id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
| label
stringclasses 2
values | prob
float64 0.5
0.73
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2006.11988 | Joseph Paul Cohen | Joseph Paul Cohen and Paul Morrison and Lan Dao and Karsten Roth and
Tim Q Duong and Marzyeh Ghassemi | COVID-19 Image Data Collection: Prospective Predictions Are the Future | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://melba-journal.org. Code for baseline
experiments can be found here: https://github.com/mlmed/covid-baselines | null | 10.59275/j.melba.2020-48g7 | null | q-bio.QM cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by/4.0/ | Across the world's coronavirus disease 2019 (COVID-19) hot spots, the need to
streamline patient diagnosis and management has become more pressing than ever.
As one of the main imaging tools, chest X-rays (CXRs) are common, fast,
non-invasive, relatively cheap, and potentially bedside to monitor the
progression of the disease. This paper describes the first public COVID-19
image data collection as well as a preliminary exploration of possible use
cases for the data. This dataset currently contains hundreds of frontal view
X-rays and is the largest public resource for COVID-19 image and prognostic
data, making it a necessary resource to develop and evaluate tools to aid in
the treatment of COVID-19. It was manually aggregated from publication figures
as well as various web based repositories into a machine learning (ML) friendly
format with accompanying dataloader code. We collected frontal and lateral view
imagery and metadata such as the time since first symptoms, intensive care unit
(ICU) status, survival status, intubation status, or hospital location. We
present multiple possible use cases for the data such as predicting the need
for the ICU, predicting patient survival, and understanding a patient's
trajectory during treatment. Data can be accessed here:
https://github.com/ieee8023/covid-chestxray-dataset
| [
{
"version": "v1",
"created": "Mon, 22 Jun 2020 03:20:36 GMT"
},
{
"version": "v2",
"created": "Thu, 1 Oct 2020 20:31:04 GMT"
},
{
"version": "v3",
"created": "Mon, 14 Dec 2020 18:52:43 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Cohen",
"Joseph Paul",
""
],
[
"Morrison",
"Paul",
""
],
[
"Dao",
"Lan",
""
],
[
"Roth",
"Karsten",
""
],
[
"Duong",
"Tim Q",
""
],
[
"Ghassemi",
"Marzyeh",
""
]
] | TITLE: COVID-19 Image Data Collection: Prospective Predictions Are the Future
ABSTRACT: Across the world's coronavirus disease 2019 (COVID-19) hot spots, the need to
streamline patient diagnosis and management has become more pressing than ever.
As one of the main imaging tools, chest X-rays (CXRs) are common, fast,
non-invasive, relatively cheap, and potentially bedside to monitor the
progression of the disease. This paper describes the first public COVID-19
image data collection as well as a preliminary exploration of possible use
cases for the data. This dataset currently contains hundreds of frontal view
X-rays and is the largest public resource for COVID-19 image and prognostic
data, making it a necessary resource to develop and evaluate tools to aid in
the treatment of COVID-19. It was manually aggregated from publication figures
as well as various web based repositories into a machine learning (ML) friendly
format with accompanying dataloader code. We collected frontal and lateral view
imagery and metadata such as the time since first symptoms, intensive care unit
(ICU) status, survival status, intubation status, or hospital location. We
present multiple possible use cases for the data such as predicting the need
for the ICU, predicting patient survival, and understanding a patient's
trajectory during treatment. Data can be accessed here:
https://github.com/ieee8023/covid-chestxray-dataset
| no_new_dataset | 0.656631 |
2007.01441 | Nalini Singh | Nalini M. Singh, Juan Eugenio Iglesias, Elfar Adalsteinsson, Adrian V.
Dalca, Polina Golland | Joint Frequency and Image Space Learning for MRI Reconstruction and
Analysis | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://www.melba-journal.org/papers/2022:018.html | null | 10.59275/j.melba.2022-16cc | null | cs.CV cs.LG eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose neural network layers that explicitly combine frequency and image
feature representations and show that they can be used as a versatile building
block for reconstruction from frequency space data. Our work is motivated by
the challenges arising in MRI acquisition where the signal is a corrupted
Fourier transform of the desired image. The proposed joint learning schemes
enable both correction of artifacts native to the frequency space and
manipulation of image space representations to reconstruct coherent image
structures at every layer of the network. This is in contrast to most current
deep learning approaches for image reconstruction that treat frequency and
image space features separately and often operate exclusively in one of the two
spaces. We demonstrate the advantages of joint convolutional learning for a
variety of tasks, including motion correction, denoising, reconstruction from
undersampled acquisitions, and combined undersampling and motion correction on
simulated and real world multicoil MRI data. The joint models produce
consistently high quality output images across all tasks and datasets. When
integrated into a state of the art unrolled optimization network with
physics-inspired data consistency constraints for undersampled reconstruction,
the proposed architectures significantly improve the optimization landscape,
which yields an order of magnitude reduction of training time. This result
suggests that joint representations are particularly well suited for MRI
signals in deep learning networks. Our code and pretrained models are publicly
available at https://github.com/nalinimsingh/interlacer.
| [
{
"version": "v1",
"created": "Thu, 2 Jul 2020 23:54:46 GMT"
},
{
"version": "v2",
"created": "Thu, 26 Nov 2020 06:22:39 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Feb 2022 15:58:29 GMT"
},
{
"version": "v4",
"created": "Sat, 18 Jun 2022 02:22:11 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Singh",
"Nalini M.",
""
],
[
"Iglesias",
"Juan Eugenio",
""
],
[
"Adalsteinsson",
"Elfar",
""
],
[
"Dalca",
"Adrian V.",
""
],
[
"Golland",
"Polina",
""
]
] | TITLE: Joint Frequency and Image Space Learning for MRI Reconstruction and
Analysis
ABSTRACT: We propose neural network layers that explicitly combine frequency and image
feature representations and show that they can be used as a versatile building
block for reconstruction from frequency space data. Our work is motivated by
the challenges arising in MRI acquisition where the signal is a corrupted
Fourier transform of the desired image. The proposed joint learning schemes
enable both correction of artifacts native to the frequency space and
manipulation of image space representations to reconstruct coherent image
structures at every layer of the network. This is in contrast to most current
deep learning approaches for image reconstruction that treat frequency and
image space features separately and often operate exclusively in one of the two
spaces. We demonstrate the advantages of joint convolutional learning for a
variety of tasks, including motion correction, denoising, reconstruction from
undersampled acquisitions, and combined undersampling and motion correction on
simulated and real world multicoil MRI data. The joint models produce
consistently high quality output images across all tasks and datasets. When
integrated into a state of the art unrolled optimization network with
physics-inspired data consistency constraints for undersampled reconstruction,
the proposed architectures significantly improve the optimization landscape,
which yields an order of magnitude reduction of training time. This result
suggests that joint representations are particularly well suited for MRI
signals in deep learning networks. Our code and pretrained models are publicly
available at https://github.com/nalinimsingh/interlacer.
| no_new_dataset | 0.7118 |
2107.09405 | Yoni Schirris | Yoni Schirris, Efstratios Gavves, Iris Nederlof, Hugo Mark Horlings,
Jonas Teuwen | DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD
classification directly from H&E whole-slide images in colorectal and breast
cancer | Main paper: 14 pages, 2 tables, 1 algorithm, 3 figures. Supplementary
material: 3 pages | Medical Image Analysis Volume 79, July 2022, 102464 | 10.1016/j.media.2022.102464 | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a Deep learning-based weak label learning method for analyzing
whole slide images (WSIs) of Hematoxylin and Eosin (H&E) stained tumor tissue
not requiring pixel-level or tile-level annotations using Self-supervised
pre-training and heterogeneity-aware deep Multiple Instance LEarning
(DeepSMILE). We apply DeepSMILE to the task of Homologous recombination
deficiency (HRD) and microsatellite instability (MSI) prediction. We utilize
contrastive self-supervised learning to pre-train a feature extractor on
histopathology tiles of cancer tissue. Additionally, we use variability-aware
deep multiple instance learning to learn the tile feature aggregation function
while modeling tumor heterogeneity. For MSI prediction in a tumor-annotated and
color normalized subset of TCGA-CRC (n=360 patients), contrastive
self-supervised learning improves the tile supervision baseline from 0.77 to
0.87 AUROC, on par with our proposed DeepSMILE method. On TCGA-BC (n=1041
patients) without any manual annotations, DeepSMILE improves HRD classification
performance from 0.77 to 0.81 AUROC compared to tile supervision with either a
self-supervised or ImageNet pre-trained feature extractor. Our proposed methods
reach the baseline performance using only 40% of the labeled data on both
datasets. These improvements suggest we can use standard self-supervised
learning techniques combined with multiple instance learning in the
histopathology domain to improve genomic label classification performance with
fewer labeled data.
| [
{
"version": "v1",
"created": "Tue, 20 Jul 2021 11:00:16 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jul 2021 11:55:58 GMT"
},
{
"version": "v3",
"created": "Wed, 28 Jun 2023 13:52:29 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Schirris",
"Yoni",
""
],
[
"Gavves",
"Efstratios",
""
],
[
"Nederlof",
"Iris",
""
],
[
"Horlings",
"Hugo Mark",
""
],
[
"Teuwen",
"Jonas",
""
]
] | TITLE: DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD
classification directly from H&E whole-slide images in colorectal and breast
cancer
ABSTRACT: We propose a Deep learning-based weak label learning method for analyzing
whole slide images (WSIs) of Hematoxylin and Eosin (H&E) stained tumor tissue
not requiring pixel-level or tile-level annotations using Self-supervised
pre-training and heterogeneity-aware deep Multiple Instance LEarning
(DeepSMILE). We apply DeepSMILE to the task of Homologous recombination
deficiency (HRD) and microsatellite instability (MSI) prediction. We utilize
contrastive self-supervised learning to pre-train a feature extractor on
histopathology tiles of cancer tissue. Additionally, we use variability-aware
deep multiple instance learning to learn the tile feature aggregation function
while modeling tumor heterogeneity. For MSI prediction in a tumor-annotated and
color normalized subset of TCGA-CRC (n=360 patients), contrastive
self-supervised learning improves the tile supervision baseline from 0.77 to
0.87 AUROC, on par with our proposed DeepSMILE method. On TCGA-BC (n=1041
patients) without any manual annotations, DeepSMILE improves HRD classification
performance from 0.77 to 0.81 AUROC compared to tile supervision with either a
self-supervised or ImageNet pre-trained feature extractor. Our proposed methods
reach the baseline performance using only 40% of the labeled data on both
datasets. These improvements suggest we can use standard self-supervised
learning techniques combined with multiple instance learning in the
histopathology domain to improve genomic label classification performance with
fewer labeled data.
| no_new_dataset | 0.712026 |
2109.02342 | Seung Su Yoon | Seung Su Yoon, Elisabeth Preuhs, Michaela Schmidt, Christoph Forman,
Teodora Chitiboi, Puneet Sharma, Juliano Lara Fernandes, Christoph Tillmanns,
Jens Wetzl, Andreas Maier | Automated Cardiac Resting Phase Detection Targeted on the Right Coronary
Artery | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://melba-journal.org/2023:001 | Machine.Learning.for.Biomedical.Imaging. 2 (2023) | 10.59275/j.melba.2023-afe2 | null | eess.IV cs.CV physics.med-ph | http://creativecommons.org/licenses/by/4.0/ | Static cardiac imaging such as late gadolinium enhancement, mapping, or 3-D
coronary angiography require prior information, e.g., the phase during a
cardiac cycle with least motion, called resting phase (RP). The purpose of this
work is to propose a fully automated framework that allows the detection of the
right coronary artery (RCA) RP within CINE series. The proposed prototype
system consists of three main steps. First, the localization of the regions of
interest (ROI) is performed. Second, the cropped ROI series are taken for
tracking motions over all time points. Third, the output motion values are used
to classify RPs. In this work, we focused on the detection of the area with the
outer edge of the cross-section of the RCA as our target. The proposed
framework was evaluated on 102 clinically acquired dataset at 1.5T and 3T. The
automatically classified RPs were compared with the reference RPs annotated
manually by a expert for testing the robustness and feasibility of the
framework. The predicted RCA RPs showed high agreement with the experts
annotated RPs with 92.7% accuracy, 90.5% sensitivity and 95.0% specificity for
the unseen study dataset. The mean absolute difference of the start and end RP
was 13.6 $\pm$ 18.6 ms for the validation study dataset (n=102). In this work,
automated RP detection has been introduced by the proposed framework and
demonstrated feasibility, robustness, and applicability for static imaging
acquisitions.
| [
{
"version": "v1",
"created": "Mon, 6 Sep 2021 10:29:52 GMT"
},
{
"version": "v2",
"created": "Wed, 1 Feb 2023 00:15:36 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yoon",
"Seung Su",
""
],
[
"Preuhs",
"Elisabeth",
""
],
[
"Schmidt",
"Michaela",
""
],
[
"Forman",
"Christoph",
""
],
[
"Chitiboi",
"Teodora",
""
],
[
"Sharma",
"Puneet",
""
],
[
"Fernandes",
"Juliano Lara",
""
],
[
"Tillmanns",
"Christoph",
""
],
[
"Wetzl",
"Jens",
""
],
[
"Maier",
"Andreas",
""
]
] | TITLE: Automated Cardiac Resting Phase Detection Targeted on the Right Coronary
Artery
ABSTRACT: Static cardiac imaging such as late gadolinium enhancement, mapping, or 3-D
coronary angiography require prior information, e.g., the phase during a
cardiac cycle with least motion, called resting phase (RP). The purpose of this
work is to propose a fully automated framework that allows the detection of the
right coronary artery (RCA) RP within CINE series. The proposed prototype
system consists of three main steps. First, the localization of the regions of
interest (ROI) is performed. Second, the cropped ROI series are taken for
tracking motions over all time points. Third, the output motion values are used
to classify RPs. In this work, we focused on the detection of the area with the
outer edge of the cross-section of the RCA as our target. The proposed
framework was evaluated on 102 clinically acquired dataset at 1.5T and 3T. The
automatically classified RPs were compared with the reference RPs annotated
manually by a expert for testing the robustness and feasibility of the
framework. The predicted RCA RPs showed high agreement with the experts
annotated RPs with 92.7% accuracy, 90.5% sensitivity and 95.0% specificity for
the unseen study dataset. The mean absolute difference of the start and end RP
was 13.6 $\pm$ 18.6 ms for the validation study dataset (n=102). In this work,
automated RP detection has been introduced by the proposed framework and
demonstrated feasibility, robustness, and applicability for static imaging
acquisitions.
| no_new_dataset | 0.702734 |
2110.13185 | Abdullah-Al-Zubaer Imran | Ayaan Haque, Abdullah-Al-Zubaer Imran, Adam Wang, Demetri Terzopoulos | Generalized Multi-Task Learning from Substantially Unlabeled
Multi-Source Medical Image Data | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://www.melba-journal.org/ | null | 10.59275/j.melba.2021-d8a3 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Deep learning-based models, when trained in a fully-supervised manner, can be
effective in performing complex image analysis tasks, although contingent upon
the availability of large labeled datasets. Especially in the medical imaging
domain, however, expert image annotation is expensive, time-consuming, and
prone to variability. Semi-supervised learning from limited quantities of
labeled data has shown promise as an alternative. Maximizing knowledge gains
from copious unlabeled data benefits semi-supervised learning models. Moreover,
learning multiple tasks within the same model further improves its
generalizability. We propose MultiMix, a new multi-task learning model that
jointly learns disease classification and anatomical segmentation in a
semi-supervised manner, while preserving explainability through a novel
saliency bridge between the two tasks. Our experiments with varying quantities
of multi-source labeled data in the training sets confirm the effectiveness of
MultiMix in the simultaneous classification of pneumonia and segmentation of
the lungs in chest X-ray images. Moreover, both in-domain and cross-domain
evaluations across these tasks further showcase the potential of our model to
adapt to challenging generalization scenarios.
| [
{
"version": "v1",
"created": "Mon, 25 Oct 2021 18:09:19 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Haque",
"Ayaan",
""
],
[
"Imran",
"Abdullah-Al-Zubaer",
""
],
[
"Wang",
"Adam",
""
],
[
"Terzopoulos",
"Demetri",
""
]
] | TITLE: Generalized Multi-Task Learning from Substantially Unlabeled
Multi-Source Medical Image Data
ABSTRACT: Deep learning-based models, when trained in a fully-supervised manner, can be
effective in performing complex image analysis tasks, although contingent upon
the availability of large labeled datasets. Especially in the medical imaging
domain, however, expert image annotation is expensive, time-consuming, and
prone to variability. Semi-supervised learning from limited quantities of
labeled data has shown promise as an alternative. Maximizing knowledge gains
from copious unlabeled data benefits semi-supervised learning models. Moreover,
learning multiple tasks within the same model further improves its
generalizability. We propose MultiMix, a new multi-task learning model that
jointly learns disease classification and anatomical segmentation in a
semi-supervised manner, while preserving explainability through a novel
saliency bridge between the two tasks. Our experiments with varying quantities
of multi-source labeled data in the training sets confirm the effectiveness of
MultiMix in the simultaneous classification of pneumonia and segmentation of
the lungs in chest X-ray images. Moreover, both in-domain and cross-domain
evaluations across these tasks further showcase the potential of our model to
adapt to challenging generalization scenarios.
| no_new_dataset | 0.709239 |
2112.10074 | Raghav Mehta | Raghav Mehta, Angelos Filos, Ujjwal Baid, Chiharu Sako, Richard
McKinley, Michael Rebsamen, Katrin Datwyler, Raphael Meier, Piotr Radojewski,
Gowtham Krishnan Murugesan, Sahil Nalawade, Chandan Ganesh, Ben Wagner, Fang
F. Yu, Baowei Fei, Ananth J. Madhuranthakam, Joseph A. Maldjian, Laura Daza,
Catalina Gomez, Pablo Arbelaez, Chengliang Dai, Shuo Wang, Hadrien Reynaud,
Yuan-han Mo, Elsa Angelini, Yike Guo, Wenjia Bai, Subhashis Banerjee, Lin-min
Pei, Murat AK, Sarahi Rosas-Gonzalez, Ilyess Zemmoura, Clovis Tauber, Minh H.
Vu, Tufve Nyholm, Tommy Lofstedt, Laura Mora Ballestar, Veronica Vilaplana,
Hugh McHugh, Gonzalo Maso Talou, Alan Wang, Jay Patel, Ken Chang, Katharina
Hoebel, Mishka Gidwani, Nishanth Arun, Sharut Gupta, Mehak Aggarwal, Praveer
Singh, Elizabeth R. Gerstner, Jayashree Kalpathy-Cramer, Nicolas Boutry,
Alexis Huard, Lasitha Vidyaratne, Md Monibor Rahman, Khan M. Iftekharuddin,
Joseph Chazalon, Elodie Puybareau, Guillaume Tochon, Jun Ma, Mariano Cabezas,
Xavier Llado, Arnau Oliver, Liliana Valencia, Sergi Valverde, Mehdi Amian,
Mohammadreza Soltaninejad, Andriy Myronenko, Ali Hatamizadeh, Xue Feng, Quan
Dou, Nicholas Tustison, Craig Meyer, Nisarg A. Shah, Sanjay Talbar,
Marc-Andre Weber, Abhishek Mahajan, Andras Jakab, Roland Wiest, Hassan M.
Fathallah-Shaykh, Arash Nazeri, Mikhail Milchenko1, Daniel Marcus, Aikaterini
Kotrotsou, Rivka Colen, John Freymann, Justin Kirby, Christos Davatzikos,
Bjoern Menze, Spyridon Bakas, Yarin Gal, Tal Arbel | QU-BraTS: MICCAI BraTS 2020 Challenge on Quantifying Uncertainty in
Brain Tumor Segmentation - Analysis of Ranking Scores and Benchmarking
Results | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA):
https://www.melba-journal.org/papers/2022:026.html | Machine.Learning.for.Biomedical.Imaging. 1 (2022) | 10.59275/j.melba.2022-354b | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Deep learning (DL) models have provided state-of-the-art performance in
various medical imaging benchmarking challenges, including the Brain Tumor
Segmentation (BraTS) challenges. However, the task of focal pathology
multi-compartment segmentation (e.g., tumor and lesion sub-regions) is
particularly challenging, and potential errors hinder translating DL models
into clinical workflows. Quantifying the reliability of DL model predictions in
the form of uncertainties could enable clinical review of the most uncertain
regions, thereby building trust and paving the way toward clinical translation.
