set to leverage the clustering
effect of the contrastive loss function. Moreover, the inclusion of rivals also
brings an inconsistency issue for the classifier and risk function due to the
intractability of the transition matrix. Consequently, an adversarial teacher
within momentum (ATM) disambiguation algorithm is proposed to cope with the
situation, allowing us to obtain a provably consistent classifier and risk
function. In addition, our method has shown high resiliency to the choice of
the label noise transition matrix. Extensive experiments demonstrate that our
method achieves promising results on the CIFAR10, CIFAR100 and CUB200 datasets.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 10:18:30 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Chen",
"Cheng",
""
],
[
"Lyu",
"Yueming",
""
],
[
"Tsang",
"Ivor W.",
""
]
] | TITLE: Adversary-Aware Partial label learning with Label distillation
ABSTRACT: To ensure that the data collected from human subjects is entrusted with a
secret, rival labels are introduced to conceal the information provided by the
participants on purpose. The corresponding learning task can be formulated as a
noisy partial-label learning problem. However, conventional partial-label
learning (PLL) methods are still vulnerable to the high ratio of noisy partial
labels, especially in a large labelling space. To learn a more robust model, we
present Adversary-Aware Partial Label Learning and introduce the
$\textit{rival}$, a set of noisy labels, to the collection of candidate labels
for each instance. By introducing the rival label, the predictive distribution
of PLL is factorised such that a handy predictive label is achieved with less
uncertainty coming from the transition matrix, assuming the rival generation
process is known. Nonetheless, the predictive accuracy is still insufficient to
produce an sufficiently accurate positive sample set to leverage the clustering
effect of the contrastive loss function. Moreover, the inclusion of rivals also
brings an inconsistency issue for the classifier and risk function due to the
intractability of the transition matrix. Consequently, an adversarial teacher
within momentum (ATM) disambiguation algorithm is proposed to cope with the
situation, allowing us to obtain a provably consistent classifier and risk
function. In addition, our method has shown high resiliency to the choice of
the label noise transition matrix. Extensive experiments demonstrate that our
method achieves promising results on the CIFAR10, CIFAR100 and CUB200 datasets.
| no_new_dataset | 0.947332 |
2304.00502 | Aristotelis Ballas | Aristotelis Ballas and Christos Diou | CNNs with Multi-Level Attention for Domain Generalization | Accepted for publication in ICMR '23 (ACM International Conference on
Multimedia Retrieval). This is a preprint of the final version | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In the past decade, deep convolutional neural networks have achieved
significant success in image classification and ranking and have therefore
found numerous applications in multimedia content retrieval. Still, these
models suffer from performance degradation when neural networks are tested on
out-of-distribution scenarios or on data originating from previously unseen
data Domains. In the present work, we focus on this problem of Domain
Generalization and propose an alternative neural network architecture for
robust, out-of-distribution image classification. We attempt to produce a model
that focuses on the causal features of the depicted class for robust image
classification in the Domain Generalization setting. To achieve this, we
propose attending to multiple-levels of information throughout a Convolutional
Neural Network and leveraging the most important attributes of an image by
employing trainable attention mechanisms. To validate our method, we evaluate
our model on four widely accepted Domain Generalization benchmarks, on which
our model is able to surpass previously reported baselines in three out of four
datasets and achieve the second best score in the fourth one.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 10:34:40 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ballas",
"Aristotelis",
""
],
[
"Diou",
"Christos",
""
]
] | TITLE: CNNs with Multi-Level Attention for Domain Generalization
ABSTRACT: In the past decade, deep convolutional neural networks have achieved
significant success in image classification and ranking and have therefore
found numerous applications in multimedia content retrieval. Still, these
models suffer from performance degradation when neural networks are tested on
out-of-distribution scenarios or on data originating from previously unseen
data Domains. In the present work, we focus on this problem of Domain
Generalization and propose an alternative neural network architecture for
robust, out-of-distribution image classification. We attempt to produce a model
that focuses on the causal features of the depicted class for robust image
classification in the Domain Generalization setting. To achieve this, we
propose attending to multiple-levels of information throughout a Convolutional
Neural Network and leveraging the most important attributes of an image by
employing trainable attention mechanisms. To validate our method, we evaluate
our model on four widely accepted Domain Generalization benchmarks, on which
our model is able to surpass previously reported baselines in three out of four
datasets and achieve the second best score in the fourth one.
| no_new_dataset | 0.951908 |
2304.00520 | Halil Faruk Karagoz | Halil Faruk Karagoz, Gulcin Baykal, Irem Arikan Eksi, Gozde Unal | Textile Pattern Generation Using Diffusion Models | Accepted at ITFC 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The problem of text-guided image generation is a complex task in Computer
Vision, with various applications, including creating visually appealing
artwork and realistic product images. One popular solution widely used for this
task is the diffusion model, a generative model that generates images through
an iterative process. Although diffusion models have demonstrated promising
results for various image generation tasks, they may only sometimes produce
satisfactory results when applied to more specific domains, such as the
generation of textile patterns based on text guidance. This study presents a
fine-tuned diffusion model specifically trained for textile pattern generation
by text guidance to address this issue. The study involves the collection of
various textile pattern images and their captioning with the help of another AI
model. The fine-tuned diffusion model is trained with this newly created
dataset, and its results are compared with the baseline models visually and
numerically. The results demonstrate that the proposed fine-tuned diffusion
model outperforms the baseline models in terms of pattern quality and
efficiency in textile pattern generation by text guidance. This study presents
a promising solution to the problem of text-guided textile pattern generation
and has the potential to simplify the design process within the textile
industry.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 12:12:24 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Karagoz",
"Halil Faruk",
""
],
[
"Baykal",
"Gulcin",
""
],
[
"Eksi",
"Irem Arikan",
""
],
[
"Unal",
"Gozde",
""
]
] | TITLE: Textile Pattern Generation Using Diffusion Models
ABSTRACT: The problem of text-guided image generation is a complex task in Computer
Vision, with various applications, including creating visually appealing
artwork and realistic product images. One popular solution widely used for this
task is the diffusion model, a generative model that generates images through
an iterative process. Although diffusion models have demonstrated promising
results for various image generation tasks, they may only sometimes produce
satisfactory results when applied to more specific domains, such as the
generation of textile patterns based on text guidance. This study presents a
fine-tuned diffusion model specifically trained for textile pattern generation
by text guidance to address this issue. The study involves the collection of
various textile pattern images and their captioning with the help of another AI
model. The fine-tuned diffusion model is trained with this newly created
dataset, and its results are compared with the baseline models visually and
numerically. The results demonstrate that the proposed fine-tuned diffusion
model outperforms the baseline models in terms of pattern quality and
efficiency in textile pattern generation by text guidance. This study presents
a promising solution to the problem of text-guided textile pattern generation
and has the potential to simplify the design process within the textile
industry.
| new_dataset | 0.957873 |
2304.00525 | Huitong Yang | Huitong Yang, Xuyang Bai, Xinge Zhu, and Yuexin Ma | One Training for Multiple Deployments: Polar-based Adaptive BEV
Perception for Autonomous Driving | 8 pages,4 figures. Accepted by IEEE International Conference on
Robotics and Automation (ICRA) | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Current on-board chips usually have different computing power, which means
multiple training processes are needed for adapting the same learning-based
algorithm to different chips, costing huge computing resources. The situation
becomes even worse for 3D perception methods with large models. Previous
vision-centric 3D perception approaches are trained with regular
grid-represented feature maps of fixed resolutions, which is not applicable to
adapt to other grid scales, limiting wider deployment. In this paper, we
leverage the Polar representation when constructing the BEV feature map from
images in order to achieve the goal of training once for multiple deployments.
Specifically, the feature along rays in Polar space can be easily adaptively
sampled and projected to the feature in Cartesian space with arbitrary
resolutions. To further improve the adaptation capability, we make multi-scale
contextual information interact with each other to enhance the feature
representation. Experiments on a large-scale autonomous driving dataset show
that our method outperforms others as for the good property of one training for
multiple deployments.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 12:37:28 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Yang",
"Huitong",
""
],
[
"Bai",
"Xuyang",
""
],
[
"Zhu",
"Xinge",
""
],
[
"Ma",
"Yuexin",
""
]
] | TITLE: One Training for Multiple Deployments: Polar-based Adaptive BEV
Perception for Autonomous Driving
ABSTRACT: Current on-board chips usually have different computing power, which means
multiple training processes are needed for adapting the same learning-based
algorithm to different chips, costing huge computing resources. The situation
becomes even worse for 3D perception methods with large models. Previous
vision-centric 3D perception approaches are trained with regular
grid-represented feature maps of fixed resolutions, which is not applicable to
adapt to other grid scales, limiting wider deployment. In this paper, we
leverage the Polar representation when constructing the BEV feature map from
images in order to achieve the goal of training once for multiple deployments.
Specifically, the feature along rays in Polar space can be easily adaptively
sampled and projected to the feature in Cartesian space with arbitrary
resolutions. To further improve the adaptation capability, we make multi-scale
contextual information interact with each other to enhance the feature
representation. Experiments on a large-scale autonomous driving dataset show
that our method outperforms others as for the good property of one training for
multiple deployments.
| no_new_dataset | 0.951774 |
2304.00531 | Zihao Zhao | Zihao Zhao, Xiaodong Ge, Zhihong Shen | S2CTrans: Building a bridge from SPARQL to Cypher | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In graph data applications, data is primarily maintained using two models:
RDF (Resource Description Framework) and property graph. The property graph
model is widely adopted by industry, leading to property graph databases
generally outperforming RDF databases in graph traversal query performance.