Several uncertainty estimation methods have recently been introduced for DL
medical image segmentation tasks. Developing scores to evaluate and compare the
performance of uncertainty measures will assist the end-user in making more
informed decisions. In this study, we explore and evaluate a score developed
during the BraTS 2019 and BraTS 2020 task on uncertainty quantification
(QU-BraTS) and designed to assess and rank uncertainty estimates for brain
tumor multi-compartment segmentation. This score (1) rewards uncertainty
estimates that produce high confidence in correct assertions and those that
assign low confidence levels at incorrect assertions, and (2) penalizes
uncertainty measures that lead to a higher percentage of under-confident
correct assertions. We further benchmark the segmentation uncertainties
generated by 14 independent participating teams of QU-BraTS 2020, all of which
also participated in the main BraTS segmentation task. Overall, our findings
confirm the importance and complementary value that uncertainty estimates
provide to segmentation algorithms, highlighting the need for uncertainty
quantification in medical image analyses. Finally, in favor of transparency and
reproducibility, our evaluation code is made publicly available at:
https://github.com/RagMeh11/QU-BraTS.
| [
{
"version": "v1",
"created": "Sun, 19 Dec 2021 06:45:55 GMT"
},
{
"version": "v2",
"created": "Tue, 23 Aug 2022 16:24:19 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Mehta",
"Raghav",
""
],
[
"Filos",
"Angelos",
""
],
[
"Baid",
"Ujjwal",
""
],
[
"Sako",
"Chiharu",
""
],
[
"McKinley",
"Richard",
""
],
[
"Rebsamen",
"Michael",
""
],
[
"Datwyler",
"Katrin",
""
],
[
"Meier",
"Raphael",
""
],
[
"Radojewski",
"Piotr",
""
],
[
"Murugesan",
"Gowtham Krishnan",
""
],
[
"Nalawade",
"Sahil",
""
],
[
"Ganesh",
"Chandan",
""
],
[
"Wagner",
"Ben",
""
],
[
"Yu",
"Fang F.",
""
],
[
"Fei",
"Baowei",
""
],
[
"Madhuranthakam",
"Ananth J.",
""
],
[
"Maldjian",
"Joseph A.",
""
],
[
"Daza",
"Laura",
""
],
[
"Gomez",
"Catalina",
""
],
[
"Arbelaez",
"Pablo",
""
],
[
"Dai",
"Chengliang",
""
],
[
"Wang",
"Shuo",
""
],
[
"Reynaud",
"Hadrien",
""
],
[
"Mo",
"Yuan-han",
""
],
[
"Angelini",
"Elsa",
""
],
[
"Guo",
"Yike",
""
],
[
"Bai",
"Wenjia",
""
],
[
"Banerjee",
"Subhashis",
""
],
[
"Pei",
"Lin-min",
""
],
[
"AK",
"Murat",
""
],
[
"Rosas-Gonzalez",
"Sarahi",
""
],
[
"Zemmoura",
"Ilyess",
""
],
[
"Tauber",
"Clovis",
""
],
[
"Vu",
"Minh H.",
""
],
[
"Nyholm",
"Tufve",
""
],
[
"Lofstedt",
"Tommy",
""
],
[
"Ballestar",
"Laura Mora",
""
],
[
"Vilaplana",
"Veronica",
""
],
[
"McHugh",
"Hugh",
""
],
[
"Talou",
"Gonzalo Maso",
""
],
[
"Wang",
"Alan",
""
],
[
"Patel",
"Jay",
""
],
[
"Chang",
"Ken",
""
],
[
"Hoebel",
"Katharina",
""
],
[
"Gidwani",
"Mishka",
""
],
[
"Arun",
"Nishanth",
""
],
[
"Gupta",
"Sharut",
""
],
[
"Aggarwal",
"Mehak",
""
],
[
"Singh",
"Praveer",
""
],
[
"Gerstner",
"Elizabeth R.",
""
],
[
"Kalpathy-Cramer",
"Jayashree",
""
],
[
"Boutry",
"Nicolas",
""
],
[
"Huard",
"Alexis",
""
],
[
"Vidyaratne",
"Lasitha",
""
],
[
"Rahman",
"Md Monibor",
""
],
[
"Iftekharuddin",
"Khan M.",
""
],
[
"Chazalon",
"Joseph",
""
],
[
"Puybareau",
"Elodie",
""
],
[
"Tochon",
"Guillaume",
""
],
[
"Ma",
"Jun",
""
],
[
"Cabezas",
"Mariano",
""
],
[
"Llado",
"Xavier",
""
],
[
"Oliver",
"Arnau",
""
],
[
"Valencia",
"Liliana",
""
],
[
"Valverde",
"Sergi",
""
],
[
"Amian",
"Mehdi",
""
],
[
"Soltaninejad",
"Mohammadreza",
""
],
[
"Myronenko",
"Andriy",
""
],
[
"Hatamizadeh",
"Ali",
""
],
[
"Feng",
"Xue",
""
],
[
"Dou",
"Quan",
""
],
[
"Tustison",
"Nicholas",
""
],
[
"Meyer",
"Craig",
""
],
[
"Shah",
"Nisarg A.",
""
],
[
"Talbar",
"Sanjay",
""
],
[
"Weber",
"Marc-Andre",
""
],
[
"Mahajan",
"Abhishek",
""
],
[
"Jakab",
"Andras",
""
],
[
"Wiest",
"Roland",
""
],
[
"Fathallah-Shaykh",
"Hassan M.",
""
],
[
"Nazeri",
"Arash",
""
],
[
"Milchenko1",
"Mikhail",
""
],
[
"Marcus",
"Daniel",
""
],
[
"Kotrotsou",
"Aikaterini",
""
],
[
"Colen",
"Rivka",
""
],
[
"Freymann",
"John",
""
],
[
"Kirby",
"Justin",
""
],
[
"Davatzikos",
"Christos",
""
],
[
"Menze",
"Bjoern",
""
],
[
"Bakas",
"Spyridon",
""
],
[
"Gal",
"Yarin",
""
],
[
"Arbel",
"Tal",
""
]
] | TITLE: QU-BraTS: MICCAI BraTS 2020 Challenge on Quantifying Uncertainty in
Brain Tumor Segmentation - Analysis of Ranking Scores and Benchmarking
Results
ABSTRACT: Deep learning (DL) models have provided state-of-the-art performance in
various medical imaging benchmarking challenges, including the Brain Tumor
Segmentation (BraTS) challenges. However, the task of focal pathology
multi-compartment segmentation (e.g., tumor and lesion sub-regions) is
particularly challenging, and potential errors hinder translating DL models
into clinical workflows. Quantifying the reliability of DL model predictions in
the form of uncertainties could enable clinical review of the most uncertain
regions, thereby building trust and paving the way toward clinical translation.
Several uncertainty estimation methods have recently been introduced for DL
medical image segmentation tasks. Developing scores to evaluate and compare the
performance of uncertainty measures will assist the end-user in making more
informed decisions. In this study, we explore and evaluate a score developed
during the BraTS 2019 and BraTS 2020 task on uncertainty quantification
(QU-BraTS) and designed to assess and rank uncertainty estimates for brain
tumor multi-compartment segmentation. This score (1) rewards uncertainty
estimates that produce high confidence in correct assertions and those that
assign low confidence levels at incorrect assertions, and (2) penalizes
uncertainty measures that lead to a higher percentage of under-confident
correct assertions. We further benchmark the segmentation uncertainties
generated by 14 independent participating teams of QU-BraTS 2020, all of which
also participated in the main BraTS segmentation task. Overall, our findings
confirm the importance and complementary value that uncertainty estimates
provide to segmentation algorithms, highlighting the need for uncertainty
quantification in medical image analyses. Finally, in favor of transparency and
reproducibility, our evaluation code is made publicly available at:
https://github.com/RagMeh11/QU-BraTS.
| no_new_dataset | 0.70978 |
2112.14869 | Dixian Zhu | Dixian Zhu, Yiming Ying and Tianbao Yang | Label Distributionally Robust Losses for Multi-class Classification:
Consistency, Robustness and Adaptivity | To appear in ICML2023; 37 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | We study a family of loss functions named label-distributionally robust (LDR)
losses for multi-class classification that are formulated from distributionally
robust optimization (DRO) perspective, where the uncertainty in the given label
information are modeled and captured by taking the worse case of distributional
weights. The benefits of this perspective are several fold: (i) it provides a
unified framework to explain the classical cross-entropy (CE) loss and SVM loss
and their variants, (ii) it includes a special family corresponding to the
temperature-scaled CE loss, which is widely adopted but poorly understood;
(iii) it allows us to achieve adaptivity to the uncertainty degree of label
information at an instance level. Our contributions include: (1) we study both
consistency and robustness by establishing top-$k$ ($\forall k\geq 1$)
consistency of LDR losses for multi-class classification, and a negative result
that a top-$1$ consistent and symmetric robust loss cannot achieve top-$k$
consistency simultaneously for all $k\geq 2$; (2) we propose a new adaptive LDR
loss that automatically adapts the individualized temperature parameter to the
noise degree of class label of each instance; (3) we demonstrate stable and
competitive performance for the proposed adaptive LDR loss on 7 benchmark
datasets under 6 noisy label and 1 clean settings against 13 loss functions,
and on one real-world noisy dataset. The code is open-sourced at
\url{https://github.com/Optimization-AI/ICML2023_LDR}.
| [
{
"version": "v1",
"created": "Thu, 30 Dec 2021 00:27:30 GMT"
},
{
"version": "v2",
"created": "Mon, 29 May 2023 04:27:33 GMT"
},
{
"version": "v3",
"created": "Tue, 6 Jun 2023 05:29:54 GMT"
},
{
"version": "v4",
"created": "Wed, 28 Jun 2023 04:53:43 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Zhu",
"Dixian",
""
],
[
"Ying",
"Yiming",
""
],
[
"Yang",
"Tianbao",
""
]
] | TITLE: Label Distributionally Robust Losses for Multi-class Classification:
Consistency, Robustness and Adaptivity
ABSTRACT: We study a family of loss functions named label-distributionally robust (LDR)
losses for multi-class classification that are formulated from distributionally
robust optimization (DRO) perspective, where the uncertainty in the given label
information are modeled and captured by taking the worse case of distributional
weights. The benefits of this perspective are several fold: (i) it provides a
unified framework to explain the classical cross-entropy (CE) loss and SVM loss
and their variants, (ii) it includes a special family corresponding to the
temperature-scaled CE loss, which is widely adopted but poorly understood;
(iii) it allows us to achieve adaptivity to the uncertainty degree of label
information at an instance level. Our contributions include: (1) we study both
consistency and robustness by establishing top-$k$ ($\forall k\geq 1$)
consistency of LDR losses for multi-class classification, and a negative result
that a top-$1$ consistent and symmetric robust loss cannot achieve top-$k$
consistency simultaneously for all $k\geq 2$; (2) we propose a new adaptive LDR
loss that automatically adapts the individualized temperature parameter to the
noise degree of class label of each instance; (3) we demonstrate stable and
competitive performance for the proposed adaptive LDR loss on 7 benchmark
datasets under 6 noisy label and 1 clean settings against 13 loss functions,
and on one real-world noisy dataset. The code is open-sourced at
\url{https://github.com/Optimization-AI/ICML2023_LDR}.
| no_new_dataset | 0.710057 |
2203.13129 | Laura Selicato | Nicoletta Del Buono, Flavia Esposito, Laura Selicato, Rafal Zdunek | Bi-level algorithm for optimizing hyperparameters in penalized
nonnegative matrix factorization | 27 pages, 25 Figures | Applied Mathematics and Computation 457 (2023) 128184 | 10.1016/j.amc.2023.128184 | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning approaches rely on hyperparameters that impact the algorithm's
performance and affect the knowledge extraction process from data. Recently,
Nonnegative Matrix Factorization (NMF) has attracted a growing interest as a
learning algorithm. This technique captures the latent information embedded in
large datasets while preserving feature properties. NMF can be formalized as a
penalized optimization task in which tuning the penalty hyperparameters is an
open issue. The current literature does not provide any general framework
addressing this task. This study proposes to express the penalty
hyperparameters problem in NMF in terms of a bi-level optimization. We design a
novel algorithm, named Alternating Bi-level (AltBi), which incorporates the
hyperparameters tuning procedure into the updates of NMF factors. Results of
the existence and convergence of numerical solutions, under appropriate
assumptions, are studied, and numerical experiments are provided.
| [
{
"version": "v1",
"created": "Thu, 24 Mar 2022 15:43:15 GMT"
},
{
"version": "v2",
"created": "Fri, 25 Mar 2022 16:27:02 GMT"
},
{
"version": "v3",
"created": "Wed, 28 Jun 2023 13:50:45 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Del Buono",
"Nicoletta",
""
],
[
"Esposito",
"Flavia",
""
],
[
"Selicato",
"Laura",
""
],
[
"Zdunek",
"Rafal",
""
]
] | TITLE: Bi-level algorithm for optimizing hyperparameters in penalized
nonnegative matrix factorization
ABSTRACT: Learning approaches rely on hyperparameters that impact the algorithm's
performance and affect the knowledge extraction process from data. Recently,
Nonnegative Matrix Factorization (NMF) has attracted a growing interest as a
learning algorithm. This technique captures the latent information embedded in
large datasets while preserving feature properties. NMF can be formalized as a
penalized optimization task in which tuning the penalty hyperparameters is an
open issue. The current literature does not provide any general framework
addressing this task. This study proposes to express the penalty
hyperparameters problem in NMF in terms of a bi-level optimization. We design a
novel algorithm, named Alternating Bi-level (AltBi), which incorporates the
hyperparameters tuning procedure into the updates of NMF factors. Results of
the existence and convergence of numerical solutions, under appropriate
assumptions, are studied, and numerical experiments are provided.
| no_new_dataset | 0.708616 |
2204.05731 | Tomer Meir | Tomer Meir, Rom Gutman, and Malka Gorfine | PyDTS: A Python Package for Discrete-Time Survival (Regularized)
Regression with Competing Risks | null | null | null | null | stat.ML cs.LG | http://creativecommons.org/licenses/by/4.0/ | Time-to-event analysis (survival analysis) is used when the response of
interest is the time until a pre-specified event occurs. Time-to-event data are
sometimes discrete either because time itself is discrete or due to grouping of
failure times into intervals or rounding off measurements. In addition, the
failure of an individual could be one of several distinct failure types, known
as competing risks (events). Most methods and software packages for survival
regression analysis assume that time is measured on a continuous scale. It is
well-known that naively applying standard continuous-time models with
discrete-time data may result in biased estimators of the discrete-time models.
The Python package PyDTS, for simulating, estimating and evaluating
semi-parametric competing-risks models for discrete-time survival data, is
introduced. The package implements a fast procedure that enables including
regularized regression methods, such as LASSO and elastic net, among others. A
simulation study showcases flexibility and accuracy of the package. The utility
of the package is demonstrated by analysing the Medical Information Mart for
Intensive Care (MIMIC) - IV dataset for prediction of hospitalization length of
stay.
| [
{
"version": "v1",
"created": "Tue, 12 Apr 2022 11:58:26 GMT"
},
{
"version": "v2",
"created": "Thu, 9 Jun 2022 17:51:34 GMT"
},
{
"version": "v3",
"created": "Thu, 4 Aug 2022 17:57:13 GMT"
},
{
"version": "v4",
"created": "Thu, 18 May 2023 16:11:17 GMT"
},
{
"version": "v5",
"created": "Tue, 27 Jun 2023 19:00:29 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Meir",
"Tomer",
""
],
[
"Gutman",
"Rom",
""
],
[
"Gorfine",
"Malka",
""
]
] | TITLE: PyDTS: A Python Package for Discrete-Time Survival (Regularized)
Regression with Competing Risks
ABSTRACT: Time-to-event analysis (survival analysis) is used when the response of
interest is the time until a pre-specified event occurs. Time-to-event data are
sometimes discrete either because time itself is discrete or due to grouping of
failure times into intervals or rounding off measurements. In addition, the
failure of an individual could be one of several distinct failure types, known
as competing risks (events). Most methods and software packages for survival
regression analysis assume that time is measured on a continuous scale. It is
well-known that naively applying standard continuous-time models with
discrete-time data may result in biased estimators of the discrete-time models.
The Python package PyDTS, for simulating, estimating and evaluating
semi-parametric competing-risks models for discrete-time survival data, is
introduced. The package implements a fast procedure that enables including
regularized regression methods, such as LASSO and elastic net, among others. A
simulation study showcases flexibility and accuracy of the package. The utility
of the package is demonstrated by analysing the Medical Information Mart for
Intensive Care (MIMIC) - IV dataset for prediction of hospitalization length of
stay.
| no_new_dataset | 0.70614 |
2204.07352 | Irene Balelli | Irene Balelli, Santiago Silva and Marco Lorenzi | A Differentially Private Probabilistic Framework for Modeling the
Variability Across Federated Datasets of Heterogeneous Multi-View
Observations | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://www.melba-journal.org | Journal of Machine Learning for Biomedical Imaging. 2022:012. pp
1-36. Special Issue: Information Processing in Medical Imaging (IPMI) 2021 | 10.59275/j.melba.2022-7175 | null | cs.LG cs.DC stat.AP stat.ME | http://creativecommons.org/licenses/by/4.0/ | We propose a novel federated learning paradigm to model data variability
among heterogeneous clients in multi-centric studies. Our method is expressed
through a hierarchical Bayesian latent variable model, where client-specific
parameters are assumed to be realization from a global distribution at the
master level, which is in turn estimated to account for data bias and
variability across clients. We show that our framework can be effectively
optimized through expectation maximization (EM) over latent master's
distribution and clients' parameters. We also introduce formal differential
privacy (DP) guarantees compatibly with our EM optimization scheme. We tested
our method on the analysis of multi-modal medical imaging data and clinical
scores from distributed clinical datasets of patients affected by Alzheimer's
disease. We demonstrate that our method is robust when data is distributed
either in iid and non-iid manners, even when local parameters perturbation is
included to provide DP guarantees. Moreover, the variability of data, views and
centers can be quantified in an interpretable manner, while guaranteeing
high-quality data reconstruction as compared to state-of-the-art autoencoding
models and federated learning schemes. The code is available at
https://gitlab.inria.fr/epione/federated-multi-views-ppca.
| [
{
"version": "v1",
"created": "Fri, 15 Apr 2022 07:20:47 GMT"
},
{
"version": "v2",
"created": "Tue, 26 Apr 2022 08:47:04 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Balelli",
"Irene",
""
],
[
"Silva",
"Santiago",
""
],
[
"Lorenzi",
"Marco",
""
]
] | TITLE: A Differentially Private Probabilistic Framework for Modeling the
Variability Across Federated Datasets of Heterogeneous Multi-View
Observations
ABSTRACT: We propose a novel federated learning paradigm to model data variability
among heterogeneous clients in multi-centric studies. Our method is expressed
through a hierarchical Bayesian latent variable model, where client-specific
parameters are assumed to be realization from a global distribution at the
master level, which is in turn estimated to account for data bias and
variability across clients. We show that our framework can be effectively
optimized through expectation maximization (EM) over latent master's
distribution and clients' parameters. We also introduce formal differential
privacy (DP) guarantees compatibly with our EM optimization scheme. We tested
our method on the analysis of multi-modal medical imaging data and clinical
scores from distributed clinical datasets of patients affected by Alzheimer's
disease. We demonstrate that our method is robust when data is distributed
either in iid and non-iid manners, even when local parameters perturbation is
included to provide DP guarantees. Moreover, the variability of data, views and
centers can be quantified in an interpretable manner, while guaranteeing
high-quality data reconstruction as compared to state-of-the-art autoencoding
models and federated learning schemes. The code is available at
https://gitlab.inria.fr/epione/federated-multi-views-ppca.
| no_new_dataset | 0.708862 |
2206.05442 | Iddo Drori | Iddo Drori, Sarah J. Zhang, Reece Shuttleworth, Sarah Zhang, Keith
Tyser, Zad Chin, Pedro Lantigua, Saisamrit Surbehera, Gregory Hunter, Derek
Austin, Leonard Tang, Yann Hicke, Sage Simhon, Sathwik Karnik, Darnell
Granberry, Madeleine Udell | From Human Days to Machine Seconds: Automatically Answering and
Generating Machine Learning Final Exams | 9 pages | null | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | A final exam in machine learning at a top institution such as MIT, Harvard,
or Cornell typically takes faculty days to write, and students hours to solve.