However, users often prefer SPARQL as their query language, as it is the W3C's
recommended standard. Consequently, exploring
SPARQL-to-Property-Graph-Query-Language translation is crucial for enhancing
graph query language interoperability and enabling effective querying of
property graphs using SPARQL. Despite the substantial differences in semantic
representation and processing logic between SPARQL and property graph query
languages like Cypher, this paper demonstrates the feasibility of translating
SPARQL to Cypher for graph traversal queries using graph relational algebra. We
present the S2CTrans framework, which achieves SPARQL-to-Cypher translation
while preserving the original semantics. Experimental results with the Berlin
SPARQL Benchmark (BSBM) datasets show that S2CTrans successfully converts most
SELECT queries in the SPARQL 1.1 specification into type-safe Cypher
statements, maintaining result consistency and improving the efficiency of data
querying using SPARQL.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 13:18:20 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Zhao",
"Zihao",
""
],
[
"Ge",
"Xiaodong",
""
],
[
"Shen",
"Zhihong",
""
]
] | TITLE: S2CTrans: Building a bridge from SPARQL to Cypher
ABSTRACT: In graph data applications, data is primarily maintained using two models:
RDF (Resource Description Framework) and property graph. The property graph
model is widely adopted by industry, leading to property graph databases
generally outperforming RDF databases in graph traversal query performance.
However, users often prefer SPARQL as their query language, as it is the W3C's
recommended standard. Consequently, exploring
SPARQL-to-Property-Graph-Query-Language translation is crucial for enhancing
graph query language interoperability and enabling effective querying of
property graphs using SPARQL. Despite the substantial differences in semantic
representation and processing logic between SPARQL and property graph query
languages like Cypher, this paper demonstrates the feasibility of translating
SPARQL to Cypher for graph traversal queries using graph relational algebra. We
present the S2CTrans framework, which achieves SPARQL-to-Cypher translation
while preserving the original semantics. Experimental results with the Berlin
SPARQL Benchmark (BSBM) datasets show that S2CTrans successfully converts most
SELECT queries in the SPARQL 1.1 specification into type-safe Cypher
statements, maintaining result consistency and improving the efficiency of data
querying using SPARQL.
| no_new_dataset | 0.940626 |
2304.00534 | Zichun Wang | Zichun Wang, Ying Fu, Ji Liu, Yulun Zhang | LG-BPN: Local and Global Blind-Patch Network for Self-Supervised
Real-World Denoising | Accepted to CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the significant results on synthetic noise under simplified
assumptions, most self-supervised denoising methods fail under real noise due
to the strong spatial noise correlation, including the advanced self-supervised
blind-spot networks (BSNs). For recent methods targeting real-world denoising,
they either suffer from ignoring this spatial correlation, or are limited by
the destruction of fine textures for under-considering the correlation. In this
paper, we present a novel method called LG-BPN for self-supervised real-world
denoising, which takes the spatial correlation statistic into our network
design for local detail restoration, and also brings the long-range
dependencies modeling ability to previously CNN-based BSN methods. First, based
on the correlation statistic, we propose a densely-sampled patch-masked
convolution module. By taking more neighbor pixels with low noise correlation
into account, we enable a denser local receptive field, preserving more useful
information for enhanced fine structure recovery. Second, we propose a dilated
Transformer block to allow distant context exploitation in BSN. This global
perception addresses the intrinsic deficiency of BSN, whose receptive field is
constrained by the blind spot requirement, which can not be fully resolved by
the previous CNN-based BSNs. These two designs enable LG-BPN to fully exploit
both the detailed structure and the global interaction in a blind manner.
Extensive results on real-world datasets demonstrate the superior performance
of our method. https://github.com/Wang-XIaoDingdd/LGBPN
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 13:32:16 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Wang",
"Zichun",
""
],
[
"Fu",
"Ying",
""
],
[
"Liu",
"Ji",
""
],
[
"Zhang",
"Yulun",
""
]
] | TITLE: LG-BPN: Local and Global Blind-Patch Network for Self-Supervised
Real-World Denoising
ABSTRACT: Despite the significant results on synthetic noise under simplified
assumptions, most self-supervised denoising methods fail under real noise due
to the strong spatial noise correlation, including the advanced self-supervised
blind-spot networks (BSNs). For recent methods targeting real-world denoising,
they either suffer from ignoring this spatial correlation, or are limited by
the destruction of fine textures for under-considering the correlation. In this
paper, we present a novel method called LG-BPN for self-supervised real-world
denoising, which takes the spatial correlation statistic into our network
design for local detail restoration, and also brings the long-range
dependencies modeling ability to previously CNN-based BSN methods. First, based
on the correlation statistic, we propose a densely-sampled patch-masked
convolution module. By taking more neighbor pixels with low noise correlation
into account, we enable a denser local receptive field, preserving more useful
information for enhanced fine structure recovery. Second, we propose a dilated
Transformer block to allow distant context exploitation in BSN. This global
perception addresses the intrinsic deficiency of BSN, whose receptive field is
constrained by the blind spot requirement, which can not be fully resolved by
the previous CNN-based BSNs. These two designs enable LG-BPN to fully exploit
both the detailed structure and the global interaction in a blind manner.