We demonstrate that large language models pass machine learning finals at a
human level, on finals available online after the models were trained, and
automatically generate new human-quality final exam questions in seconds.
Previous work has developed program synthesis and few-shot learning methods to
solve university-level problem set questions in mathematics and STEM courses.
In this work, we develop and compare methods that solve final exams, which
differ from problem sets in several ways: the questions are longer, have
multiple parts, are more complicated, and span a broader set of topics. We
curate a dataset and benchmark of questions from machine learning final exams
available online and code for answering these questions and generating new
questions. We show how to generate new questions from other questions and
course notes. For reproducibility and future research on this final exam
benchmark, we use automatic checkers for multiple-choice, numeric, and
questions with expression answers. We perform ablation studies comparing
zero-shot learning with few-shot learning and chain-of-thought prompting using
GPT-3, OPT, Codex, and ChatGPT across machine learning topics and find that
few-shot learning methods perform best. We highlight the transformative
potential of language models to streamline the writing and solution of
large-scale assessments, significantly reducing the workload from human days to
mere machine seconds. Our results suggest that rather than banning large
language models such as ChatGPT in class, instructors should teach students to
harness them by asking students meta-questions about correctness, completeness,
and originality of the responses generated, encouraging critical thinking in
academic studies.
| [
{
"version": "v1",
"created": "Sat, 11 Jun 2022 06:38:06 GMT"
},
{
"version": "v2",
"created": "Mon, 29 Aug 2022 23:56:52 GMT"
},
{
"version": "v3",
"created": "Mon, 19 Dec 2022 19:37:45 GMT"
},
{
"version": "v4",
"created": "Thu, 22 Dec 2022 18:59:36 GMT"
},
{
"version": "v5",
"created": "Fri, 23 Dec 2022 13:41:18 GMT"
},
{
"version": "v6",
"created": "Thu, 15 Jun 2023 03:32:23 GMT"
},
{
"version": "v7",
"created": "Wed, 28 Jun 2023 04:42:05 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Drori",
"Iddo",
""
],
[
"Zhang",
"Sarah J.",
""
],
[
"Shuttleworth",
"Reece",
""
],
[
"Zhang",
"Sarah",
""
],
[
"Tyser",
"Keith",
""
],
[
"Chin",
"Zad",
""
],
[
"Lantigua",
"Pedro",
""
],
[
"Surbehera",
"Saisamrit",
""
],
[
"Hunter",
"Gregory",
""
],
[
"Austin",
"Derek",
""
],
[
"Tang",
"Leonard",
""
],
[
"Hicke",
"Yann",
""
],
[
"Simhon",
"Sage",
""
],
[
"Karnik",
"Sathwik",
""
],
[
"Granberry",
"Darnell",
""
],
[
"Udell",
"Madeleine",
""
]
] | TITLE: From Human Days to Machine Seconds: Automatically Answering and
Generating Machine Learning Final Exams
ABSTRACT: A final exam in machine learning at a top institution such as MIT, Harvard,
or Cornell typically takes faculty days to write, and students hours to solve.
We demonstrate that large language models pass machine learning finals at a
human level, on finals available online after the models were trained, and
automatically generate new human-quality final exam questions in seconds.
Previous work has developed program synthesis and few-shot learning methods to
solve university-level problem set questions in mathematics and STEM courses.
In this work, we develop and compare methods that solve final exams, which
differ from problem sets in several ways: the questions are longer, have
multiple parts, are more complicated, and span a broader set of topics. We
curate a dataset and benchmark of questions from machine learning final exams
available online and code for answering these questions and generating new
questions. We show how to generate new questions from other questions and
course notes. For reproducibility and future research on this final exam
benchmark, we use automatic checkers for multiple-choice, numeric, and
questions with expression answers. We perform ablation studies comparing
zero-shot learning with few-shot learning and chain-of-thought prompting using
GPT-3, OPT, Codex, and ChatGPT across machine learning topics and find that
few-shot learning methods perform best. We highlight the transformative
potential of language models to streamline the writing and solution of
large-scale assessments, significantly reducing the workload from human days to
mere machine seconds. Our results suggest that rather than banning large
language models such as ChatGPT in class, instructors should teach students to
harness them by asking students meta-questions about correctness, completeness,
and originality of the responses generated, encouraging critical thinking in
academic studies.
| new_dataset | 0.714236 |
2206.12902 | Wendi Xu | Wendi Xu, Xianpeng Wang | ETO Meets Scheduling: Learning Key Knowledge from Single-Objective
Problems to Multi-Objective Problem | For improvement | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by/4.0/ | Evolutionary transfer optimization(ETO) serves as "a new frontier in
evolutionary computation research", which will avoid zero reuse of experience
and knowledge from solved problems in traditional evolutionary computation. In
scheduling applications via ETO, a highly competitive "meeting" framework
between them could be constituted towards both intelligent scheduling and green
scheduling, especially for carbon neutrality within the context of China. To
the best of our knowledge, our study on scheduling here, is the 1st work of ETO
for complex optimization when multiobjective problem "meets" single-objective
problems in combinatorial case (not multitasking optimization). More
specifically, key knowledge like positional building blocks clustered, could be
learned and transferred for permutation flow shop scheduling problem (PFSP).
Empirical studies on well-studied benchmarks validate relatively firm
effectiveness and great potential of our proposed ETO-PFSP framework.
| [
{
"version": "v1",
"created": "Sun, 26 Jun 2022 15:29:57 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 16:51:25 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Xu",
"Wendi",
""
],
[
"Wang",
"Xianpeng",
""
]
] | TITLE: ETO Meets Scheduling: Learning Key Knowledge from Single-Objective
Problems to Multi-Objective Problem
ABSTRACT: Evolutionary transfer optimization(ETO) serves as "a new frontier in
evolutionary computation research", which will avoid zero reuse of experience
and knowledge from solved problems in traditional evolutionary computation. In
scheduling applications via ETO, a highly competitive "meeting" framework
between them could be constituted towards both intelligent scheduling and green
scheduling, especially for carbon neutrality within the context of China. To
the best of our knowledge, our study on scheduling here, is the 1st work of ETO
for complex optimization when multiobjective problem "meets" single-objective
problems in combinatorial case (not multitasking optimization). More
specifically, key knowledge like positional building blocks clustered, could be
learned and transferred for permutation flow shop scheduling problem (PFSP).
Empirical studies on well-studied benchmarks validate relatively firm
effectiveness and great potential of our proposed ETO-PFSP framework.
| no_new_dataset | 0.708231 |
2206.12906 | Wendi Xu | Xu Wendi, Wang Xianpeng, Guo Qingxin, Song Xiangman, Zhao Ren, Zhao
Guodong, Yang Yang, Xu Te, He Dakuo | Towards KAB2S: Learning Key Knowledge from Single-Objective Problems to
Multi-Objective Problem | For improvement | null | null | null | cs.NE cs.AI | http://creativecommons.org/licenses/by/4.0/ | As "a new frontier in evolutionary computation research", evolutionary
transfer optimization(ETO) will overcome the traditional paradigm of zero reuse
of related experience and knowledge from solved past problems in researches of
evolutionary computation. In scheduling applications via ETO, a quite appealing
and highly competitive framework "meeting" between them could be formed for
both intelligent scheduling and green scheduling, especially for international
pledge of "carbon neutrality" from China. To the best of our knowledge, our
paper on scheduling here, serves as the 1st work of a class of ETO frameworks
when multiobjective optimization problem "meets" single-objective optimization
problems in discrete case (not multitasking optimization). More specifically,
key knowledge conveyed for industrial applications, like positional building
blocks with genetic algorithm based settings, could be used via the new core
transfer mechanism and learning techniques for permutation flow shop scheduling
problem(PFSP). Extensive studies on well-studied benchmarks validate firm
effectiveness and great universality of our proposed ETO-PFSP framework
empirically. Our investigations (1) enrich the ETO frameworks, (2) contribute
to the classical and fundamental theory of building block for genetic
algorithms and memetic algorithms, and (3) head towards the paradigm shift of
evolutionary scheduling via learning by proposal and practice of paradigm of
"knowledge and building-block based scheduling" (KAB2S) for "industrial
intelligence" in China.
| [
{
"version": "v1",
"created": "Sun, 26 Jun 2022 15:36:05 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 12:25:04 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Wendi",
"Xu",
""
],
[
"Xianpeng",
"Wang",
""
],
[
"Qingxin",
"Guo",
""
],
[
"Xiangman",
"Song",
""
],
[
"Ren",
"Zhao",
""
],
[
"Guodong",
"Zhao",
""
],
[
"Yang",
"Yang",
""
],
[
"Te",
"Xu",
""
],
[
"Dakuo",
"He",
""
]
] | TITLE: Towards KAB2S: Learning Key Knowledge from Single-Objective Problems to
Multi-Objective Problem
ABSTRACT: As "a new frontier in evolutionary computation research", evolutionary
transfer optimization(ETO) will overcome the traditional paradigm of zero reuse
of related experience and knowledge from solved past problems in researches of
evolutionary computation. In scheduling applications via ETO, a quite appealing
and highly competitive framework "meeting" between them could be formed for
both intelligent scheduling and green scheduling, especially for international
pledge of "carbon neutrality" from China. To the best of our knowledge, our
paper on scheduling here, serves as the 1st work of a class of ETO frameworks
when multiobjective optimization problem "meets" single-objective optimization
problems in discrete case (not multitasking optimization). More specifically,
key knowledge conveyed for industrial applications, like positional building
blocks with genetic algorithm based settings, could be used via the new core
transfer mechanism and learning techniques for permutation flow shop scheduling
problem(PFSP). Extensive studies on well-studied benchmarks validate firm
effectiveness and great universality of our proposed ETO-PFSP framework
empirically. Our investigations (1) enrich the ETO frameworks, (2) contribute
to the classical and fundamental theory of building block for genetic
algorithms and memetic algorithms, and (3) head towards the paradigm shift of
evolutionary scheduling via learning by proposal and practice of paradigm of
"knowledge and building-block based scheduling" (KAB2S) for "industrial
intelligence" in China.
| no_new_dataset | 0.708767 |
2208.05438 | Hongyang Du | Hongyang Du, Jiazhen Liu, Dusit Niyato, Jiawen Kang, Zehui Xiong,
Junshan Zhang, and Dong In Kim | Attention-aware Resource Allocation and QoE Analysis for Metaverse
xURLLC Services | Accepted by IEEE Journal on Selected Areas in Communications (IEEE
JSAC) | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | Metaverse encapsulates our expectations of the next-generation Internet,
while bringing new key performance indicators (KPIs). Although conventional
ultra-reliable and low-latency communications (URLLC) can satisfy objective
KPIs, it is difficult to provide a personalized immersive experience that is a
distinctive feature of the Metaverse. Since the quality of experience (QoE) can
be regarded as a comprehensive KPI, the URLLC is evolved towards the next
generation URLLC (xURLLC) with a personalized resource allocation scheme to
achieve higher QoE. To deploy Metaverse xURLLC services, we study the
interaction between the Metaverse service provider (MSP) and the network
infrastructure provider (InP), and provide an optimal contract design
framework. Specifically, the utility of the MSP, defined as a function of
Metaverse users' QoE, is to be maximized, while ensuring the incentives of the
InP. To model the QoE mathematically, we propose a novel metric named
Meta-Immersion that incorporates both the objective KPIs and subjective
feelings of Metaverse users. Furthermore, we develop an attention-aware
rendering capacity allocation scheme to improve QoE in xURLLC. Using a
user-object-attention level dataset, we validate that the xURLLC can achieve an
average of 20.1% QoE improvement compared to the conventional URLLC with a
uniform resource allocation scheme. The code for this paper is available at
https://github.com/HongyangDu/AttentionQoE
| [
{
"version": "v1",
"created": "Wed, 10 Aug 2022 16:51:27 GMT"
},
{
"version": "v2",
"created": "Thu, 11 Aug 2022 11:55:07 GMT"
},
{
"version": "v3",
"created": "Thu, 8 Sep 2022 02:55:33 GMT"
},
{
"version": "v4",
"created": "Thu, 2 Feb 2023 06:43:35 GMT"
},
{
"version": "v5",
"created": "Mon, 27 Mar 2023 02:20:48 GMT"
},
{
"version": "v6",
"created": "Wed, 28 Jun 2023 11:38:11 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Du",
"Hongyang",
""
],
[
"Liu",
"Jiazhen",
""
],
[
"Niyato",
"Dusit",
""
],
[
"Kang",
"Jiawen",
""
],
[
"Xiong",
"Zehui",
""
],
[
"Zhang",
"Junshan",
""
],
[
"Kim",
"Dong In",
""
]
] | TITLE: Attention-aware Resource Allocation and QoE Analysis for Metaverse
xURLLC Services
ABSTRACT: Metaverse encapsulates our expectations of the next-generation Internet,
while bringing new key performance indicators (KPIs). Although conventional
ultra-reliable and low-latency communications (URLLC) can satisfy objective
KPIs, it is difficult to provide a personalized immersive experience that is a
distinctive feature of the Metaverse. Since the quality of experience (QoE) can
be regarded as a comprehensive KPI, the URLLC is evolved towards the next
generation URLLC (xURLLC) with a personalized resource allocation scheme to
achieve higher QoE. To deploy Metaverse xURLLC services, we study the
interaction between the Metaverse service provider (MSP) and the network
infrastructure provider (InP), and provide an optimal contract design
framework. Specifically, the utility of the MSP, defined as a function of
Metaverse users' QoE, is to be maximized, while ensuring the incentives of the
InP. To model the QoE mathematically, we propose a novel metric named
Meta-Immersion that incorporates both the objective KPIs and subjective
feelings of Metaverse users. Furthermore, we develop an attention-aware
rendering capacity allocation scheme to improve QoE in xURLLC. Using a
user-object-attention level dataset, we validate that the xURLLC can achieve an
average of 20.1% QoE improvement compared to the conventional URLLC with a
uniform resource allocation scheme. The code for this paper is available at
https://github.com/HongyangDu/AttentionQoE
| no_new_dataset | 0.706792 |
2211.07730 | Zifeng Wang | Zifeng Wang, Zizhao Zhang, Jacob Devlin, Chen-Yu Lee, Guolong Su, Hao
Zhang, Jennifer Dy, Vincent Perot, Tomas Pfister | QueryForm: A Simple Zero-shot Form Entity Query Framework | Accepted to Findings of ACL 2023 | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Zero-shot transfer learning for document understanding is a crucial yet
under-investigated scenario to help reduce the high cost involved in annotating
document entities. We present a novel query-based framework, QueryForm, that
extracts entity values from form-like documents in a zero-shot fashion.
QueryForm contains a dual prompting mechanism that composes both the document
schema and a specific entity type into a query, which is used to prompt a
Transformer model to perform a single entity extraction task. Furthermore, we
propose to leverage large-scale query-entity pairs generated from form-like
webpages with weak HTML annotations to pre-train QueryForm. By unifying
pre-training and fine-tuning into the same query-based framework, QueryForm
enables models to learn from structured documents containing various entities
and layouts, leading to better generalization to target document types without
the need for target-specific training data. QueryForm sets new state-of-the-art
average F1 score on both the XFUND (+4.6%~10.1%) and the Payment (+3.2%~9.5%)
zero-shot benchmark, with a smaller model size and no additional image input.
| [
{
"version": "v1",
"created": "Mon, 14 Nov 2022 20:02:02 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 00:48:46 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Wang",
"Zifeng",
""
],
[
"Zhang",
"Zizhao",
""
],
[
"Devlin",
"Jacob",
""
],
[
"Lee",
"Chen-Yu",
""
],
[
"Su",
"Guolong",
""
],
[
"Zhang",
"Hao",
""
],
[
"Dy",
"Jennifer",
""
],
[
"Perot",
"Vincent",
""
],
[
"Pfister",
"Tomas",
""
]
] | TITLE: QueryForm: A Simple Zero-shot Form Entity Query Framework
ABSTRACT: Zero-shot transfer learning for document understanding is a crucial yet
under-investigated scenario to help reduce the high cost involved in annotating
document entities. We present a novel query-based framework, QueryForm, that
extracts entity values from form-like documents in a zero-shot fashion.
QueryForm contains a dual prompting mechanism that composes both the document
schema and a specific entity type into a query, which is used to prompt a
Transformer model to perform a single entity extraction task. Furthermore, we
propose to leverage large-scale query-entity pairs generated from form-like
webpages with weak HTML annotations to pre-train QueryForm. By unifying
pre-training and fine-tuning into the same query-based framework, QueryForm
enables models to learn from structured documents containing various entities
and layouts, leading to better generalization to target document types without
the need for target-specific training data. QueryForm sets new state-of-the-art
average F1 score on both the XFUND (+4.6%~10.1%) and the Payment (+3.2%~9.5%)
zero-shot benchmark, with a smaller model size and no additional image input.
| no_new_dataset | 0.709856 |
2212.01055 | Erik G\"artner | Erik G\"artner, Luke Metz, Mykhaylo Andriluka, C. Daniel Freeman,
Cristian Sminchisescu | Transformer-Based Learned Optimization | Accepted to the IEEE/CVF Conference on Computer Vision and Pattern
Recognition 2023 (CVPR) in Vancouver, Canada | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a new approach to learned optimization where we represent the
computation of an optimizer's update step using a neural network. The
parameters of the optimizer are then learned by training on a set of
optimization tasks with the objective to perform minimization efficiently. Our
innovation is a new neural network architecture, Optimus, for the learned
optimizer inspired by the classic BFGS algorithm. As in BFGS, we estimate a
preconditioning matrix as a sum of rank-one updates but use a Transformer-based
neural network to predict these updates jointly with the step length and
direction. In contrast to several recent learned optimization-based approaches,
our formulation allows for conditioning across the dimensions of the parameter
space of the target problem while remaining applicable to optimization tasks of
variable dimensionality without retraining. We demonstrate the advantages of
our approach on a benchmark composed of objective functions traditionally used
for the evaluation of optimization algorithms, as well as on the real
world-task of physics-based visual reconstruction of articulated 3d human
motion.
| [
{
"version": "v1",
"created": "Fri, 2 Dec 2022 09:47:08 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Apr 2023 16:46:59 GMT"
},
{
"version": "v3",
"created": "Wed, 24 May 2023 14:40:47 GMT"
},
{
"version": "v4",
"created": "Wed, 28 Jun 2023 09:23:08 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Gärtner",
"Erik",
""
],
[
"Metz",
"Luke",
""
],
[
"Andriluka",
"Mykhaylo",
""
],
[
"Freeman",
"C. Daniel",
""
],
[
"Sminchisescu",
"Cristian",
""
]
] | TITLE: Transformer-Based Learned Optimization
ABSTRACT: We propose a new approach to learned optimization where we represent the
computation of an optimizer's update step using a neural network. The
parameters of the optimizer are then learned by training on a set of
optimization tasks with the objective to perform minimization efficiently. Our
innovation is a new neural network architecture, Optimus, for the learned
optimizer inspired by the classic BFGS algorithm. As in BFGS, we estimate a
preconditioning matrix as a sum of rank-one updates but use a Transformer-based
neural network to predict these updates jointly with the step length and
direction. In contrast to several recent learned optimization-based approaches,
our formulation allows for conditioning across the dimensions of the parameter
space of the target problem while remaining applicable to optimization tasks of
variable dimensionality without retraining. We demonstrate the advantages of
our approach on a benchmark composed of objective functions traditionally used
for the evaluation of optimization algorithms, as well as on the real
world-task of physics-based visual reconstruction of articulated 3d human
motion.
| no_new_dataset | 0.708389 |
2212.01476 | Chao Zhao | Chao Zhao, Faeze Brahman, Kaiqiang Song, Wenlin Yao, Dian Yu, Snigdha
Chaturvedi | NarraSum: A Large-Scale Dataset for Abstractive Narrative Summarization | EMNLP Findings 2022 | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Narrative summarization aims to produce a distilled version of a narrative to
describe its most salient events and characters. Summarizing a narrative is
challenging as it requires an understanding of event causality and character
behaviors. To encourage research in this direction, we propose NarraSum, a
large-scale narrative summarization dataset. It contains 122K narrative
documents, which are collected from plot descriptions of movies and TV episodes
with diverse genres, and their corresponding abstractive summaries. Experiments
show that there is a large performance gap between humans and the
state-of-the-art summarization models on NarraSum. We hope that this dataset
will promote future research in summarization, as well as broader studies of
natural language understanding and generation. The dataset is available at
https://github.com/zhaochaocs/narrasum.
| [
{
"version": "v1",
"created": "Fri, 2 Dec 2022 22:51:51 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 04:08:20 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Zhao",
"Chao",
""
],
[
"Brahman",
"Faeze",
""
],
[
"Song",
"Kaiqiang",
""
],
[
"Yao",
"Wenlin",
""
],
[
"Yu",
"Dian",
""
],
[
"Chaturvedi",
"Snigdha",
""
]
] | TITLE: NarraSum: A Large-Scale Dataset for Abstractive Narrative Summarization
ABSTRACT: Narrative summarization aims to produce a distilled version of a narrative to
describe its most salient events and characters. Summarizing a narrative is
challenging as it requires an understanding of event causality and character
behaviors. To encourage research in this direction, we propose NarraSum, a
large-scale narrative summarization dataset. It contains 122K narrative
documents, which are collected from plot descriptions of movies and TV episodes
with diverse genres, and their corresponding abstractive summaries. Experiments
show that there is a large performance gap between humans and the
state-of-the-art summarization models on NarraSum. We hope that this dataset
will promote future research in summarization, as well as broader studies of
natural language understanding and generation. The dataset is available at
https://github.com/zhaochaocs/narrasum.
| new_dataset | 0.715809 |
2301.03465 | Yankun Xu | Yankun Xu, Jie Yang, Wenjie Ming, Shuang Wang, and Mohamad Sawan | Shorter Latency of Real-time Epileptic Seizure Detection via
Probabilistic Prediction | 9 figures, 4 tables. Submitted to Elsevier ESWA | null | null | null | eess.SP cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although recent studies have proposed seizure detection algorithms with good
sensitivity performance, there is a remained challenge that they were hard to
achieve significantly short detection latency in real-time scenarios. In this
manuscript, we propose a novel deep learning framework intended for shortening
epileptic seizure detection latency via probabilistic prediction. We are the
first to convert the seizure detection task from traditional binary
classification to probabilistic prediction by introducing a crossing period
from seizure-oriented EEG recording and proposing a labeling rule using
soft-label for crossing period samples. And, a novel multiscale STFT-based
feature extraction method combined with 3D-CNN architecture is proposed to
accurately capture predictive probabilities of samples. Furthermore, we also
propose rectified weighting strategy to enhance predictive probabilities, and
accumulative decision-making rule to achieve significantly shorter detection
latency. We implement the proposed framework on two prevalent datasets --
CHB-MIT scalp EEG dataset and SWEC-ETHZ intracranial EEG dataset in
patient-specific leave-one-seizure-out cross-validation scheme. Eventually, the
proposed algorithm successfully detected 94 out of 99 seizures during crossing
period and 100% seizures detected after EEG onset, averaged 14.84% rectified
predictive ictal probability (RPIP) errors of crossing samples, 2.3 s detection
latency, 0.08/h false detection rate (FDR) on CHB-MIT dataset. Meanwhile, 84
out of 89 detected seizures during crossing period, 100% detected seizures
after EEG onset, 16.17% RPIP errors, 4.7 s detection latency, and 0.08/h FDR
are achieved on SWEC-ETHZ dataset. The obtained detection latencies are at
least 50% shorter than state-of-the-art results reported in previous studies.
| [
{
"version": "v1",
"created": "Wed, 4 Jan 2023 08:45:47 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 04:40:48 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Xu",
"Yankun",
""
],
[
"Yang",
"Jie",
""
],
[
"Ming",
"Wenjie",
""
],
[
"Wang",
"Shuang",
""
],
[
"Sawan",
"Mohamad",
""
]
] | TITLE: Shorter Latency of Real-time Epileptic Seizure Detection via
Probabilistic Prediction
ABSTRACT: Although recent studies have proposed seizure detection algorithms with good
sensitivity performance, there is a remained challenge that they were hard to
achieve significantly short detection latency in real-time scenarios. In this
manuscript, we propose a novel deep learning framework intended for shortening
epileptic seizure detection latency via probabilistic prediction. We are the
first to convert the seizure detection task from traditional binary
classification to probabilistic prediction by introducing a crossing period
from seizure-oriented EEG recording and proposing a labeling rule using
soft-label for crossing period samples. And, a novel multiscale STFT-based
feature extraction method combined with 3D-CNN architecture is proposed to
accurately capture predictive probabilities of samples. Furthermore, we also
propose rectified weighting strategy to enhance predictive probabilities, and
accumulative decision-making rule to achieve significantly shorter detection
latency. We implement the proposed framework on two prevalent datasets --
CHB-MIT scalp EEG dataset and SWEC-ETHZ intracranial EEG dataset in
patient-specific leave-one-seizure-out cross-validation scheme. Eventually, the
proposed algorithm successfully detected 94 out of 99 seizures during crossing
period and 100% seizures detected after EEG onset, averaged 14.84% rectified
predictive ictal probability (RPIP) errors of crossing samples, 2.3 s detection
latency, 0.08/h false detection rate (FDR) on CHB-MIT dataset. Meanwhile, 84
out of 89 detected seizures during crossing period, 100% detected seizures
after EEG onset, 16.17% RPIP errors, 4.7 s detection latency, and 0.08/h FDR
are achieved on SWEC-ETHZ dataset. The obtained detection latencies are at
least 50% shorter than state-of-the-art results reported in previous studies.
| no_new_dataset | 0.712995 |
2303.09519 | Bobby Huggins | Bobby Huggins, Chengkun Li, Marlon Tobaben, Mikko J. Aarnos, Luigi
Acerbi | PyVBMC: Efficient Bayesian inference in Python | 6 pages, 1 figure. Published in The Journal of Open Source Software.
Documentation is available at https://acerbilab.github.io/pyvbmc and source
code is available at https://github.com/acerbilab/pyvbmc | The Journal of Open Source Software, 8(86), 2023, 5428 | 10.21105/joss.05428 | null | stat.ML cs.LG stat.CO stat.ME | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | PyVBMC is a Python implementation of the Variational Bayesian Monte Carlo
(VBMC) algorithm for posterior and model inference for black-box computational
models (Acerbi, 2018, 2020). VBMC is an approximate inference method designed
for efficient parameter estimation and model assessment when model evaluations
are mildly-to-very expensive (e.g., a second or more) and/or noisy.
Specifically, VBMC computes:
- a flexible (non-Gaussian) approximate posterior distribution of the model
parameters, from which statistics and posterior samples can be easily
extracted;
- an approximation of the model evidence or marginal likelihood, a metric
used for Bayesian model selection.
PyVBMC can be applied to any computational or statistical model with up to
roughly 10-15 continuous parameters, with the only requirement that the user
can provide a Python function that computes the target log likelihood of the
model, or an approximation thereof (e.g., an estimate of the likelihood
obtained via simulation or Monte Carlo methods). PyVBMC is particularly
effective when the model takes more than about a second per evaluation, with
dramatic speed-ups of 1-2 orders of magnitude when compared to traditional
approximate inference methods.
Extensive benchmarks on both artificial test problems and a large number of
real models from the computational sciences, particularly computational and
cognitive neuroscience, show that VBMC generally - and often vastly -
outperforms alternative methods for sample-efficient Bayesian inference, and is
applicable to both exact and simulator-based models (Acerbi, 2018, 2019, 2020).
PyVBMC brings this state-of-the-art inference algorithm to Python, along with
an easy-to-use Pythonic interface for running the algorithm and manipulating
and visualizing its results.
| [
{
"version": "v1",
"created": "Thu, 16 Mar 2023 17:37:22 GMT"
},
{
"version": "v2",
"created": "Tue, 27 Jun 2023 20:14:04 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Huggins",
"Bobby",
""
],
[
"Li",
"Chengkun",
""
],
[
"Tobaben",
"Marlon",
""
],
[
"Aarnos",
"Mikko J.",
""
],
[
"Acerbi",
"Luigi",
""
]
] | TITLE: PyVBMC: Efficient Bayesian inference in Python
ABSTRACT: PyVBMC is a Python implementation of the Variational Bayesian Monte Carlo
(VBMC) algorithm for posterior and model inference for black-box computational
models (Acerbi, 2018, 2020). VBMC is an approximate inference method designed
for efficient parameter estimation and model assessment when model evaluations
are mildly-to-very expensive (e.g., a second or more) and/or noisy.
Specifically, VBMC computes:
- a flexible (non-Gaussian) approximate posterior distribution of the model
parameters, from which statistics and posterior samples can be easily
extracted;
- an approximation of the model evidence or marginal likelihood, a metric
used for Bayesian model selection.
PyVBMC can be applied to any computational or statistical model with up to
roughly 10-15 continuous parameters, with the only requirement that the user
can provide a Python function that computes the target log likelihood of the
model, or an approximation thereof (e.g., an estimate of the likelihood
obtained via simulation or Monte Carlo methods). PyVBMC is particularly
effective when the model takes more than about a second per evaluation, with
dramatic speed-ups of 1-2 orders of magnitude when compared to traditional
approximate inference methods.
Extensive benchmarks on both artificial test problems and a large number of
real models from the computational sciences, particularly computational and
cognitive neuroscience, show that VBMC generally - and often vastly -
outperforms alternative methods for sample-efficient Bayesian inference, and is
applicable to both exact and simulator-based models (Acerbi, 2018, 2019, 2020).
PyVBMC brings this state-of-the-art inference algorithm to Python, along with
an easy-to-use Pythonic interface for running the algorithm and manipulating
and visualizing its results.
| no_new_dataset | 0.706007 |
2303.13391 | Chantal Pellegrini | Chantal Pellegrini, Matthias Keicher, Ege \"Ozsoy, Petra Jiraskova,
Rickmer Braren, Nassir Navab | Xplainer: From X-Ray Observations to Explainable Zero-Shot Diagnosis | provisionally accepted for publication at MICCAI 2023, 9 pages, 2
figures, 6 tables | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automated diagnosis prediction from medical images is a valuable resource to
support clinical decision-making. However, such systems usually need to be
trained on large amounts of annotated data, which often is scarce in the
medical domain. Zero-shot methods address this challenge by allowing a flexible
adaption to new settings with different clinical findings without relying on
labeled data. Further, to integrate automated diagnosis in the clinical
workflow, methods should be transparent and explainable, increasing medical
professionals' trust and facilitating correctness verification. In this work,
we introduce Xplainer, a novel framework for explainable zero-shot diagnosis in
the clinical setting. Xplainer adapts the classification-by-description
approach of contrastive vision-language models to the multi-label medical
diagnosis task. Specifically, instead of directly predicting a diagnosis, we
prompt the model to classify the existence of descriptive observations, which a
radiologist would look for on an X-Ray scan, and use the descriptor
probabilities to estimate the likelihood of a diagnosis. Our model is
explainable by design, as the final diagnosis prediction is directly based on
the prediction of the underlying descriptors. We evaluate Xplainer on two chest
X-ray datasets, CheXpert and ChestX-ray14, and demonstrate its effectiveness in
improving the performance and explainability of zero-shot diagnosis. Our
results suggest that Xplainer provides a more detailed understanding of the
decision-making process and can be a valuable tool for clinical diagnosis.
| [
{
"version": "v1",
"created": "Thu, 23 Mar 2023 16:07:31 GMT"
},
{
"version": "v2",
"created": "Mon, 27 Mar 2023 11:36:12 GMT"
},
{
"version": "v3",
"created": "Wed, 28 Jun 2023 10:26:46 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Pellegrini",
"Chantal",
""
],
[
"Keicher",
"Matthias",
""
],
[
"Özsoy",
"Ege",
""
],
[
"Jiraskova",
"Petra",
""
],
[
"Braren",
"Rickmer",
""
],
[
"Navab",
"Nassir",
""
]
] | TITLE: Xplainer: From X-Ray Observations to Explainable Zero-Shot Diagnosis
ABSTRACT: Automated diagnosis prediction from medical images is a valuable resource to
support clinical decision-making. However, such systems usually need to be
trained on large amounts of annotated data, which often is scarce in the
medical domain. Zero-shot methods address this challenge by allowing a flexible
adaption to new settings with different clinical findings without relying on
labeled data. Further, to integrate automated diagnosis in the clinical
workflow, methods should be transparent and explainable, increasing medical
professionals' trust and facilitating correctness verification. In this work,
we introduce Xplainer, a novel framework for explainable zero-shot diagnosis in
the clinical setting. Xplainer adapts the classification-by-description
approach of contrastive vision-language models to the multi-label medical
diagnosis task. Specifically, instead of directly predicting a diagnosis, we
prompt the model to classify the existence of descriptive observations, which a
radiologist would look for on an X-Ray scan, and use the descriptor
probabilities to estimate the likelihood of a diagnosis. Our model is
explainable by design, as the final diagnosis prediction is directly based on
the prediction of the underlying descriptors. We evaluate Xplainer on two chest
X-ray datasets, CheXpert and ChestX-ray14, and demonstrate its effectiveness in
improving the performance and explainability of zero-shot diagnosis. Our
results suggest that Xplainer provides a more detailed understanding of the
decision-making process and can be a valuable tool for clinical diagnosis.
| no_new_dataset | 0.709044 |
2303.14307 | Pingchuan Ma | Pingchuan Ma, Alexandros Haliassos, Adriana Fernandez-Lopez, Honglie
Chen, Stavros Petridis, Maja Pantic | Auto-AVSR: Audio-Visual Speech Recognition with Automatic Labels | Accepted to ICASSP 2023 | null | 10.1109/ICASSP49357.2023.10096889 | null | cs.CV cs.SD eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Audio-visual speech recognition has received a lot of attention due to its
robustness against acoustic noise. Recently, the performance of automatic,
visual, and audio-visual speech recognition (ASR, VSR, and AV-ASR,
respectively) has been substantially improved, mainly due to the use of larger
models and training sets. However, accurate labelling of datasets is
time-consuming and expensive. Hence, in this work, we investigate the use of
automatically-generated transcriptions of unlabelled datasets to increase the
training set size. For this purpose, we use publicly-available pre-trained ASR
models to automatically transcribe unlabelled datasets such as AVSpeech and
VoxCeleb2. Then, we train ASR, VSR and AV-ASR models on the augmented training
set, which consists of the LRS2 and LRS3 datasets as well as the additional
automatically-transcribed data. We demonstrate that increasing the size of the
training set, a recent trend in the literature, leads to reduced WER despite
using noisy transcriptions. The proposed model achieves new state-of-the-art
performance on AV-ASR on LRS2 and LRS3. In particular, it achieves a WER of
0.9% on LRS3, a relative improvement of 30% over the current state-of-the-art
approach, and outperforms methods that have been trained on non-publicly
available datasets with 26 times more training data.
| [
{
"version": "v1",
"created": "Sat, 25 Mar 2023 00:37:34 GMT"
},
{
"version": "v2",
"created": "Fri, 16 Jun 2023 16:22:36 GMT"
},
{
"version": "v3",
"created": "Wed, 28 Jun 2023 14:41:17 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Ma",
"Pingchuan",
""
],
[
"Haliassos",
"Alexandros",
""
],
[
"Fernandez-Lopez",
"Adriana",
""
],
[
"Chen",
"Honglie",
""
],
[
"Petridis",
"Stavros",
""
],
[
"Pantic",
"Maja",
""
]
] | TITLE: Auto-AVSR: Audio-Visual Speech Recognition with Automatic Labels
ABSTRACT: Audio-visual speech recognition has received a lot of attention due to its
robustness against acoustic noise. Recently, the performance of automatic,
visual, and audio-visual speech recognition (ASR, VSR, and AV-ASR,
respectively) has been substantially improved, mainly due to the use of larger
models and training sets. However, accurate labelling of datasets is
time-consuming and expensive. Hence, in this work, we investigate the use of
automatically-generated transcriptions of unlabelled datasets to increase the
training set size. For this purpose, we use publicly-available pre-trained ASR
models to automatically transcribe unlabelled datasets such as AVSpeech and
VoxCeleb2. Then, we train ASR, VSR and AV-ASR models on the augmented training
set, which consists of the LRS2 and LRS3 datasets as well as the additional
automatically-transcribed data. We demonstrate that increasing the size of the
training set, a recent trend in the literature, leads to reduced WER despite
using noisy transcriptions. The proposed model achieves new state-of-the-art
performance on AV-ASR on LRS2 and LRS3. In particular, it achieves a WER of
0.9% on LRS3, a relative improvement of 30% over the current state-of-the-art
approach, and outperforms methods that have been trained on non-publicly
available datasets with 26 times more training data.
| no_new_dataset | 0.709636 |
2304.12988 | Xiao Qi | Xiao Qi, David J. Foran, John L. Nosher, and Ilker Hacihaliloglu | Multi-Scale Feature Fusion using Parallel-Attention Block for COVID-19
Chest X-ray Diagnosis | Accepted for publication at the Journal of Machine Learning for
Biomedical Imaging (MELBA) https://melba-journal.org/2023:008 | Machine.Learning.for.Biomedical.Imaging. 2 (2023) | 10.59275/j.melba.2023-7e96 | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Under the global COVID-19 crisis, accurate diagnosis of COVID-19 from Chest
X-ray (CXR) images is critical. To reduce intra- and inter-observer
variability, during the radiological assessment, computer-aided diagnostic
tools have been utilized to supplement medical decision-making and subsequent
disease management. Computational methods with high accuracy and robustness are
required for rapid triaging of patients and aiding radiologists in the
interpretation of the collected data. In this study, we propose a novel
multi-feature fusion network using parallel attention blocks to fuse the
original CXR images and local-phase feature-enhanced CXR images at
multi-scales. We examine our model on various COVID-19 datasets acquired from
different organizations to assess the generalization ability. Our experiments
demonstrate that our method achieves state-of-art performance and has improved
generalization capability, which is crucial for widespread deployment.
| [
{
"version": "v1",
"created": "Tue, 25 Apr 2023 16:56:12 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Qi",
"Xiao",
""
],
[
"Foran",
"David J.",
""
],
[
"Nosher",
"John L.",
""
],
[
"Hacihaliloglu",
"Ilker",
""
]
] | TITLE: Multi-Scale Feature Fusion using Parallel-Attention Block for COVID-19
Chest X-ray Diagnosis
ABSTRACT: Under the global COVID-19 crisis, accurate diagnosis of COVID-19 from Chest
X-ray (CXR) images is critical. To reduce intra- and inter-observer
variability, during the radiological assessment, computer-aided diagnostic
tools have been utilized to supplement medical decision-making and subsequent
disease management. Computational methods with high accuracy and robustness are
required for rapid triaging of patients and aiding radiologists in the
interpretation of the collected data. In this study, we propose a novel
multi-feature fusion network using parallel attention blocks to fuse the
original CXR images and local-phase feature-enhanced CXR images at
multi-scales. We examine our model on various COVID-19 datasets acquired from
different organizations to assess the generalization ability. Our experiments
demonstrate that our method achieves state-of-art performance and has improved
generalization capability, which is crucial for widespread deployment.