Extensive results on real-world datasets demonstrate the superior performance
of our method. https://github.com/Wang-XIaoDingdd/LGBPN
| no_new_dataset | 0.950365 |
2304.00545 | Qijiong Liu | Qijiong Liu, Jieming Zhu, Jiahao Wu, Tiandeng Wu, Zhenhua Dong,
Xiao-Ming Wu | FANS: Fast Non-Autoregressive Sequence Generation for Item List
Continuation | 10 pages, ACM The Web Conference 2023 accepted paper | null | 10.1145/3543507.3583430 | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | User-curated item lists, such as video-based playlists on Youtube and
book-based lists on Goodreads, have become prevalent for content sharing on
online platforms. Item list continuation is proposed to model the overall trend
of a list and predict subsequent items. Recently, Transformer-based models have
shown promise in comprehending contextual information and capturing item
relationships in a list. However, deploying them in real-time industrial
applications is challenging, mainly because the autoregressive generation
mechanism used in them is time-consuming. In this paper, we propose a novel
fast non-autoregressive sequence generation model, namely FANS, to enhance
inference efficiency and quality for item list continuation. First, we use a
non-autoregressive generation mechanism to decode next $K$ items simultaneously
instead of one by one in existing models. Then, we design a two-stage
classifier to replace the vanilla classifier used in current transformer-based
models to further reduce the decoding time. Moreover, to improve the quality of
non-autoregressive generation, we employ a curriculum learning strategy to
optimize training. Experimental results on four real-world item list
continuation datasets including Zhihu, Spotify, AotM, and Goodreads show that
our FANS model can significantly improve inference efficiency (up to 8.7x)
while achieving competitive or better generation quality for item list
continuation compared with the state-of-the-art autoregressive models. We also
validate the efficiency of FANS in an industrial setting. Our source code and
data will be available at MindSpore/models and Github.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 14:41:17 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Liu",
"Qijiong",
""
],
[
"Zhu",
"Jieming",
""
],
[
"Wu",
"Jiahao",
""
],
[
"Wu",
"Tiandeng",
""
],
[
"Dong",
"Zhenhua",
""
],
[
"Wu",
"Xiao-Ming",
""
]
] | TITLE: FANS: Fast Non-Autoregressive Sequence Generation for Item List
Continuation
ABSTRACT: User-curated item lists, such as video-based playlists on Youtube and
book-based lists on Goodreads, have become prevalent for content sharing on
online platforms. Item list continuation is proposed to model the overall trend
of a list and predict subsequent items. Recently, Transformer-based models have
shown promise in comprehending contextual information and capturing item
relationships in a list. However, deploying them in real-time industrial
applications is challenging, mainly because the autoregressive generation
mechanism used in them is time-consuming. In this paper, we propose a novel
fast non-autoregressive sequence generation model, namely FANS, to enhance
inference efficiency and quality for item list continuation. First, we use a
non-autoregressive generation mechanism to decode next $K$ items simultaneously
instead of one by one in existing models. Then, we design a two-stage
classifier to replace the vanilla classifier used in current transformer-based
models to further reduce the decoding time. Moreover, to improve the quality of
non-autoregressive generation, we employ a curriculum learning strategy to
optimize training. Experimental results on four real-world item list
continuation datasets including Zhihu, Spotify, AotM, and Goodreads show that
our FANS model can significantly improve inference efficiency (up to 8.7x)
while achieving competitive or better generation quality for item list
continuation compared with the state-of-the-art autoregressive models. We also
validate the efficiency of FANS in an industrial setting. Our source code and
data will be available at MindSpore/models and Github.
| no_new_dataset | 0.952131 |
2304.00546 | Pranav Rajpurkar | Alexander Ke, Shih-Cheng Huang, Chloe P O'Connell, Michal Klimont,
Serena Yeung, Pranav Rajpurkar | Video Pretraining Advances 3D Deep Learning on Chest CT Tasks | Accepted at MIDL 2023 | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Pretraining on large natural image classification datasets such as ImageNet
has aided model development on data-scarce 2D medical tasks. 3D medical tasks
often have much less data than 2D medical tasks, prompting practitioners to
rely on pretrained 2D models to featurize slices. However, these 2D models have
been surpassed by 3D models on 3D computer vision benchmarks since they do not
natively leverage cross-sectional or temporal information. In this study, we
explore whether natural video pretraining for 3D models can enable higher
performance on smaller datasets for 3D medical tasks. We demonstrate video
pretraining improves the average performance of seven 3D models on two chest CT
datasets, regardless of finetuning dataset size, and that video pretraining
allows 3D models to outperform 2D baselines. Lastly, we observe that
pretraining on the large-scale out-of-domain Kinetics dataset improves
performance more than pretraining on a typically-sized in-domain CT dataset.