| no_new_dataset | 0.709799 |
2305.03617 | Yuanyuan Peng | Yuanyuan Peng, Pengpeng Luan, Zixu Zhang | MAF-Net: Multiple attention-guided fusion network for fundus vascular
image segmentation | 19 pages,9 figures | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Accurately segmenting blood vessels in retinal fundus images is crucial in
the early screening, diagnosing, and evaluating some ocular diseases, yet it
poses a nontrivial uncertainty for the segmentation task due to various factors
such as significant light variations, uneven curvilinear structures, and
non-uniform contrast. As a result, a multiple attention-guided fusion network
(MAF-Net) is proposed to accurately detect blood vessels in retinal fundus
images. Currently, traditional UNet-based models may lose partial information
due to explicitly modeling long-distance dependencies, which may lead to
unsatisfactory results. To enrich contextual information for the loss of scene
information compensation, an attention fusion mechanism that combines the
channel attention with spatial attention mechanisms constructed by Transformer
is employed to extract various features of blood vessels from retinal fundus
images. Subsequently, a unique spatial attention mechanism is applied in the
skip connection to filter out redundant information and noise from low-level
features, thus enabling better integration with high-level features. In
addition, a DropOut layer is employed to randomly discard some neurons, which
can prevent overfitting of the deep learning network and improve its
generalization performance. Experimental results were verified in public
datasets DRIVE, STARE and CHASEDB1 with F1 scores of 0.818, 0.836 and 0.811,
and Acc values of 0.968, 0.973 and 0.973, respectively. Both visual inspection
and quantitative evaluation demonstrate that our method produces satisfactory
results compared to some state-of-the-art methods.
| [
{
"version": "v1",
"created": "Fri, 5 May 2023 15:22:20 GMT"
},
{
"version": "v2",
"created": "Wed, 14 Jun 2023 10:33:09 GMT"
},
{
"version": "v3",
"created": "Wed, 28 Jun 2023 13:49:32 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Peng",
"Yuanyuan",
""
],
[
"Luan",
"Pengpeng",
""
],
[
"Zhang",
"Zixu",
""
]
] | TITLE: MAF-Net: Multiple attention-guided fusion network for fundus vascular
image segmentation
ABSTRACT: Accurately segmenting blood vessels in retinal fundus images is crucial in
the early screening, diagnosing, and evaluating some ocular diseases, yet it
poses a nontrivial uncertainty for the segmentation task due to various factors
such as significant light variations, uneven curvilinear structures, and
non-uniform contrast. As a result, a multiple attention-guided fusion network
(MAF-Net) is proposed to accurately detect blood vessels in retinal fundus
images. Currently, traditional UNet-based models may lose partial information
due to explicitly modeling long-distance dependencies, which may lead to
unsatisfactory results. To enrich contextual information for the loss of scene
information compensation, an attention fusion mechanism that combines the
channel attention with spatial attention mechanisms constructed by Transformer
is employed to extract various features of blood vessels from retinal fundus
images. Subsequently, a unique spatial attention mechanism is applied in the
skip connection to filter out redundant information and noise from low-level
features, thus enabling better integration with high-level features. In
addition, a DropOut layer is employed to randomly discard some neurons, which
can prevent overfitting of the deep learning network and improve its
generalization performance. Experimental results were verified in public
datasets DRIVE, STARE and CHASEDB1 with F1 scores of 0.818, 0.836 and 0.811,
and Acc values of 0.968, 0.973 and 0.973, respectively. Both visual inspection
and quantitative evaluation demonstrate that our method produces satisfactory
results compared to some state-of-the-art methods.
| no_new_dataset | 0.710616 |
2305.18418 | Lama Alssum | Lama Alssum, Juan Leon Alcazar, Merey Ramazanova, Chen Zhao, Bernard
Ghanem | Just a Glimpse: Rethinking Temporal Information for Video Continual
Learning | Accepted at CLVision Workshop - CVPR23 (Best Paper Award) | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Class-incremental learning is one of the most important settings for the
study of Continual Learning, as it closely resembles real-world application
scenarios. With constrained memory sizes, catastrophic forgetting arises as the
number of classes/tasks increases. Studying continual learning in the video
domain poses even more challenges, as video data contains a large number of
frames, which places a higher burden on the replay memory. The current common
practice is to sub-sample frames from the video stream and store them in the
replay memory. In this paper, we propose SMILE a novel replay mechanism for
effective video continual learning based on individual/single frames. Through
extensive experimentation, we show that under extreme memory constraints, video
diversity plays a more significant role than temporal information. Therefore,
our method focuses on learning from a small number of frames that represent a
large number of unique videos. On three representative video datasets,
Kinetics, UCF101, and ActivityNet, the proposed method achieves
state-of-the-art performance, outperforming the previous state-of-the-art by up
to 21.49%.
| [
{
"version": "v1",
"created": "Sun, 28 May 2023 19:14:25 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 12:34:22 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Alssum",
"Lama",
""
],
[
"Alcazar",
"Juan Leon",
""
],
[
"Ramazanova",
"Merey",
""
],
[
"Zhao",
"Chen",
""
],
[
"Ghanem",
"Bernard",
""
]
] | TITLE: Just a Glimpse: Rethinking Temporal Information for Video Continual
Learning
ABSTRACT: Class-incremental learning is one of the most important settings for the
study of Continual Learning, as it closely resembles real-world application
scenarios. With constrained memory sizes, catastrophic forgetting arises as the
number of classes/tasks increases. Studying continual learning in the video
domain poses even more challenges, as video data contains a large number of
frames, which places a higher burden on the replay memory. The current common
practice is to sub-sample frames from the video stream and store them in the
replay memory. In this paper, we propose SMILE a novel replay mechanism for
effective video continual learning based on individual/single frames. Through
extensive experimentation, we show that under extreme memory constraints, video
diversity plays a more significant role than temporal information. Therefore,
our method focuses on learning from a small number of frames that represent a
large number of unique videos. On three representative video datasets,
Kinetics, UCF101, and ActivityNet, the proposed method achieves
state-of-the-art performance, outperforming the previous state-of-the-art by up
to 21.49%.
| no_new_dataset | 0.70866 |
2306.01340 | Zehui Liao | Zehui Liao, Yutong Xie, Shishuai Hu, Yong Xia | Transformer-based Annotation Bias-aware Medical Image Segmentation | MICCAI 2023 Early Accept (Top 14%), 11 pages, 2 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Manual medical image segmentation is subjective and suffers from
annotator-related bias, which can be mimicked or amplified by deep learning
methods. Recently, researchers have suggested that such bias is the combination
of the annotator preference and stochastic error, which are modeled by
convolution blocks located after decoder and pixel-wise independent Gaussian
distribution, respectively. It is unlikely that convolution blocks can
effectively model the varying degrees of preference at the full resolution
level. Additionally, the independent pixel-wise Gaussian distribution
disregards pixel correlations, leading to a discontinuous boundary. This paper
proposes a Transformer-based Annotation Bias-aware (TAB) medical image
segmentation model, which tackles the annotator-related bias via modeling
annotator preference and stochastic errors. TAB employs the Transformer with
learnable queries to extract the different preference-focused features. This
enables TAB to produce segmentation with various preferences simultaneously
using a single segmentation head. Moreover, TAB takes the multivariant normal
distribution assumption that models pixel correlations, and learns the
annotation distribution to disentangle the stochastic error. We evaluated our
TAB on an OD/OC segmentation benchmark annotated by six annotators. Our results
suggest that TAB outperforms existing medical image segmentation models which
take into account the annotator-related bias.
| [
{
"version": "v1",
"created": "Fri, 2 Jun 2023 08:09:53 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 16:08:39 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Liao",
"Zehui",
""
],
[
"Xie",
"Yutong",
""
],
[
"Hu",
"Shishuai",
""
],
[
"Xia",
"Yong",
""
]
] | TITLE: Transformer-based Annotation Bias-aware Medical Image Segmentation
ABSTRACT: Manual medical image segmentation is subjective and suffers from
annotator-related bias, which can be mimicked or amplified by deep learning
methods. Recently, researchers have suggested that such bias is the combination
of the annotator preference and stochastic error, which are modeled by
convolution blocks located after decoder and pixel-wise independent Gaussian
distribution, respectively. It is unlikely that convolution blocks can
effectively model the varying degrees of preference at the full resolution
level. Additionally, the independent pixel-wise Gaussian distribution
disregards pixel correlations, leading to a discontinuous boundary. This paper
proposes a Transformer-based Annotation Bias-aware (TAB) medical image
segmentation model, which tackles the annotator-related bias via modeling
annotator preference and stochastic errors. TAB employs the Transformer with
learnable queries to extract the different preference-focused features. This
enables TAB to produce segmentation with various preferences simultaneously
using a single segmentation head. Moreover, TAB takes the multivariant normal
distribution assumption that models pixel correlations, and learns the
annotation distribution to disentangle the stochastic error. We evaluated our
TAB on an OD/OC segmentation benchmark annotated by six annotators. Our results
suggest that TAB outperforms existing medical image segmentation models which
take into account the annotator-related bias.
| no_new_dataset | 0.710176 |
2306.05143 | Zehui Li | Zehui Li, Akashaditya Das, William A V Beardall, Yiren Zhao, Guy-Bart
Stan | Genomic Interpreter: A Hierarchical Genomic Deep Neural Network with 1D
Shifted Window Transformer | 40th International Conference on Machine Learning (ICML 2023)
Workshop on Computational Biology (WCB) | null | null | null | cs.LG q-bio.GN | http://creativecommons.org/licenses/by/4.0/ | Given the increasing volume and quality of genomics data, extracting new
insights requires interpretable machine-learning models. This work presents
Genomic Interpreter: a novel architecture for genomic assay prediction. This
model outperforms the state-of-the-art models for genomic assay prediction
tasks. Our model can identify hierarchical dependencies in genomic sites. This
is achieved through the integration of 1D-Swin, a novel Transformer-based block
designed by us for modelling long-range hierarchical data. Evaluated on a
dataset containing 38,171 DNA segments of 17K base pairs, Genomic Interpreter
demonstrates superior performance in chromatin accessibility and gene
expression prediction and unmasks the underlying `syntax' of gene regulation.
| [
{
"version": "v1",
"created": "Thu, 8 Jun 2023 12:10:13 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 08:17:32 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Li",
"Zehui",
""
],
[
"Das",
"Akashaditya",
""
],
[
"Beardall",
"William A V",
""
],
[
"Zhao",
"Yiren",
""
],
[
"Stan",
"Guy-Bart",
""
]
] | TITLE: Genomic Interpreter: A Hierarchical Genomic Deep Neural Network with 1D
Shifted Window Transformer
ABSTRACT: Given the increasing volume and quality of genomics data, extracting new
insights requires interpretable machine-learning models. This work presents
Genomic Interpreter: a novel architecture for genomic assay prediction. This
model outperforms the state-of-the-art models for genomic assay prediction
tasks. Our model can identify hierarchical dependencies in genomic sites. This
is achieved through the integration of 1D-Swin, a novel Transformer-based block
designed by us for modelling long-range hierarchical data. Evaluated on a
dataset containing 38,171 DNA segments of 17K base pairs, Genomic Interpreter
demonstrates superior performance in chromatin accessibility and gene
expression prediction and unmasks the underlying `syntax' of gene regulation.
| no_new_dataset | 0.708175 |
2306.06238 | Harvey Dam | Harvey Dam, Vinu Joseph, Aditya Bhaskara, Ganesh Gopalakrishnan,
Saurav Muralidharan, Michael Garland | Understanding the Effect of the Long Tail on Neural Network Compression | null | null | null | null | cs.LG cs.AI cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Network compression is now a mature sub-field of neural network research:
over the last decade, significant progress has been made towards reducing the
size of models and speeding up inference, while maintaining the classification
accuracy. However, many works have observed that focusing on just the overall
accuracy can be misguided. E.g., it has been shown that mismatches between the
full and compressed models can be biased towards under-represented classes.
This raises the important research question, can we achieve network compression
while maintaining "semantic equivalence" with the original network? In this
work, we study this question in the context of the "long tail" phenomenon in
computer vision datasets observed by Feldman, et al. They argue that
memorization of certain inputs (appropriately defined) is essential to
achieving good generalization. As compression limits the capacity of a network
(and hence also its ability to memorize), we study the question: are mismatches
between the full and compressed models correlated with the memorized training
data? We present positive evidence in this direction for image classification
tasks, by considering different base architectures and compression schemes.
| [
{
"version": "v1",
"created": "Fri, 9 Jun 2023 20:18:05 GMT"
},
{
"version": "v2",
"created": "Mon, 19 Jun 2023 09:46:09 GMT"
},
{
"version": "v3",
"created": "Tue, 27 Jun 2023 23:14:16 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Dam",
"Harvey",
""
],
[
"Joseph",
"Vinu",
""
],
[
"Bhaskara",
"Aditya",
""
],
[
"Gopalakrishnan",
"Ganesh",
""
],
[
"Muralidharan",
"Saurav",
""
],
[
"Garland",
"Michael",
""
]
] | TITLE: Understanding the Effect of the Long Tail on Neural Network Compression
ABSTRACT: Network compression is now a mature sub-field of neural network research:
over the last decade, significant progress has been made towards reducing the
size of models and speeding up inference, while maintaining the classification
accuracy. However, many works have observed that focusing on just the overall
accuracy can be misguided. E.g., it has been shown that mismatches between the
full and compressed models can be biased towards under-represented classes.
This raises the important research question, can we achieve network compression
while maintaining "semantic equivalence" with the original network? In this
work, we study this question in the context of the "long tail" phenomenon in
computer vision datasets observed by Feldman, et al. They argue that
memorization of certain inputs (appropriately defined) is essential to
achieving good generalization. As compression limits the capacity of a network
(and hence also its ability to memorize), we study the question: are mismatches
between the full and compressed models correlated with the memorized training
data? We present positive evidence in this direction for image classification
tasks, by considering different base architectures and compression schemes.
| no_new_dataset | 0.707979 |
2306.07089 | Jiancheng Yang | Ziqiao Weng, Jiancheng Yang, Dongnan Liu, Weidong Cai | Topology Repairing of Disconnected Pulmonary Airways and Vessels:
Baselines and a Dataset | MICCAI 2023 Early Accepted | null | null | null | eess.IV cs.AI cs.CV | http://creativecommons.org/licenses/by/4.0/ | Accurate segmentation of pulmonary airways and vessels is crucial for the
diagnosis and treatment of pulmonary diseases. However, current deep learning
approaches suffer from disconnectivity issues that hinder their clinical
usefulness. To address this challenge, we propose a post-processing approach
that leverages a data-driven method to repair the topology of disconnected
pulmonary tubular structures. Our approach formulates the problem as a keypoint
detection task, where a neural network is trained to predict keypoints that can
bridge disconnected components. We use a training data synthesis pipeline that
generates disconnected data from complete pulmonary structures. Moreover, the
new Pulmonary Tree Repairing (PTR) dataset is publicly available, which
comprises 800 complete 3D models of pulmonary airways, arteries, and veins, as
well as the synthetic disconnected data. Our code and data are available at
https://github.com/M3DV/pulmonary-tree-repairing.
| [
{
"version": "v1",
"created": "Mon, 12 Jun 2023 13:01:50 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 17:33:27 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Weng",
"Ziqiao",
""
],
[
"Yang",
"Jiancheng",
""
],
[
"Liu",
"Dongnan",
""
],
[
"Cai",
"Weidong",
""
]
] | TITLE: Topology Repairing of Disconnected Pulmonary Airways and Vessels:
Baselines and a Dataset
ABSTRACT: Accurate segmentation of pulmonary airways and vessels is crucial for the
diagnosis and treatment of pulmonary diseases. However, current deep learning
approaches suffer from disconnectivity issues that hinder their clinical
usefulness. To address this challenge, we propose a post-processing approach
that leverages a data-driven method to repair the topology of disconnected
pulmonary tubular structures. Our approach formulates the problem as a keypoint
detection task, where a neural network is trained to predict keypoints that can
bridge disconnected components. We use a training data synthesis pipeline that
generates disconnected data from complete pulmonary structures. Moreover, the
new Pulmonary Tree Repairing (PTR) dataset is publicly available, which
comprises 800 complete 3D models of pulmonary airways, arteries, and veins, as
well as the synthetic disconnected data. Our code and data are available at
https://github.com/M3DV/pulmonary-tree-repairing.
| new_dataset | 0.7138 |
2306.08640 | Difei Gao | Difei Gao, Lei Ji, Luowei Zhou, Kevin Qinghong Lin, Joya Chen, Zihan
Fan, Mike Zheng Shou | AssistGPT: A General Multi-modal Assistant that can Plan, Execute,
Inspect, and Learn | Project page: https://showlab.github.io/assistgpt/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recent research on Large Language Models (LLMs) has led to remarkable
advancements in general NLP AI assistants. Some studies have further explored
the use of LLMs for planning and invoking models or APIs to address more
general multi-modal user queries. Despite this progress, complex visual-based
tasks still remain challenging due to the diverse nature of visual tasks. This
diversity is reflected in two aspects: 1) Reasoning paths. For many real-life
applications, it is hard to accurately decompose a query simply by examining
the query itself. Planning based on the specific visual content and the results
of each step is usually required. 2) Flexible inputs and intermediate results.
Input forms could be flexible for in-the-wild cases, and involves not only a
single image or video but a mixture of videos and images, e.g., a user-view
image with some reference videos. Besides, a complex reasoning process will
also generate diverse multimodal intermediate results, e.g., video narrations,
segmented video clips, etc. To address such general cases, we propose a
multi-modal AI assistant, AssistGPT, with an interleaved code and language
reasoning approach called Plan, Execute, Inspect, and Learn (PEIL) to integrate
LLMs with various tools. Specifically, the Planner is capable of using natural
language to plan which tool in Executor should do next based on the current
reasoning progress. Inspector is an efficient memory manager to assist the
Planner to feed proper visual information into a specific tool. Finally, since
the entire reasoning process is complex and flexible, a Learner is designed to
enable the model to autonomously explore and discover the optimal solution. We
conducted experiments on A-OKVQA and NExT-QA benchmarks, achieving
state-of-the-art results. Moreover, showcases demonstrate the ability of our
system to handle questions far more complex than those found in the benchmarks.
| [
{
"version": "v1",
"created": "Wed, 14 Jun 2023 17:12:56 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 05:00:35 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Gao",
"Difei",
""
],
[
"Ji",
"Lei",
""
],
[
"Zhou",
"Luowei",
""
],
[
"Lin",
"Kevin Qinghong",
""
],
[
"Chen",
"Joya",
""
],
[
"Fan",
"Zihan",
""
],
[
"Shou",
"Mike Zheng",
""
]
] | TITLE: AssistGPT: A General Multi-modal Assistant that can Plan, Execute,
Inspect, and Learn
ABSTRACT: Recent research on Large Language Models (LLMs) has led to remarkable
advancements in general NLP AI assistants. Some studies have further explored
the use of LLMs for planning and invoking models or APIs to address more
general multi-modal user queries. Despite this progress, complex visual-based
tasks still remain challenging due to the diverse nature of visual tasks. This
diversity is reflected in two aspects: 1) Reasoning paths. For many real-life
applications, it is hard to accurately decompose a query simply by examining
the query itself. Planning based on the specific visual content and the results
of each step is usually required. 2) Flexible inputs and intermediate results.
Input forms could be flexible for in-the-wild cases, and involves not only a
single image or video but a mixture of videos and images, e.g., a user-view
image with some reference videos. Besides, a complex reasoning process will
also generate diverse multimodal intermediate results, e.g., video narrations,
segmented video clips, etc. To address such general cases, we propose a
multi-modal AI assistant, AssistGPT, with an interleaved code and language
reasoning approach called Plan, Execute, Inspect, and Learn (PEIL) to integrate
LLMs with various tools. Specifically, the Planner is capable of using natural
language to plan which tool in Executor should do next based on the current
reasoning progress. Inspector is an efficient memory manager to assist the
Planner to feed proper visual information into a specific tool. Finally, since
the entire reasoning process is complex and flexible, a Learner is designed to
enable the model to autonomously explore and discover the optimal solution. We
conducted experiments on A-OKVQA and NExT-QA benchmarks, achieving
state-of-the-art results. Moreover, showcases demonstrate the ability of our
system to handle questions far more complex than those found in the benchmarks.
| no_new_dataset | 0.706437 |
2306.11429 | Giovanni Cioffi | Giovanni Cioffi, Leonard Bauersfeld, Davide Scaramuzza | HDVIO: Improving Localization and Disturbance Estimation with Hybrid
Dynamics VIO | null | Robotics: Science and Systems (RSS) 2023 | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Visual-inertial odometry (VIO) is the most common approach for estimating the
state of autonomous micro aerial vehicles using only onboard sensors. Existing
methods improve VIO performance by including a dynamics model in the estimation
pipeline. However, such methods degrade in the presence of low-fidelity vehicle
models and continuous external disturbances, such as wind. Our proposed method,
HDVIO, overcomes these limitations by using a hybrid dynamics model that
combines a point-mass vehicle model with a learning-based component that
captures complex aerodynamic effects. HDVIO estimates the external force and
the full robot state by leveraging the discrepancy between the actual motion
and the predicted motion of the hybrid dynamics model. Our hybrid dynamics
model uses a history of thrust and IMU measurements to predict the vehicle
dynamics. To demonstrate the performance of our method, we present results on
both public and novel drone dynamics datasets and show real-world experiments
of a quadrotor flying in strong winds up to 25 km/h. The results show that our
approach improves the motion and external force estimation compared to the
state-of-the-art by up to 33% and 40%, respectively. Furthermore, differently
from existing methods, we show that it is possible to predict the vehicle
dynamics accurately while having no explicit knowledge of its full state.
| [
{
"version": "v1",
"created": "Tue, 20 Jun 2023 10:19:16 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 14:06:22 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Cioffi",
"Giovanni",
""
],
[
"Bauersfeld",
"Leonard",
""
],
[
"Scaramuzza",
"Davide",
""
]
] | TITLE: HDVIO: Improving Localization and Disturbance Estimation with Hybrid
Dynamics VIO
ABSTRACT: Visual-inertial odometry (VIO) is the most common approach for estimating the
state of autonomous micro aerial vehicles using only onboard sensors. Existing
methods improve VIO performance by including a dynamics model in the estimation
pipeline. However, such methods degrade in the presence of low-fidelity vehicle
models and continuous external disturbances, such as wind. Our proposed method,
HDVIO, overcomes these limitations by using a hybrid dynamics model that
combines a point-mass vehicle model with a learning-based component that
captures complex aerodynamic effects. HDVIO estimates the external force and
the full robot state by leveraging the discrepancy between the actual motion
and the predicted motion of the hybrid dynamics model. Our hybrid dynamics
model uses a history of thrust and IMU measurements to predict the vehicle
dynamics. To demonstrate the performance of our method, we present results on
both public and novel drone dynamics datasets and show real-world experiments
of a quadrotor flying in strong winds up to 25 km/h. The results show that our
approach improves the motion and external force estimation compared to the
state-of-the-art by up to 33% and 40%, respectively. Furthermore, differently
from existing methods, we show that it is possible to predict the vehicle
dynamics accurately while having no explicit knowledge of its full state.
| no_new_dataset | 0.707847 |
2306.14203 | Shester Gueuwou | Shester Gueuwou and Rose-Mary Owusuaa Mensah Gyening | Stance Prediction and Analysis of Twitter data : A case study of Ghana
2020 Presidential Elections | Not adequate | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | On December 7, 2020, Ghanaians participated in the polls to determine their
president for the next four years. To gain insights from this presidential
election, we conducted stance analysis (which is not always equivalent to
sentiment analysis) to understand how Twitter, a popular social media platform,
reflected the opinions of its users regarding the two main presidential
candidates. We collected a total of 99,356 tweets using the Twitter API
(Tweepy) and manually annotated 3,090 tweets into three classes: Against,
Neutral, and Support. We then performed preprocessing on the tweets. The
resulting dataset was evaluated using two lexicon-based approaches, VADER and
TextBlob, as well as five supervised machine learning-based approaches: Support
Vector Machine (SVM), Logistic Regression (LR), Multinomial Na\"ive Bayes
(MNB), Stochastic Gradient Descent (SGD), and Random Forest (RF), based on
metrics such as accuracy, precision, recall, and F1-score. The best performance
was achieved by Logistic Regression with an accuracy of 71.13%. We utilized
Logistic Regression to classify all the extracted tweets and subsequently
conducted an analysis and discussion of the results. For access to our data and
code, please visit:
https://github.com/ShesterG/Stance-Detection-Ghana-2020-Elections.git
| [
{
"version": "v1",
"created": "Sun, 25 Jun 2023 10:54:24 GMT"
},
{
"version": "v2",
"created": "Tue, 27 Jun 2023 18:39:53 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Gueuwou",
"Shester",
""
],
[
"Gyening",
"Rose-Mary Owusuaa Mensah",
""
]
] | TITLE: Stance Prediction and Analysis of Twitter data : A case study of Ghana
2020 Presidential Elections
ABSTRACT: On December 7, 2020, Ghanaians participated in the polls to determine their
president for the next four years. To gain insights from this presidential
election, we conducted stance analysis (which is not always equivalent to
sentiment analysis) to understand how Twitter, a popular social media platform,
reflected the opinions of its users regarding the two main presidential
candidates. We collected a total of 99,356 tweets using the Twitter API
(Tweepy) and manually annotated 3,090 tweets into three classes: Against,
Neutral, and Support. We then performed preprocessing on the tweets. The
resulting dataset was evaluated using two lexicon-based approaches, VADER and
TextBlob, as well as five supervised machine learning-based approaches: Support
Vector Machine (SVM), Logistic Regression (LR), Multinomial Na\"ive Bayes
(MNB), Stochastic Gradient Descent (SGD), and Random Forest (RF), based on
metrics such as accuracy, precision, recall, and F1-score. The best performance
was achieved by Logistic Regression with an accuracy of 71.13%. We utilized
Logistic Regression to classify all the extracted tweets and subsequently
conducted an analysis and discussion of the results. For access to our data and
code, please visit:
https://github.com/ShesterG/Stance-Detection-Ghana-2020-Elections.git
| no_new_dataset | 0.70576 |
2306.14764 | Surendrabikram Thapa | Farhan Ahmad Jafri, Mohammad Aman Siddiqui, Surendrabikram Thapa,
Kritesh Rauniyar, Usman Naseem, Imran Razzak | Uncovering Political Hate Speech During Indian Election Campaign: A New
Low-Resource Dataset and Baselines | Accepted to ICWSM Workshop (MEDIATE) | null | null | null | cs.CL | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The detection of hate speech in political discourse is a critical issue, and
this becomes even more challenging in low-resource languages. To address this
issue, we introduce a new dataset named IEHate, which contains 11,457 manually
annotated Hindi tweets related to the Indian Assembly Election Campaign from
November 1, 2021, to March 9, 2022. We performed a detailed analysis of the
dataset, focusing on the prevalence of hate speech in political communication
and the different forms of hateful language used. Additionally, we benchmark
the dataset using a range of machine learning, deep learning, and
transformer-based algorithms. Our experiments reveal that the performance of
these models can be further improved, highlighting the need for more advanced
techniques for hate speech detection in low-resource languages. In particular,
the relatively higher score of human evaluation over algorithms emphasizes the
importance of utilizing both human and automated approaches for effective hate
speech moderation. Our IEHate dataset can serve as a valuable resource for
researchers and practitioners working on developing and evaluating hate speech
detection techniques in low-resource languages. Overall, our work underscores
the importance of addressing the challenges of identifying and mitigating hate
speech in political discourse, particularly in the context of low-resource
languages. The dataset and resources for this work are made available at
https://github.com/Farhan-jafri/Indian-Election.
| [
{
"version": "v1",
"created": "Mon, 26 Jun 2023 15:17:54 GMT"
},
{
"version": "v2",
"created": "Tue, 27 Jun 2023 16:55:14 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Jafri",
"Farhan Ahmad",
""
],
[
"Siddiqui",
"Mohammad Aman",
""
],
[
"Thapa",
"Surendrabikram",
""
],
[
"Rauniyar",
"Kritesh",
""
],
[
"Naseem",
"Usman",
""
],
[
"Razzak",
"Imran",
""
]
] | TITLE: Uncovering Political Hate Speech During Indian Election Campaign: A New
Low-Resource Dataset and Baselines
ABSTRACT: The detection of hate speech in political discourse is a critical issue, and
this becomes even more challenging in low-resource languages. To address this
issue, we introduce a new dataset named IEHate, which contains 11,457 manually
annotated Hindi tweets related to the Indian Assembly Election Campaign from
November 1, 2021, to March 9, 2022. We performed a detailed analysis of the
dataset, focusing on the prevalence of hate speech in political communication
and the different forms of hateful language used. Additionally, we benchmark
the dataset using a range of machine learning, deep learning, and
transformer-based algorithms. Our experiments reveal that the performance of
these models can be further improved, highlighting the need for more advanced
techniques for hate speech detection in low-resource languages. In particular,
the relatively higher score of human evaluation over algorithms emphasizes the
importance of utilizing both human and automated approaches for effective hate
speech moderation. Our IEHate dataset can serve as a valuable resource for
researchers and practitioners working on developing and evaluating hate speech
detection techniques in low-resource languages. Overall, our work underscores
the importance of addressing the challenges of identifying and mitigating hate
speech in political discourse, particularly in the context of low-resource
languages. The dataset and resources for this work are made available at
https://github.com/Farhan-jafri/Indian-Election.
| new_dataset | 0.715182 |
2306.15377 | Mehmet Kerim Yucel | Evangelos Skartados, Konstantinos Georgiadis, Mehmet Kerim Yucel,
Koskinas Ioannis, Armando Domi, Anastasios Drosou, Bruno Manganelli, Albert
Saa-Garriga | TrickVOS: A Bag of Tricks for Video Object Segmentation | Accepted to ICIP 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Space-time memory (STM) network methods have been dominant in semi-supervised
video object segmentation (SVOS) due to their remarkable performance. In this
work, we identify three key aspects where we can improve such methods; i)
supervisory signal, ii) pretraining and iii) spatial awareness. We then propose
TrickVOS; a generic, method-agnostic bag of tricks addressing each aspect with
i) a structure-aware hybrid loss, ii) a simple decoder pretraining regime and
iii) a cheap tracker that imposes spatial constraints in model predictions.
Finally, we propose a lightweight network and show that when trained with
TrickVOS, it achieves competitive results to state-of-the-art methods on DAVIS
and YouTube benchmarks, while being one of the first STM-based SVOS methods
that can run in real-time on a mobile device.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 10:54:08 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Jun 2023 12:57:11 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Skartados",
"Evangelos",
""
],
[
"Georgiadis",
"Konstantinos",
""
],
[
"Yucel",
"Mehmet Kerim",
""
],
[
"Ioannis",
"Koskinas",
""
],
[
"Domi",
"Armando",
""
],
[
"Drosou",
"Anastasios",
""
],
[
"Manganelli",
"Bruno",
""
],
[
"Saa-Garriga",
"Albert",
""
]
] | TITLE: TrickVOS: A Bag of Tricks for Video Object Segmentation
ABSTRACT: Space-time memory (STM) network methods have been dominant in semi-supervised
video object segmentation (SVOS) due to their remarkable performance. In this
work, we identify three key aspects where we can improve such methods; i)
supervisory signal, ii) pretraining and iii) spatial awareness. We then propose
TrickVOS; a generic, method-agnostic bag of tricks addressing each aspect with
i) a structure-aware hybrid loss, ii) a simple decoder pretraining regime and
iii) a cheap tracker that imposes spatial constraints in model predictions.
Finally, we propose a lightweight network and show that when trained with
TrickVOS, it achieves competitive results to state-of-the-art methods on DAVIS
and YouTube benchmarks, while being one of the first STM-based SVOS methods
that can run in real-time on a mobile device.
| no_new_dataset | 0.707588 |
2306.15692 | Arnav Kartikeya | Evan Kellener, Ihina Nath, An Ngo, Thomas Nguyen, Joshua Schuman, Coen
Adler, Arnav Kartikeya | Utilizing Segment Anything Model For Assessing Localization of GRAD-CAM
in Medical Imaging | 11 pages, 14 figures, 1 table | null | null | null | eess.IV cs.CV | http://creativecommons.org/publicdomain/zero/1.0/ | The introduction of saliency map algorithms as an approach for assessing the
interoperability of images has allowed for a deeper understanding of current
black-box models with Artificial Intelligence. Their rise in popularity has led
to these algorithms being applied in multiple fields, including medical
imaging. With a classification task as important as those in the medical
domain, a need for rigorous testing of their capabilities arises. Current works
examine capabilities through assessing the localization of saliency maps upon
medical abnormalities within an image, through comparisons with human
annotations. We propose utilizing Segment Anything Model (SAM) to both further
the accuracy of such existing metrics, while also generalizing beyond the need
for human annotations. Our results show both high degrees of similarity to
existing metrics while also highlighting the capabilities of this methodology
to beyond human-annotation. Furthermore, we explore the applications (and
challenges) of SAM within the medical domain, including image pre-processing
before segmenting, natural language proposals to SAM in the form of CLIP-SAM,
and SAM accuracy across multiple medical imaging datasets.
| [
{
"version": "v1",
"created": "Sat, 24 Jun 2023 19:54:50 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Kellener",
"Evan",
""
],
[
"Nath",
"Ihina",
""
],
[
"Ngo",
"An",
""
],
[
"Nguyen",
"Thomas",
""
],
[
"Schuman",
"Joshua",
""
],
[
"Adler",
"Coen",
""
],
[
"Kartikeya",
"Arnav",
""
]
] | TITLE: Utilizing Segment Anything Model For Assessing Localization of GRAD-CAM
in Medical Imaging
ABSTRACT: The introduction of saliency map algorithms as an approach for assessing the
interoperability of images has allowed for a deeper understanding of current
black-box models with Artificial Intelligence. Their rise in popularity has led
to these algorithms being applied in multiple fields, including medical
imaging. With a classification task as important as those in the medical
domain, a need for rigorous testing of their capabilities arises. Current works
examine capabilities through assessing the localization of saliency maps upon
medical abnormalities within an image, through comparisons with human
annotations. We propose utilizing Segment Anything Model (SAM) to both further
the accuracy of such existing metrics, while also generalizing beyond the need
for human annotations. Our results show both high degrees of similarity to
existing metrics while also highlighting the capabilities of this methodology
to beyond human-annotation. Furthermore, we explore the applications (and
challenges) of SAM within the medical domain, including image pre-processing
before segmenting, natural language proposals to SAM in the form of CLIP-SAM,
and SAM accuracy across multiple medical imaging datasets.
| no_new_dataset | 0.711425 |
2306.15704 | Rui He | Yuanxi Sun, Rui He, Youzeng Li, Zuwei Huang, Feng Hu, Xu Cheng, Jie
Tang | MAE-GEBD:Winning the CVPR'2023 LOVEU-GEBD Challenge | Winner of CVPR2023 LOVEU GEBD Challenge | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The Generic Event Boundary Detection (GEBD) task aims to build a model for
segmenting videos into segments by detecting general event boundaries
applicable to various classes. In this paper, based on last year's MAE-GEBD
method, we have improved our model performance on the GEBD task by adjusting
the data processing strategy and loss function. Based on last year's approach,
we extended the application of pseudo-label to a larger dataset and made many
experimental attempts. In addition, we applied focal loss to concentrate more
on difficult samples and improved our model performance. Finally, we improved
the segmentation alignment strategy used last year, and dynamically adjusted
the segmentation alignment method according to the boundary density and
duration of the video, so that our model can be more flexible and fully
applicable in different situations. With our method, we achieve an F1 score of
86.03% on the Kinetics-GEBD test set, which is a 0.09% improvement in the F1
score compared to our 2022 Kinetics-GEBD method.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 02:35:19 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Sun",
"Yuanxi",
""
],
[
"He",
"Rui",
""
],
[
"Li",
"Youzeng",
""
],
[
"Huang",
"Zuwei",
""
],
[
"Hu",
"Feng",
""
],
[
"Cheng",
"Xu",
""
],
[
"Tang",
"Jie",
""
]
] | TITLE: MAE-GEBD:Winning the CVPR'2023 LOVEU-GEBD Challenge
ABSTRACT: The Generic Event Boundary Detection (GEBD) task aims to build a model for
segmenting videos into segments by detecting general event boundaries
applicable to various classes. In this paper, based on last year's MAE-GEBD
method, we have improved our model performance on the GEBD task by adjusting
the data processing strategy and loss function. Based on last year's approach,
we extended the application of pseudo-label to a larger dataset and made many
experimental attempts. In addition, we applied focal loss to concentrate more
on difficult samples and improved our model performance. Finally, we improved
the segmentation alignment strategy used last year, and dynamically adjusted
the segmentation alignment method according to the boundary density and
duration of the video, so that our model can be more flexible and fully
applicable in different situations. With our method, we achieve an F1 score of
86.03% on the Kinetics-GEBD test set, which is a 0.09% improvement in the F1
score compared to our 2022 Kinetics-GEBD method.
| no_new_dataset | 0.711268 |
2306.15713 | Chris Zhang | Chris Zhang, Runsheng Guo, Wenyuan Zeng, Yuwen Xiong, Binbin Dai, Rui
Hu, Mengye Ren, Raquel Urtasun | Rethinking Closed-loop Training for Autonomous Driving | ECCV 2022 | null | 10.1007/978-3-031-19842-7_16 | null | cs.RO cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advances in high-fidelity simulators have enabled closed-loop training
of autonomous driving agents, potentially solving the distribution shift in
training v.s. deployment and allowing training to be scaled both safely and
cheaply. However, there is a lack of understanding of how to build effective
training benchmarks for closed-loop training. In this work, we present the
first empirical study which analyzes the effects of different training
benchmark designs on the success of learning agents, such as how to design
traffic scenarios and scale training environments. Furthermore, we show that
many popular RL algorithms cannot achieve satisfactory performance in the
context of autonomous driving, as they lack long-term planning and take an
extremely long time to train. To address these issues, we propose trajectory
value learning (TRAVL), an RL-based driving agent that performs planning with
multistep look-ahead and exploits cheaply generated imagined data for efficient
learning. Our experiments show that TRAVL can learn much faster and produce
safer maneuvers compared to all the baselines. For more information, visit the
project website: https://waabi.ai/research/travl
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 17:58:39 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Zhang",
"Chris",
""
],
[
"Guo",
"Runsheng",
""
],
[
"Zeng",
"Wenyuan",
""
],
[
"Xiong",
"Yuwen",
""
],
[
"Dai",
"Binbin",
""
],
[
"Hu",
"Rui",
""
],
[
"Ren",
"Mengye",
""
],
[
"Urtasun",
"Raquel",
""
]
] | TITLE: Rethinking Closed-loop Training for Autonomous Driving
ABSTRACT: Recent advances in high-fidelity simulators have enabled closed-loop training
of autonomous driving agents, potentially solving the distribution shift in
training v.s. deployment and allowing training to be scaled both safely and
cheaply. However, there is a lack of understanding of how to build effective
training benchmarks for closed-loop training. In this work, we present the
first empirical study which analyzes the effects of different training
benchmark designs on the success of learning agents, such as how to design
traffic scenarios and scale training environments. Furthermore, we show that
many popular RL algorithms cannot achieve satisfactory performance in the
context of autonomous driving, as they lack long-term planning and take an
extremely long time to train. To address these issues, we propose trajectory
value learning (TRAVL), an RL-based driving agent that performs planning with
multistep look-ahead and exploits cheaply generated imagined data for efficient
learning. Our experiments show that TRAVL can learn much faster and produce
safer maneuvers compared to all the baselines. For more information, visit the
project website: https://waabi.ai/research/travl
| no_new_dataset | 0.708389 |
2306.15732 | Michael Yoder | Michael Miller Yoder, Ahmad Diab, David West Brown, Kathleen M. Carley | A Weakly Supervised Classifier and Dataset of White Supremacist Language | ACL 2023 short | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | We present a dataset and classifier for detecting the language of white
supremacist extremism, a growing issue in online hate speech. Our weakly
supervised classifier is trained on large datasets of text from explicitly
white supremacist domains paired with neutral and anti-racist data from similar
domains. We demonstrate that this approach improves generalization performance
to new domains. Incorporating anti-racist texts as counterexamples to white
supremacist language mitigates bias.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 18:19:32 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yoder",
"Michael Miller",
""
],
[
"Diab",
"Ahmad",
""
],
[
"Brown",
"David West",
""
],
[
"Carley",
"Kathleen M.",
""
]
] | TITLE: A Weakly Supervised Classifier and Dataset of White Supremacist Language
ABSTRACT: We present a dataset and classifier for detecting the language of white
supremacist extremism, a growing issue in online hate speech. Our weakly
supervised classifier is trained on large datasets of text from explicitly
white supremacist domains paired with neutral and anti-racist data from similar
domains. We demonstrate that this approach improves generalization performance
to new domains. Incorporating anti-racist texts as counterexamples to white
supremacist language mitigates bias.