Our results show consistent benefits of video pretraining across a wide array
of architectures, tasks, and training dataset sizes, supporting a shift from
small-scale in-domain pretraining to large-scale out-of-domain pretraining for
3D medical tasks. Our code is available at:
https://github.com/rajpurkarlab/chest-ct-pretraining
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 14:46:58 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Ke",
"Alexander",
""
],
[
"Huang",
"Shih-Cheng",
""
],
[
"O'Connell",
"Chloe P",
""
],
[
"Klimont",
"Michal",
""
],
[
"Yeung",
"Serena",
""
],
[
"Rajpurkar",
"Pranav",
""
]
] | TITLE: Video Pretraining Advances 3D Deep Learning on Chest CT Tasks
ABSTRACT: Pretraining on large natural image classification datasets such as ImageNet
has aided model development on data-scarce 2D medical tasks. 3D medical tasks
often have much less data than 2D medical tasks, prompting practitioners to
rely on pretrained 2D models to featurize slices. However, these 2D models have
been surpassed by 3D models on 3D computer vision benchmarks since they do not
natively leverage cross-sectional or temporal information. In this study, we
explore whether natural video pretraining for 3D models can enable higher
performance on smaller datasets for 3D medical tasks. We demonstrate video
pretraining improves the average performance of seven 3D models on two chest CT
datasets, regardless of finetuning dataset size, and that video pretraining
allows 3D models to outperform 2D baselines. Lastly, we observe that
pretraining on the large-scale out-of-domain Kinetics dataset improves
performance more than pretraining on a typically-sized in-domain CT dataset.
Our results show consistent benefits of video pretraining across a wide array
of architectures, tasks, and training dataset sizes, supporting a shift from
small-scale in-domain pretraining to large-scale out-of-domain pretraining for
3D medical tasks. Our code is available at:
https://github.com/rajpurkarlab/chest-ct-pretraining
| no_new_dataset | 0.940243 |
2304.00557 | Giang Nguyen | Viet H. Pham, Thang M. Pham, Giang Nguyen, Long Nguyen, Dien Dinh | Semi-supervised Neural Machine Translation with Consistency
Regularization for Low-Resource Languages | TMP and GN contributed equally | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | The advent of deep learning has led to a significant gain in machine
translation. However, most of the studies required a large parallel dataset
which is scarce and expensive to construct and even unavailable for some
languages. This paper presents a simple yet effective method to tackle this
problem for low-resource languages by augmenting high-quality sentence pairs
and training NMT models in a semi-supervised manner. Specifically, our approach
combines the cross-entropy loss for supervised learning with KL Divergence for
unsupervised fashion given pseudo and augmented target sentences derived from
the model. We also introduce a SentenceBERT-based filter to enhance the quality
of augmenting data by retaining semantically similar sentence pairs.
Experimental results show that our approach significantly improves NMT
baselines, especially on low-resource datasets with 0.46--2.03 BLEU scores. We
also demonstrate that using unsupervised training for augmented data is more
efficient than reusing the ground-truth target sentences for supervised
learning.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 15:24:08 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Pham",
"Viet H.",
""
],
[
"Pham",
"Thang M.",
""
],
[
"Nguyen",
"Giang",
""
],
[
"Nguyen",
"Long",
""
],
[
"Dinh",
"Dien",
""
]
] | TITLE: Semi-supervised Neural Machine Translation with Consistency
Regularization for Low-Resource Languages
ABSTRACT: The advent of deep learning has led to a significant gain in machine
translation. However, most of the studies required a large parallel dataset
which is scarce and expensive to construct and even unavailable for some
languages. This paper presents a simple yet effective method to tackle this
problem for low-resource languages by augmenting high-quality sentence pairs
and training NMT models in a semi-supervised manner. Specifically, our approach
combines the cross-entropy loss for supervised learning with KL Divergence for
unsupervised fashion given pseudo and augmented target sentences derived from
the model. We also introduce a SentenceBERT-based filter to enhance the quality
of augmenting data by retaining semantically similar sentence pairs.
Experimental results show that our approach significantly improves NMT
baselines, especially on low-resource datasets with 0.46--2.03 BLEU scores. We
also demonstrate that using unsupervised training for augmented data is more
efficient than reusing the ground-truth target sentences for supervised
learning.
| no_new_dataset | 0.949669 |
2304.00584 | Afagh Mehri Shervedani | Afagh Mehri Shervedani, Siyu Li, Natawut Monaikul, Bahareh Abbasi,
Barbara Di Eugenio, Milos Zefran | An End-to-End Human Simulator for Task-Oriented Multimodal Human-Robot
Collaboration | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a neural network-based user simulator that can provide a
multimodal interactive environment for training Reinforcement Learning (RL)
agents in collaborative tasks involving multiple modes of communication. The
simulator is trained on the existing ELDERLY-AT-HOME corpus and accommodates
multiple modalities such as language, pointing gestures, and haptic-ostensive
actions. The paper also presents a novel multimodal data augmentation approach,
which addresses the challenge of using a limited dataset due to the expensive
and time-consuming nature of collecting human demonstrations. Overall, the
study highlights the potential for using RL and multimodal user simulators in
developing and improving domestic assistive robots.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 18:02:26 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Shervedani",
"Afagh Mehri",
""
],
[
"Li",
"Siyu",
""
],
[
"Monaikul",
"Natawut",
""
],
[
"Abbasi",
"Bahareh",
""
],
[
"Di Eugenio",
"Barbara",
""
],
[
"Zefran",
"Milos",
""
]
] | TITLE: An End-to-End Human Simulator for Task-Oriented Multimodal Human-Robot
Collaboration
ABSTRACT: This paper proposes a neural network-based user simulator that can provide a
multimodal interactive environment for training Reinforcement Learning (RL)
agents in collaborative tasks involving multiple modes of communication. The
simulator is trained on the existing ELDERLY-AT-HOME corpus and accommodates
multiple modalities such as language, pointing gestures, and haptic-ostensive
actions. The paper also presents a novel multimodal data augmentation approach,
which addresses the challenge of using a limited dataset due to the expensive
and time-consuming nature of collecting human demonstrations. Overall, the
study highlights the potential for using RL and multimodal user simulators in
developing and improving domestic assistive robots.
| no_new_dataset | 0.948917 |
2304.00592 | Cheng Deng | Cheng Deng, Bo Tong, Luoyi Fu, Jiaxin Ding, Dexing Cao, Xinbing Wang,
Chenghu Zhou | PK-Chat: Pointer Network Guided Knowledge Driven Generative Dialogue
Model | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the research of end-to-end dialogue systems, using real-world knowledge to
generate natural, fluent, and human-like utterances with correct answers is
crucial. However, domain-specific conversational dialogue systems may be
incoherent and introduce erroneous external information to answer questions due
to the out-of-vocabulary issue or the wrong knowledge from the parameters of
the neural network. In this work, we propose PK-Chat, a Pointer network guided
Knowledge-driven generative dialogue model, incorporating a unified pretrained
language model and a pointer network over knowledge graphs. The words generated
by PK-Chat in the dialogue are derived from the prediction of word lists and
the direct prediction of the external knowledge graph knowledge. Moreover,
based on the PK-Chat, a dialogue system is built for academic scenarios in the
case of geosciences. Finally, an academic dialogue benchmark is constructed to
evaluate the quality of dialogue systems in academic scenarios and the source
code is available online.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 18:23:13 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Deng",
"Cheng",
""
],
[
"Tong",
"Bo",
""
],
[
"Fu",
"Luoyi",
""
],
[
"Ding",
"Jiaxin",
""
],
[
"Cao",
"Dexing",
""
],
[
"Wang",
"Xinbing",
""
],
[
"Zhou",
"Chenghu",
""
]
] | TITLE: PK-Chat: Pointer Network Guided Knowledge Driven Generative Dialogue
Model
ABSTRACT: In the research of end-to-end dialogue systems, using real-world knowledge to
generate natural, fluent, and human-like utterances with correct answers is
crucial. However, domain-specific conversational dialogue systems may be
incoherent and introduce erroneous external information to answer questions due
to the out-of-vocabulary issue or the wrong knowledge from the parameters of
the neural network. In this work, we propose PK-Chat, a Pointer network guided
Knowledge-driven generative dialogue model, incorporating a unified pretrained
language model and a pointer network over knowledge graphs. The words generated
by PK-Chat in the dialogue are derived from the prediction of word lists and
the direct prediction of the external knowledge graph knowledge. Moreover,
based on the PK-Chat, a dialogue system is built for academic scenarios in the
case of geosciences. Finally, an academic dialogue benchmark is constructed to
evaluate the quality of dialogue systems in academic scenarios and the source
code is available online.
| no_new_dataset | 0.949763 |
2304.00600 | Paul Micaelli | Paul Micaelli, Arash Vahdat, Hongxu Yin, Jan Kautz, Pavlo Molchanov | Recurrence without Recurrence: Stable Video Landmark Detection with Deep
Equilibrium Models | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Cascaded computation, whereby predictions are recurrently refined over
several stages, has been a persistent theme throughout the development of
landmark detection models. In this work, we show that the recently proposed
Deep Equilibrium Model (DEQ) can be naturally adapted to this form of
computation. Our Landmark DEQ (LDEQ) achieves state-of-the-art performance on
the challenging WFLW facial landmark dataset, reaching $3.92$ NME with fewer
parameters and a training memory cost of $\mathcal{O}(1)$ in the number of
recurrent modules. Furthermore, we show that DEQs are particularly suited for
landmark detection in videos. In this setting, it is typical to train on still
images due to the lack of labelled videos. This can lead to a ``flickering''
effect at inference time on video, whereby a model can rapidly oscillate
between different plausible solutions across consecutive frames. By rephrasing
DEQs as a constrained optimization, we emulate recurrence at inference time,
despite not having access to temporal data at training time. This Recurrence
without Recurrence (RwR) paradigm helps in reducing landmark flicker, which we
demonstrate by introducing a new metric, normalized mean flicker (NMF), and
contributing a new facial landmark video dataset (WFLW-V) targeting landmark
uncertainty. On the WFLW-V hard subset made up of $500$ videos, our LDEQ with
RwR improves the NME and NMF by $10$ and $13\%$ respectively, compared to the
strongest previously published model using a hand-tuned conventional filter.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 19:08:02 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Micaelli",
"Paul",
""
],
[
"Vahdat",
"Arash",
""
],
[
"Yin",
"Hongxu",
""
],
[
"Kautz",
"Jan",
""
],
[
"Molchanov",
"Pavlo",
""
]
] | TITLE: Recurrence without Recurrence: Stable Video Landmark Detection with Deep
Equilibrium Models
ABSTRACT: Cascaded computation, whereby predictions are recurrently refined over
several stages, has been a persistent theme throughout the development of
landmark detection models. In this work, we show that the recently proposed
Deep Equilibrium Model (DEQ) can be naturally adapted to this form of
computation. Our Landmark DEQ (LDEQ) achieves state-of-the-art performance on
the challenging WFLW facial landmark dataset, reaching $3.92$ NME with fewer
parameters and a training memory cost of $\mathcal{O}(1)$ in the number of
recurrent modules. Furthermore, we show that DEQs are particularly suited for
landmark detection in videos. In this setting, it is typical to train on still
images due to the lack of labelled videos. This can lead to a ``flickering''
effect at inference time on video, whereby a model can rapidly oscillate
between different plausible solutions across consecutive frames. By rephrasing
DEQs as a constrained optimization, we emulate recurrence at inference time,
despite not having access to temporal data at training time. This Recurrence
without Recurrence (RwR) paradigm helps in reducing landmark flicker, which we
demonstrate by introducing a new metric, normalized mean flicker (NMF), and
contributing a new facial landmark video dataset (WFLW-V) targeting landmark
uncertainty. On the WFLW-V hard subset made up of $500$ videos, our LDEQ with
RwR improves the NME and NMF by $10$ and $13\%$ respectively, compared to the
strongest previously published model using a hand-tuned conventional filter.