| new_dataset | 0.714267 |
2306.15733 | Marija Ivanovska | Marija Ivanovska, Vitomir \v{S}truc | Face Morphing Attack Detection with Denoising Diffusion Probabilistic
Models | Published at IWBF 2023 | 2023 11th International Workshop on Biometrics and Forensics
(IWBF) | 10.1109/IWBF57495.2023.10156877 | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Morphed face images have recently become a growing concern for existing face
verification systems, as they are relatively easy to generate and can be used
to impersonate someone's identity for various malicious purposes. Efficient
Morphing Attack Detection (MAD) that generalizes well across different morphing
techniques is, therefore, of paramount importance. Existing MAD techniques
predominantly rely on discriminative models that learn from examples of bona
fide and morphed images and, as a result, often exhibit sub-optimal
generalization performance when confronted with unknown types of morphing
attacks. To address this problem, we propose a novel, diffusion-based MAD
method in this paper that learns only from the characteristics of bona fide
images. Various forms of morphing attacks are then detected by our model as
out-of-distribution samples. We perform rigorous experiments over four
different datasets (CASIA-WebFace, FRLL-Morphs, FERET-Morphs and FRGC-Morphs)
and compare the proposed solution to both discriminatively-trained and
once-class MAD models. The experimental results show that our MAD model
achieves highly competitive results on all considered datasets.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 18:19:45 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Ivanovska",
"Marija",
""
],
[
"Štruc",
"Vitomir",
""
]
] | TITLE: Face Morphing Attack Detection with Denoising Diffusion Probabilistic
Models
ABSTRACT: Morphed face images have recently become a growing concern for existing face
verification systems, as they are relatively easy to generate and can be used
to impersonate someone's identity for various malicious purposes. Efficient
Morphing Attack Detection (MAD) that generalizes well across different morphing
techniques is, therefore, of paramount importance. Existing MAD techniques
predominantly rely on discriminative models that learn from examples of bona
fide and morphed images and, as a result, often exhibit sub-optimal
generalization performance when confronted with unknown types of morphing
attacks. To address this problem, we propose a novel, diffusion-based MAD
method in this paper that learns only from the characteristics of bona fide
images. Various forms of morphing attacks are then detected by our model as
out-of-distribution samples. We perform rigorous experiments over four
different datasets (CASIA-WebFace, FRLL-Morphs, FERET-Morphs and FRGC-Morphs)
and compare the proposed solution to both discriminatively-trained and
once-class MAD models. The experimental results show that our MAD model
achieves highly competitive results on all considered datasets.
| no_new_dataset | 0.71081 |
2306.15742 | Zelun Luo | Zelun Luo, Yuliang Zou, Yijin Yang, Zane Durante, De-An Huang, Zhiding
Yu, Chaowei Xiao, Li Fei-Fei, Animashree Anandkumar | Differentially Private Video Activity Recognition | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In recent years, differential privacy has seen significant advancements in
image classification; however, its application to video activity recognition
remains under-explored. This paper addresses the challenges of applying
differential privacy to video activity recognition, which primarily stem from:
(1) a discrepancy between the desired privacy level for entire videos and the
nature of input data processed by contemporary video architectures, which are
typically short, segmented clips; and (2) the complexity and sheer size of
video datasets relative to those in image classification, which render
traditional differential privacy methods inadequate. To tackle these issues, we
propose Multi-Clip DP-SGD, a novel framework for enforcing video-level
differential privacy through clip-based classification models. This method
samples multiple clips from each video, averages their gradients, and applies
gradient clipping in DP-SGD without incurring additional privacy loss.
Moreover, we incorporate a parameter-efficient transfer learning strategy to
make the model scalable for large-scale video datasets. Through extensive
evaluations on the UCF-101 and HMDB-51 datasets, our approach exhibits
impressive performance, achieving 81% accuracy with a privacy budget of
epsilon=5 on UCF-101, marking a 76% improvement compared to a direct
application of DP-SGD. Furthermore, we demonstrate that our transfer learning
strategy is versatile and can enhance differentially private image
classification across an array of datasets including CheXpert, ImageNet,
CIFAR-10, and CIFAR-100.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 18:47:09 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Luo",
"Zelun",
""
],
[
"Zou",
"Yuliang",
""
],
[
"Yang",
"Yijin",
""
],
[
"Durante",
"Zane",
""
],
[
"Huang",
"De-An",
""
],
[
"Yu",
"Zhiding",
""
],
[
"Xiao",
"Chaowei",
""
],
[
"Fei-Fei",
"Li",
""
],
[
"Anandkumar",
"Animashree",
""
]
] | TITLE: Differentially Private Video Activity Recognition
ABSTRACT: In recent years, differential privacy has seen significant advancements in
image classification; however, its application to video activity recognition
remains under-explored. This paper addresses the challenges of applying
differential privacy to video activity recognition, which primarily stem from:
(1) a discrepancy between the desired privacy level for entire videos and the
nature of input data processed by contemporary video architectures, which are
typically short, segmented clips; and (2) the complexity and sheer size of
video datasets relative to those in image classification, which render
traditional differential privacy methods inadequate. To tackle these issues, we
propose Multi-Clip DP-SGD, a novel framework for enforcing video-level
differential privacy through clip-based classification models. This method
samples multiple clips from each video, averages their gradients, and applies
gradient clipping in DP-SGD without incurring additional privacy loss.
Moreover, we incorporate a parameter-efficient transfer learning strategy to
make the model scalable for large-scale video datasets. Through extensive
evaluations on the UCF-101 and HMDB-51 datasets, our approach exhibits
impressive performance, achieving 81% accuracy with a privacy budget of
epsilon=5 on UCF-101, marking a 76% improvement compared to a direct
application of DP-SGD. Furthermore, we demonstrate that our transfer learning
strategy is versatile and can enhance differentially private image
classification across an array of datasets including CheXpert, ImageNet,
CIFAR-10, and CIFAR-100.
| no_new_dataset | 0.710051 |
2306.15744 | Ayush Sekhari | Badih Ghazi, Pritish Kamath, Ravi Kumar, Pasin Manurangsi, Ayush
Sekhari, Chiyuan Zhang | Ticketed Learning-Unlearning Schemes | Conference on Learning Theory (COLT) 2023 | null | null | null | cs.LG cs.DS stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the learning--unlearning paradigm defined as follows. First given
a dataset, the goal is to learn a good predictor, such as one minimizing a
certain loss. Subsequently, given any subset of examples that wish to be
unlearnt, the goal is to learn, without the knowledge of the original training
dataset, a good predictor that is identical to the predictor that would have
been produced when learning from scratch on the surviving examples.
We propose a new ticketed model for learning--unlearning wherein the learning
algorithm can send back additional information in the form of a small-sized
(encrypted) ``ticket'' to each participating training example, in addition to
retaining a small amount of ``central'' information for later. Subsequently,
the examples that wish to be unlearnt present their tickets to the unlearning
algorithm, which additionally uses the central information to return a new
predictor. We provide space-efficient ticketed learning--unlearning schemes for
a broad family of concept classes, including thresholds, parities,
intersection-closed classes, among others.
En route, we introduce the count-to-zero problem, where during unlearning,
the goal is to simply know if there are any examples that survived. We give a
ticketed learning--unlearning scheme for this problem that relies on the
construction of Sperner families with certain properties, which might be of
independent interest.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 18:54:40 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Ghazi",
"Badih",
""
],
[
"Kamath",
"Pritish",
""
],
[
"Kumar",
"Ravi",
""
],
[
"Manurangsi",
"Pasin",
""
],
[
"Sekhari",
"Ayush",
""
],
[
"Zhang",
"Chiyuan",
""
]
] | TITLE: Ticketed Learning-Unlearning Schemes
ABSTRACT: We consider the learning--unlearning paradigm defined as follows. First given
a dataset, the goal is to learn a good predictor, such as one minimizing a
certain loss. Subsequently, given any subset of examples that wish to be
unlearnt, the goal is to learn, without the knowledge of the original training
dataset, a good predictor that is identical to the predictor that would have
been produced when learning from scratch on the surviving examples.
We propose a new ticketed model for learning--unlearning wherein the learning
algorithm can send back additional information in the form of a small-sized
(encrypted) ``ticket'' to each participating training example, in addition to
retaining a small amount of ``central'' information for later. Subsequently,
the examples that wish to be unlearnt present their tickets to the unlearning
algorithm, which additionally uses the central information to return a new
predictor. We provide space-efficient ticketed learning--unlearning schemes for
a broad family of concept classes, including thresholds, parities,
intersection-closed classes, among others.
En route, we introduce the count-to-zero problem, where during unlearning,
the goal is to simply know if there are any examples that survived. We give a
ticketed learning--unlearning scheme for this problem that relies on the
construction of Sperner families with certain properties, which might be of
independent interest.
| no_new_dataset | 0.710239 |
2306.15765 | Hari Mohan Pandey | Santosh Kumar Yadav, Muhtashim Rafiqi, Egna Praneeth Gummana, Kamlesh
Tiwari, Hari Mohan Pandey, Shaik Ali Akbara | A Novel Two Stream Decision Level Fusion of Vision and Inertial Sensors
Data for Automatic Multimodal Human Activity Recognition System | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper presents a novel multimodal human activity recognition system. It
uses a two-stream decision level fusion of vision and inertial sensors. In the
first stream, raw RGB frames are passed to a part affinity field-based pose
estimation network to detect the keypoints of the user. These keypoints are
then pre-processed and inputted in a sliding window fashion to a specially
designed convolutional neural network for the spatial feature extraction
followed by regularized LSTMs to calculate the temporal features. The outputs
of LSTM networks are then inputted to fully connected layers for
classification. In the second stream, data obtained from inertial sensors are
pre-processed and inputted to regularized LSTMs for the feature extraction
followed by fully connected layers for the classification. At this stage, the
SoftMax scores of two streams are then fused using the decision level fusion
which gives the final prediction. Extensive experiments are conducted to
evaluate the performance. Four multimodal standard benchmark datasets (UP-Fall
detection, UTD-MHAD, Berkeley-MHAD, and C-MHAD) are used for experimentations.
The accuracies obtained by the proposed system are 96.9 %, 97.6 %, 98.7 %, and
95.9 % respectively on the UP-Fall Detection, UTDMHAD, Berkeley-MHAD, and
C-MHAD datasets. These results are far superior than the current
state-of-the-art methods.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 19:29:35 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yadav",
"Santosh Kumar",
""
],
[
"Rafiqi",
"Muhtashim",
""
],
[
"Gummana",
"Egna Praneeth",
""
],
[
"Tiwari",
"Kamlesh",
""
],
[
"Pandey",
"Hari Mohan",
""
],
[
"Akbara",
"Shaik Ali",
""
]
] | TITLE: A Novel Two Stream Decision Level Fusion of Vision and Inertial Sensors
Data for Automatic Multimodal Human Activity Recognition System
ABSTRACT: This paper presents a novel multimodal human activity recognition system. It
uses a two-stream decision level fusion of vision and inertial sensors. In the
first stream, raw RGB frames are passed to a part affinity field-based pose
estimation network to detect the keypoints of the user. These keypoints are
then pre-processed and inputted in a sliding window fashion to a specially
designed convolutional neural network for the spatial feature extraction
followed by regularized LSTMs to calculate the temporal features. The outputs
of LSTM networks are then inputted to fully connected layers for
classification. In the second stream, data obtained from inertial sensors are
pre-processed and inputted to regularized LSTMs for the feature extraction
followed by fully connected layers for the classification. At this stage, the
SoftMax scores of two streams are then fused using the decision level fusion
which gives the final prediction. Extensive experiments are conducted to
evaluate the performance. Four multimodal standard benchmark datasets (UP-Fall
detection, UTD-MHAD, Berkeley-MHAD, and C-MHAD) are used for experimentations.
The accuracies obtained by the proposed system are 96.9 %, 97.6 %, 98.7 %, and
95.9 % respectively on the UP-Fall Detection, UTDMHAD, Berkeley-MHAD, and
C-MHAD datasets. These results are far superior than the current
state-of-the-art methods.
| no_new_dataset | 0.712001 |
2306.15768 | Hari Mohan Pandey | Santosh Kumar Yadav, Apurv Shukla, Kamlesh Tiwari, Hari Mohan Pandey,
Shaik Ali Akbar | An Efficient Deep Convolutional Neural Network Model For Yoga Pose
Recognition Using Single Images | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Pose recognition deals with designing algorithms to locate human body joints
in a 2D/3D space and run inference on the estimated joint locations for
predicting the poses. Yoga poses consist of some very complex postures. It
imposes various challenges on the computer vision algorithms like occlusion,
inter-class similarity, intra-class variability, viewpoint complexity, etc.
This paper presents YPose, an efficient deep convolutional neural network (CNN)
model to recognize yoga asanas from RGB images. The proposed model consists of
four steps as follows: (a) first, the region of interest (ROI) is segmented
using segmentation based approaches to extract the ROI from the original
images; (b) second, these refined images are passed to a CNN architecture based
on the backbone of EfficientNets for feature extraction; (c) third, dense
refinement blocks, adapted from the architecture of densely connected networks
are added to learn more diversified features; and (d) fourth, global average
pooling and fully connected layers are applied for the classification of the
multi-level hierarchy of the yoga poses. The proposed model has been tested on
the Yoga-82 dataset. It is a publicly available benchmark dataset for yoga pose
recognition. Experimental results show that the proposed model achieves the
state-of-the-art on this dataset. The proposed model obtained an accuracy of
93.28%, which is an improvement over the earlier state-of-the-art (79.35%) with
a margin of approximately 13.9%. The code will be made publicly available.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 19:34:46 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yadav",
"Santosh Kumar",
""
],
[
"Shukla",
"Apurv",
""
],
[
"Tiwari",
"Kamlesh",
""
],
[
"Pandey",
"Hari Mohan",
""
],
[
"Akbar",
"Shaik Ali",
""
]
] | TITLE: An Efficient Deep Convolutional Neural Network Model For Yoga Pose
Recognition Using Single Images
ABSTRACT: Pose recognition deals with designing algorithms to locate human body joints
in a 2D/3D space and run inference on the estimated joint locations for
predicting the poses. Yoga poses consist of some very complex postures. It
imposes various challenges on the computer vision algorithms like occlusion,
inter-class similarity, intra-class variability, viewpoint complexity, etc.
This paper presents YPose, an efficient deep convolutional neural network (CNN)
model to recognize yoga asanas from RGB images. The proposed model consists of
four steps as follows: (a) first, the region of interest (ROI) is segmented
using segmentation based approaches to extract the ROI from the original
images; (b) second, these refined images are passed to a CNN architecture based
on the backbone of EfficientNets for feature extraction; (c) third, dense
refinement blocks, adapted from the architecture of densely connected networks
are added to learn more diversified features; and (d) fourth, global average
pooling and fully connected layers are applied for the classification of the
multi-level hierarchy of the yoga poses. The proposed model has been tested on
the Yoga-82 dataset. It is a publicly available benchmark dataset for yoga pose
recognition. Experimental results show that the proposed model achieves the
state-of-the-art on this dataset. The proposed model obtained an accuracy of
93.28%, which is an improvement over the earlier state-of-the-art (79.35%) with
a margin of approximately 13.9%. The code will be made publicly available.
| no_new_dataset | 0.687814 |
2306.15773 | Naveen Namashivayam | Naveen Namashivayam, Krishna Kandalla, James B White III, Larry
Kaplan, Mark Pagel | Exploring Fully Offloaded GPU Stream-Aware Message Passing | 12 pages, 17 figures | null | null | null | cs.DC cs.NI cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern heterogeneous supercomputing systems are comprised of CPUs, GPUs, and
high-speed network interconnects. Communication libraries supporting efficient
data transfers involving memory buffers from the GPU memory typically require
the CPU to orchestrate the data transfer operations. A new offload-friendly
communication strategy, stream-triggered (ST) communication, was explored to
allow offloading the synchronization and data movement operations from the CPU
to the GPU. A Message Passing Interface (MPI) one-sided active target
synchronization based implementation was used as an exemplar to illustrate the
proposed strategy. A latency-sensitive nearest neighbor microbenchmark was used
to explore the various performance aspects of the implementation. The offloaded
implementation shows significant on-node performance advantages over standard
MPI active RMA (36%) and point-to-point (61%) communication. The current
multi-node improvement is less (23% faster than standard active RMA but 11%
slower than point-to-point), but plans are in progress to purse further
improvements.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 19:49:30 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Namashivayam",
"Naveen",
""
],
[
"Kandalla",
"Krishna",
""
],
[
"White",
"James B",
"III"
],
[
"Kaplan",
"Larry",
""
],
[
"Pagel",
"Mark",
""
]
] | TITLE: Exploring Fully Offloaded GPU Stream-Aware Message Passing
ABSTRACT: Modern heterogeneous supercomputing systems are comprised of CPUs, GPUs, and
high-speed network interconnects. Communication libraries supporting efficient
data transfers involving memory buffers from the GPU memory typically require
the CPU to orchestrate the data transfer operations. A new offload-friendly
communication strategy, stream-triggered (ST) communication, was explored to
allow offloading the synchronization and data movement operations from the CPU
to the GPU. A Message Passing Interface (MPI) one-sided active target
synchronization based implementation was used as an exemplar to illustrate the
proposed strategy. A latency-sensitive nearest neighbor microbenchmark was used
to explore the various performance aspects of the implementation. The offloaded
implementation shows significant on-node performance advantages over standard
MPI active RMA (36%) and point-to-point (61%) communication. The current
multi-node improvement is less (23% faster than standard active RMA but 11%
slower than point-to-point), but plans are in progress to purse further
improvements.
| no_new_dataset | 0.710007 |
2306.15777 | Alicia Parrish | Alicia Parrish, Sarah Laszlo, Lora Aroyo | "Is a picture of a bird a bird": Policy recommendations for dealing with
ambiguity in machine vision models | null | null | null | null | cs.CY cs.CV cs.HC | http://creativecommons.org/licenses/by-sa/4.0/ | Many questions that we ask about the world do not have a single clear answer,
yet typical human annotation set-ups in machine learning assume there must be a
single ground truth label for all examples in every task. The divergence
between reality and practice is stark, especially in cases with inherent
ambiguity and where the range of different subjective judgments is wide. Here,
we examine the implications of subjective human judgments in the behavioral
task of labeling images used to train machine vision models. We identify three
primary sources of ambiguity arising from (i) depictions of labels in the
images, (ii) raters' backgrounds, and (iii) the task definition. On the basis
of the empirical results, we suggest best practices for handling label
ambiguity in machine learning datasets.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 19:59:33 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Parrish",
"Alicia",
""
],
[
"Laszlo",
"Sarah",
""
],
[
"Aroyo",
"Lora",
""
]
] | TITLE: "Is a picture of a bird a bird": Policy recommendations for dealing with
ambiguity in machine vision models
ABSTRACT: Many questions that we ask about the world do not have a single clear answer,
yet typical human annotation set-ups in machine learning assume there must be a
single ground truth label for all examples in every task. The divergence
between reality and practice is stark, especially in cases with inherent
ambiguity and where the range of different subjective judgments is wide. Here,
we examine the implications of subjective human judgments in the behavioral
task of labeling images used to train machine vision models. We identify three
primary sources of ambiguity arising from (i) depictions of labels in the
images, (ii) raters' backgrounds, and (iii) the task definition. On the basis
of the empirical results, we suggest best practices for handling label
ambiguity in machine learning datasets.