| no_new_dataset | 0.776284 |
2304.00629 | Boyang Lyu | Boyang Lyu, Thuan Nguyen, Matthias Scheutz, Prakash Ishwar, Shuchin
Aeron | A principled approach to model validation in domain generalization | Accepted to ICASSP 2023 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Domain generalization aims to learn a model with good generalization ability,
that is, the learned model should not only perform well on several seen domains
but also on unseen domains with different data distributions. State-of-the-art
domain generalization methods typically train a representation function
followed by a classifier jointly to minimize both the classification risk and
the domain discrepancy. However, when it comes to model selection, most of
these methods rely on traditional validation routines that select models solely
based on the lowest classification risk on the validation set. In this paper,
we theoretically demonstrate a trade-off between minimizing classification risk
and mitigating domain discrepancy, i.e., it is impossible to achieve the
minimum of these two objectives simultaneously. Motivated by this theoretical
result, we propose a novel model selection method suggesting that the
validation process should account for both the classification risk and the
domain discrepancy. We validate the effectiveness of the proposed method by
numerical results on several domain generalization datasets.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 21:12:13 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Lyu",
"Boyang",
""
],
[
"Nguyen",
"Thuan",
""
],
[
"Scheutz",
"Matthias",
""
],
[
"Ishwar",
"Prakash",
""
],
[
"Aeron",
"Shuchin",
""
]
] | TITLE: A principled approach to model validation in domain generalization
ABSTRACT: Domain generalization aims to learn a model with good generalization ability,
that is, the learned model should not only perform well on several seen domains
but also on unseen domains with different data distributions. State-of-the-art
domain generalization methods typically train a representation function
followed by a classifier jointly to minimize both the classification risk and
the domain discrepancy. However, when it comes to model selection, most of
these methods rely on traditional validation routines that select models solely
based on the lowest classification risk on the validation set. In this paper,
we theoretically demonstrate a trade-off between minimizing classification risk
and mitigating domain discrepancy, i.e., it is impossible to achieve the
minimum of these two objectives simultaneously. Motivated by this theoretical
result, we propose a novel model selection method suggesting that the
validation process should account for both the classification risk and the
domain discrepancy. We validate the effectiveness of the proposed method by
numerical results on several domain generalization datasets.
| no_new_dataset | 0.949949 |
2304.00634 | Dwip Dalal | Dwip Dalal, Vivek Srivastava, Mayank Singh | MMT: A Multilingual and Multi-Topic Indian Social Media Dataset | null | EACL Workshop C3NLP 2023 | null | null | cs.CL cs.LG cs.SI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Social media plays a significant role in cross-cultural communication. A vast
amount of this occurs in code-mixed and multilingual form, posing a significant
challenge to Natural Language Processing (NLP) tools for processing such
information, like language identification, topic modeling, and named-entity
recognition. To address this, we introduce a large-scale multilingual, and
multi-topic dataset (MMT) collected from Twitter (1.7 million Tweets),
encompassing 13 coarse-grained and 63 fine-grained topics in the Indian
context. We further annotate a subset of 5,346 tweets from the MMT dataset with
various Indian languages and their code-mixed counterparts. Also, we
demonstrate that the currently existing tools fail to capture the linguistic
diversity in MMT on two downstream tasks, i.e., topic modeling and language
identification. To facilitate future research, we will make the anonymized and
annotated dataset available in the public domain.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 21:39:00 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Dalal",
"Dwip",
""
],
[
"Srivastava",
"Vivek",
""
],
[
"Singh",
"Mayank",
""
]
] | TITLE: MMT: A Multilingual and Multi-Topic Indian Social Media Dataset
ABSTRACT: Social media plays a significant role in cross-cultural communication. A vast
amount of this occurs in code-mixed and multilingual form, posing a significant
challenge to Natural Language Processing (NLP) tools for processing such
information, like language identification, topic modeling, and named-entity
recognition. To address this, we introduce a large-scale multilingual, and
multi-topic dataset (MMT) collected from Twitter (1.7 million Tweets),
encompassing 13 coarse-grained and 63 fine-grained topics in the Indian
context. We further annotate a subset of 5,346 tweets from the MMT dataset with
various Indian languages and their code-mixed counterparts. Also, we
demonstrate that the currently existing tools fail to capture the linguistic
diversity in MMT on two downstream tasks, i.e., topic modeling and language
identification. To facilitate future research, we will make the anonymized and
annotated dataset available in the public domain.
| new_dataset | 0.962778 |
2304.00636 | Rabia Bounaama | Rabia Bounaama, Mohammed El Amine Abderrahim | Classifying COVID-19 Related Tweets for Fake News Detection and
Sentiment Analysis with BERT-based Models | CERIST'22: CERIST NLP Challenge 2022, March 29, 2023, Algeria,
Algiers | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | The present paper is about the participation of our team "techno" on
CERIST'22 shared tasks. We used an available dataset "task1.c" related to
covid-19 pandemic. It comprises 4128 tweets for sentiment analysis task and
8661 tweets for fake news detection task. We used natural language processing
tools with the combination of the most renowned pre-trained language models
BERT (Bidirectional Encoder Representations from Transformers). The results
shows the efficacy of pre-trained language models as we attained an accuracy of
0.93 for the sentiment analysis task and 0.90 for the fake news detection task.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 22:00:27 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Bounaama",
"Rabia",
""
],
[
"Abderrahim",
"Mohammed El Amine",
""
]
] | TITLE: Classifying COVID-19 Related Tweets for Fake News Detection and
Sentiment Analysis with BERT-based Models
ABSTRACT: The present paper is about the participation of our team "techno" on
CERIST'22 shared tasks. We used an available dataset "task1.c" related to
covid-19 pandemic. It comprises 4128 tweets for sentiment analysis task and
8661 tweets for fake news detection task. We used natural language processing
tools with the combination of the most renowned pre-trained language models
BERT (Bidirectional Encoder Representations from Transformers). The results
shows the efficacy of pre-trained language models as we attained an accuracy of
0.93 for the sentiment analysis task and 0.90 for the fake news detection task.
| no_new_dataset | 0.876211 |
2304.00653 | Sudath Heiyanthuduwage PhD | Sudath Rohitha Heiyanthuduwage, Md Anisur Rahman and Md Zahidul Islam | Enhancing Cluster Quality of Numerical Datasets with Domain Ontology | 6 Pages, IEEE CSDE2022 Conference Paper | null | null | null | cs.DB cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Ontology-based clustering has gained attention in recent years due to the
potential benefits of ontology. Current ontology-based clustering approaches
have mainly been applied to reduce the dimensionality of attributes in text
document clustering. Reduction in dimensionality of attributes using ontology
helps to produce high quality clusters for a dataset. However, ontology-based
approaches in clustering numerical datasets have not been gained enough
attention. Moreover, some literature mentions that ontology-based clustering
can produce either high quality or low-quality clusters from a dataset.
Therefore, in this paper we present a clustering approach that is based on
domain ontology to reduce the dimensionality of attributes in a numerical
dataset using domain ontology and to produce high quality clusters. For every
dataset, we produce three datasets using domain ontology. We then cluster these
datasets using a genetic algorithm-based clustering technique called
GenClust++. The clusters of each dataset are evaluated in terms of Sum of
Squared-Error (SSE). We use six numerical datasets to evaluate the performance
of our ontology-based approach. The experimental results of our approach
indicate that cluster quality gradually improves from lower to the higher
levels of a domain ontology.
| [
{
"version": "v1",
"created": "Sun, 2 Apr 2023 23:40:17 GMT"
}
] | 2023-04-04T00:00:00 | [
[
"Heiyanthuduwage",
"Sudath Rohitha",
""
],
[
"Rahman",
"Md Anisur",
""
],
[
"Islam",
"Md Zahidul",
""
]
] | TITLE: Enhancing Cluster Quality of Numerical Datasets with Domain Ontology
ABSTRACT: Ontology-based clustering has gained attention in recent years due to the
potential benefits of ontology. Current ontology-based clustering approaches
have mainly been applied to reduce the dimensionality of attributes in text
document clustering. Reduction in dimensionality of attributes using ontology
helps to produce high quality clusters for a dataset. However, ontology-based
approaches in clustering numerical datasets have not been gained enough
attention. Moreover, some literature mentions that ontology-based clustering
can produce either high quality or low-quality clusters from a dataset.
Therefore, in this paper we present a clustering approach that is based on
domain ontology to reduce the dimensionality of attributes in a numerical
dataset using domain ontology and to produce high quality clusters. For every
dataset, we produce three datasets using domain ontology. We then cluster these
datasets using a genetic algorithm-based clustering technique called
GenClust++. The clusters of each dataset are evaluated in terms of Sum of
Squared-Error (SSE). We use six numerical datasets to evaluate the performance
of our ontology-based approach. The experimental results of our approach
indicate that cluster quality gradually improves from lower to the higher
levels of a domain ontology.
| no_new_dataset | 0.954009 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.