| no_new_dataset | 0.711005 |
2306.15790 | Edward Raff | Tyler LeBlond, Joseph Munoz, Fred Lu, Maya Fuchs, Elliott
Zaresky-Williams, Edward Raff, Brian Testa | Probing the Transition to Dataset-Level Privacy in ML Models Using an
Output-Specific and Data-Resolved Privacy Profile | Approved for Public Release; Distribution Unlimited. PA
#:AFRL-2022-3639 | null | null | null | cs.LG cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Differential privacy (DP) is the prevailing technique for protecting user
data in machine learning models. However, deficits to this framework include a
lack of clarity for selecting the privacy budget $\epsilon$ and a lack of
quantification for the privacy leakage for a particular data row by a
particular trained model. We make progress toward these limitations and a new
perspective by which to visualize DP results by studying a privacy metric that
quantifies the extent to which a model trained on a dataset using a DP
mechanism is ``covered" by each of the distributions resulting from training on
neighboring datasets. We connect this coverage metric to what has been
established in the literature and use it to rank the privacy of individual
samples from the training set in what we call a privacy profile. We
additionally show that the privacy profile can be used to probe an observed
transition to indistinguishability that takes place in the neighboring
distributions as $\epsilon$ decreases, which we suggest is a tool that can
enable the selection of $\epsilon$ by the ML practitioner wishing to make use
of DP.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 20:39:07 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"LeBlond",
"Tyler",
""
],
[
"Munoz",
"Joseph",
""
],
[
"Lu",
"Fred",
""
],
[
"Fuchs",
"Maya",
""
],
[
"Zaresky-Williams",
"Elliott",
""
],
[
"Raff",
"Edward",
""
],
[
"Testa",
"Brian",
""
]
] | TITLE: Probing the Transition to Dataset-Level Privacy in ML Models Using an
Output-Specific and Data-Resolved Privacy Profile
ABSTRACT: Differential privacy (DP) is the prevailing technique for protecting user
data in machine learning models. However, deficits to this framework include a
lack of clarity for selecting the privacy budget $\epsilon$ and a lack of
quantification for the privacy leakage for a particular data row by a
particular trained model. We make progress toward these limitations and a new
perspective by which to visualize DP results by studying a privacy metric that
quantifies the extent to which a model trained on a dataset using a DP
mechanism is ``covered" by each of the distributions resulting from training on
neighboring datasets. We connect this coverage metric to what has been
established in the literature and use it to rank the privacy of individual
samples from the training set in what we call a privacy profile. We
additionally show that the privacy profile can be used to probe an observed
transition to indistinguishability that takes place in the neighboring
distributions as $\epsilon$ decreases, which we suggest is a tool that can
enable the selection of $\epsilon$ by the ML practitioner wishing to make use
of DP.
| no_new_dataset | 0.710842 |
2306.15796 | Yakun Yu | Yakun Yu, Mingjun Zhao, Shi-ang Qi, Feiran Sun, Baoxun Wang, Weidong
Guo, Xiaoli Wang, Lei Yang, Di Niu | ConKI: Contrastive Knowledge Injection for Multimodal Sentiment Analysis | Accepted by ACL Findings 2023 | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multimodal Sentiment Analysis leverages multimodal signals to detect the
sentiment of a speaker. Previous approaches concentrate on performing
multimodal fusion and representation learning based on general knowledge
obtained from pretrained models, which neglects the effect of domain-specific
knowledge. In this paper, we propose Contrastive Knowledge Injection (ConKI)
for multimodal sentiment analysis, where specific-knowledge representations for
each modality can be learned together with general knowledge representations
via knowledge injection based on an adapter architecture. In addition, ConKI
uses a hierarchical contrastive learning procedure performed between knowledge
types within every single modality, across modalities within each sample, and
across samples to facilitate the effective learning of the proposed
representations, hence improving multimodal sentiment predictions. The
experiments on three popular multimodal sentiment analysis benchmarks show that
ConKI outperforms all prior methods on a variety of performance metrics.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 20:51:03 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yu",
"Yakun",
""
],
[
"Zhao",
"Mingjun",
""
],
[
"Qi",
"Shi-ang",
""
],
[
"Sun",
"Feiran",
""
],
[
"Wang",
"Baoxun",
""
],
[
"Guo",
"Weidong",
""
],
[
"Wang",
"Xiaoli",
""
],
[
"Yang",
"Lei",
""
],
[
"Niu",
"Di",
""
]
] | TITLE: ConKI: Contrastive Knowledge Injection for Multimodal Sentiment Analysis
ABSTRACT: Multimodal Sentiment Analysis leverages multimodal signals to detect the
sentiment of a speaker. Previous approaches concentrate on performing
multimodal fusion and representation learning based on general knowledge
obtained from pretrained models, which neglects the effect of domain-specific
knowledge. In this paper, we propose Contrastive Knowledge Injection (ConKI)
for multimodal sentiment analysis, where specific-knowledge representations for
each modality can be learned together with general knowledge representations
via knowledge injection based on an adapter architecture. In addition, ConKI
uses a hierarchical contrastive learning procedure performed between knowledge
types within every single modality, across modalities within each sample, and
across samples to facilitate the effective learning of the proposed
representations, hence improving multimodal sentiment predictions. The
experiments on three popular multimodal sentiment analysis benchmarks show that
ConKI outperforms all prior methods on a variety of performance metrics.
| no_new_dataset | 0.710208 |
2306.15803 | Jordi Planes | Ram\'on B\'ejar and Ant\'onio Morgado and Jordi Planes and Joao
Marques-Silva | On Logic-Based Explainability with Partially Specified Inputs | 14 pages | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | In the practical deployment of machine learning (ML) models, missing data
represents a recurring challenge. Missing data is often addressed when training
ML models. But missing data also needs to be addressed when deciding
predictions and when explaining those predictions. Missing data represents an
opportunity to partially specify the inputs of the prediction to be explained.
This paper studies the computation of logic-based explanations in the presence
of partially specified inputs. The paper shows that most of the algorithms
proposed in recent years for computing logic-based explanations can be
generalized for computing explanations given the partially specified inputs.
One related result is that the complexity of computing logic-based explanations
remains unchanged. A similar result is proved in the case of logic-based
explainability subject to input constraints. Furthermore, the proposed solution
for computing explanations given partially specified inputs is applied to
classifiers obtained from well-known public datasets, thereby illustrating a
number of novel explainability use cases.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 21:09:25 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Béjar",
"Ramón",
""
],
[
"Morgado",
"António",
""
],
[
"Planes",
"Jordi",
""
],
[
"Marques-Silva",
"Joao",
""
]
] | TITLE: On Logic-Based Explainability with Partially Specified Inputs
ABSTRACT: In the practical deployment of machine learning (ML) models, missing data
represents a recurring challenge. Missing data is often addressed when training
ML models. But missing data also needs to be addressed when deciding
predictions and when explaining those predictions. Missing data represents an
opportunity to partially specify the inputs of the prediction to be explained.
This paper studies the computation of logic-based explanations in the presence
of partially specified inputs. The paper shows that most of the algorithms
proposed in recent years for computing logic-based explanations can be
generalized for computing explanations given the partially specified inputs.
One related result is that the complexity of computing logic-based explanations
remains unchanged. A similar result is proved in the case of logic-based
explainability subject to input constraints. Furthermore, the proposed solution
for computing explanations given partially specified inputs is applied to
classifiers obtained from well-known public datasets, thereby illustrating a
number of novel explainability use cases.
| no_new_dataset | 0.710069 |
2306.15804 | Huimin Xu | Huimin Xu, Meijun Liu, Yi Bu, Shujing Sun, Yi Zhang, Chenwei Zhang,
Daniel E. Acuna, Steven Gray, Eric Meyer, Ying Ding | The Impact of Heterogeneous Shared Leadership in Scientific Teams | null | null | null | null | physics.soc-ph cs.CY | http://creativecommons.org/licenses/by/4.0/ | Leadership is evolving dynamically from an individual endeavor to shared
efforts. This paper aims to advance our understanding of shared leadership in
scientific teams. We define three kinds of leaders, junior (10-15), mid
(15-20), and senior (20+) based on career age. By considering the combinations
of any two leaders, we distinguish shared leadership as heterogeneous when
leaders are in different age cohorts and homogeneous when leaders are in the
same age cohort. Drawing on 1,845,351 CS, 254,039 Sociology, and 193,338
Business teams with two leaders in the OpenAlex dataset, we identify that
heterogeneous shared leadership brings higher citation impact for teams than
homogeneous shared leadership. Specifically, when junior leaders are paired
with senior leaders, it significantly increases team citation ranking by 1-2%,
in comparison with two leaders of similar age. We explore the patterns between
homogeneous leaders and heterogeneous leaders from team scale, expertise
composition, and knowledge recency perspectives. Compared with homogeneous
leaders, heterogeneous leaders are more adaptive in large teams, have more
diverse expertise, and trace both the newest and oldest references.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 21:10:00 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Xu",
"Huimin",
""
],
[
"Liu",
"Meijun",
""
],
[
"Bu",
"Yi",
""
],
[
"Sun",
"Shujing",
""
],
[
"Zhang",
"Yi",
""
],
[
"Zhang",
"Chenwei",
""
],
[
"Acuna",
"Daniel E.",
""
],
[
"Gray",
"Steven",
""
],
[
"Meyer",
"Eric",
""
],
[
"Ding",
"Ying",
""
]
] | TITLE: The Impact of Heterogeneous Shared Leadership in Scientific Teams
ABSTRACT: Leadership is evolving dynamically from an individual endeavor to shared
efforts. This paper aims to advance our understanding of shared leadership in
scientific teams. We define three kinds of leaders, junior (10-15), mid
(15-20), and senior (20+) based on career age. By considering the combinations
of any two leaders, we distinguish shared leadership as heterogeneous when
leaders are in different age cohorts and homogeneous when leaders are in the
same age cohort. Drawing on 1,845,351 CS, 254,039 Sociology, and 193,338
Business teams with two leaders in the OpenAlex dataset, we identify that
heterogeneous shared leadership brings higher citation impact for teams than
homogeneous shared leadership. Specifically, when junior leaders are paired
with senior leaders, it significantly increases team citation ranking by 1-2%,
in comparison with two leaders of similar age. We explore the patterns between
homogeneous leaders and heterogeneous leaders from team scale, expertise
composition, and knowledge recency perspectives. Compared with homogeneous
leaders, heterogeneous leaders are more adaptive in large teams, have more
diverse expertise, and trace both the newest and oldest references.
| no_new_dataset | 0.711036 |
2306.15826 | Tianyi Chen | Zhehua Zhong, Tianyi Chen, Zhen Wang | MAT: Mixed-Strategy Game of Adversarial Training in Fine-tuning | Will appear on IJCAI 2023 | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | Fine-tuning large-scale pre-trained language models has been demonstrated
effective for various natural language processing (NLP) tasks. Previous studies
have established that incorporating adversarial training during the fine-tuning
stage can significantly enhance model generalization and robustness. However,
from the perspective of game theory, such utilizations of adversarial training
correspond to pure-strategy games, which are inherently limited in terms of the
scope of their strategies, thereby still having room for improvement. In order
to push the performance boundaries, we propose a novel Mixed-strategy
Adversarial Training algorithm (MAT). Methodologically, we derive the Nash
equilibrium of a mixed-strategy game for adversarial training using Entropy
Mirror Descent to establish MAT by sampling method. To verify the effectiveness
of MAT, we conducted extensive benchmark experiments on large-scale pre-trained
models, such as BERT and RoBERTa. MAT significantly outperforms the
state-of-the-art methods on both the GLUE and ANLI benchmarks in terms of
generalization and robustness.
| [
{
"version": "v1",
"created": "Tue, 27 Jun 2023 23:19:53 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Zhong",
"Zhehua",
""
],
[
"Chen",
"Tianyi",
""
],
[
"Wang",
"Zhen",
""
]
] | TITLE: MAT: Mixed-Strategy Game of Adversarial Training in Fine-tuning
ABSTRACT: Fine-tuning large-scale pre-trained language models has been demonstrated
effective for various natural language processing (NLP) tasks. Previous studies
have established that incorporating adversarial training during the fine-tuning
stage can significantly enhance model generalization and robustness. However,
from the perspective of game theory, such utilizations of adversarial training
correspond to pure-strategy games, which are inherently limited in terms of the
scope of their strategies, thereby still having room for improvement. In order
to push the performance boundaries, we propose a novel Mixed-strategy
Adversarial Training algorithm (MAT). Methodologically, we derive the Nash
equilibrium of a mixed-strategy game for adversarial training using Entropy
Mirror Descent to establish MAT by sampling method. To verify the effectiveness
of MAT, we conducted extensive benchmark experiments on large-scale pre-trained
models, such as BERT and RoBERTa. MAT significantly outperforms the
state-of-the-art methods on both the GLUE and ANLI benchmarks in terms of
generalization and robustness.
| no_new_dataset | 0.710823 |
2306.15850 | Santhosh Kumar Ramakrishnan | Santhosh Kumar Ramakrishnan, Ziad Al-Halah, Kristen Grauman | SpotEM: Efficient Video Search for Episodic Memory | Published in ICML 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The goal in episodic memory (EM) is to search a long egocentric video to
answer a natural language query (e.g., "where did I leave my purse?"). Existing
EM methods exhaustively extract expensive fixed-length clip features to look
everywhere in the video for the answer, which is infeasible for long
wearable-camera videos that span hours or even days. We propose SpotEM, an
approach to achieve efficiency for a given EM method while maintaining good
accuracy. SpotEM consists of three key ideas: 1) a novel clip selector that
learns to identify promising video regions to search conditioned on the
language query; 2) a set of low-cost semantic indexing features that capture
the context of rooms, objects, and interactions that suggest where to look; and
3) distillation losses that address the optimization issues arising from
end-to-end joint training of the clip selector and EM model. Our experiments on
200+ hours of video from the Ego4D EM Natural Language Queries benchmark and
three different EM models demonstrate the effectiveness of our approach:
computing only 10% - 25% of the clip features, we preserve 84% - 97% of the
original EM model's accuracy. Project page:
https://vision.cs.utexas.edu/projects/spotem
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 00:52:49 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Ramakrishnan",
"Santhosh Kumar",
""
],
[
"Al-Halah",
"Ziad",
""
],
[
"Grauman",
"Kristen",
""
]
] | TITLE: SpotEM: Efficient Video Search for Episodic Memory
ABSTRACT: The goal in episodic memory (EM) is to search a long egocentric video to
answer a natural language query (e.g., "where did I leave my purse?"). Existing
EM methods exhaustively extract expensive fixed-length clip features to look
everywhere in the video for the answer, which is infeasible for long
wearable-camera videos that span hours or even days. We propose SpotEM, an
approach to achieve efficiency for a given EM method while maintaining good
accuracy. SpotEM consists of three key ideas: 1) a novel clip selector that
learns to identify promising video regions to search conditioned on the
language query; 2) a set of low-cost semantic indexing features that capture
the context of rooms, objects, and interactions that suggest where to look; and
3) distillation losses that address the optimization issues arising from
end-to-end joint training of the clip selector and EM model. Our experiments on
200+ hours of video from the Ego4D EM Natural Language Queries benchmark and
three different EM models demonstrate the effectiveness of our approach:
computing only 10% - 25% of the clip features, we preserve 84% - 97% of the
original EM model's accuracy. Project page:
https://vision.cs.utexas.edu/projects/spotem
| no_new_dataset | 0.709982 |
2306.15852 | Meenakshi Sarkar | Meenakshi Sarkar, Vinayak Honkote, Dibyendu Das and Debasish Ghose | Action-conditioned Deep Visual Prediction with RoAM, a new Indoor Human
Motion Dataset for Autonomous Robots | null | null | null | null | cs.RO cs.CV | http://creativecommons.org/licenses/by/4.0/ | With the increasing adoption of robots across industries, it is crucial to
focus on developing advanced algorithms that enable robots to anticipate,
comprehend, and plan their actions effectively in collaboration with humans. We
introduce the Robot Autonomous Motion (RoAM) video dataset, which is collected
with a custom-made turtlebot3 Burger robot in a variety of indoor environments
recording various human motions from the robot's ego-vision. The dataset also
includes synchronized records of the LiDAR scan and all control actions taken
by the robot as it navigates around static and moving human agents. The unique
dataset provides an opportunity to develop and benchmark new visual prediction
frameworks that can predict future image frames based on the action taken by
the recording agent in partially observable scenarios or cases where the
imaging sensor is mounted on a moving platform. We have benchmarked the dataset
on our novel deep visual prediction framework called ACPNet where the
approximated future image frames are also conditioned on action taken by the
robot and demonstrated its potential for incorporating robot dynamics into the
video prediction paradigm for mobile robotics and autonomous navigation
research.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 00:58:44 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Sarkar",
"Meenakshi",
""
],
[
"Honkote",
"Vinayak",
""
],
[
"Das",
"Dibyendu",
""
],
[
"Ghose",
"Debasish",
""
]
] | TITLE: Action-conditioned Deep Visual Prediction with RoAM, a new Indoor Human
Motion Dataset for Autonomous Robots
ABSTRACT: With the increasing adoption of robots across industries, it is crucial to
focus on developing advanced algorithms that enable robots to anticipate,
comprehend, and plan their actions effectively in collaboration with humans. We
introduce the Robot Autonomous Motion (RoAM) video dataset, which is collected
with a custom-made turtlebot3 Burger robot in a variety of indoor environments
recording various human motions from the robot's ego-vision. The dataset also
includes synchronized records of the LiDAR scan and all control actions taken
by the robot as it navigates around static and moving human agents. The unique
dataset provides an opportunity to develop and benchmark new visual prediction
frameworks that can predict future image frames based on the action taken by
the recording agent in partially observable scenarios or cases where the
imaging sensor is mounted on a moving platform. We have benchmarked the dataset
on our novel deep visual prediction framework called ACPNet where the
approximated future image frames are also conditioned on action taken by the
robot and demonstrated its potential for incorporating robot dynamics into the
video prediction paradigm for mobile robotics and autonomous navigation
research.
| new_dataset | 0.717179 |
2306.15853 | Marjan Shahi | Marjan Shahi, David Clausi, Alexander Wong | GoalieNet: A Multi-Stage Network for Joint Goalie, Equipment, and Net
Pose Estimation in Ice Hockey | null | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In the field of computer vision-driven ice hockey analytics, one of the most
challenging and least studied tasks is goalie pose estimation. Unlike general
human pose estimation, goalie pose estimation is much more complex as it
involves not only the detection of keypoints corresponding to the joints of the
goalie concealed under thick padding and mask, but also a large number of
non-human keypoints corresponding to the large leg pads and gloves worn, the
stick, as well as the hockey net. To tackle this challenge, we introduce
GoalieNet, a multi-stage deep neural network for jointly estimating the pose of
the goalie, their equipment, and the net. Experimental results using NHL
benchmark data demonstrate that the proposed GoalieNet can achieve an average
of 84\% accuracy across all keypoints, where 22 out of 29 keypoints are
detected with more than 80\% accuracy. This indicates that such a joint pose
estimation approach can be a promising research direction.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 01:00:36 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Shahi",
"Marjan",
""
],
[
"Clausi",
"David",
""
],
[
"Wong",
"Alexander",
""
]
] | TITLE: GoalieNet: A Multi-Stage Network for Joint Goalie, Equipment, and Net
Pose Estimation in Ice Hockey
ABSTRACT: In the field of computer vision-driven ice hockey analytics, one of the most
challenging and least studied tasks is goalie pose estimation. Unlike general
human pose estimation, goalie pose estimation is much more complex as it
involves not only the detection of keypoints corresponding to the joints of the
goalie concealed under thick padding and mask, but also a large number of
non-human keypoints corresponding to the large leg pads and gloves worn, the
stick, as well as the hockey net. To tackle this challenge, we introduce
GoalieNet, a multi-stage deep neural network for jointly estimating the pose of
the goalie, their equipment, and the net. Experimental results using NHL
benchmark data demonstrate that the proposed GoalieNet can achieve an average
of 84\% accuracy across all keypoints, where 22 out of 29 keypoints are
detected with more than 80\% accuracy. This indicates that such a joint pose
estimation approach can be a promising research direction.
| no_new_dataset | 0.708458 |
2306.15870 | Zheyan Jin | Zheyan Jin, Shiqi Chen, Yueting Chen, Zhihai Xu, Huajun Feng | Let Segment Anything Help Image Dehaze | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | The large language model and high-level vision model have achieved impressive
performance improvements with large datasets and model sizes. However,
low-level computer vision tasks, such as image dehaze and blur removal, still
rely on a small number of datasets and small-sized models, which generally
leads to overfitting and local optima. Therefore, we propose a framework to
integrate large-model prior into low-level computer vision tasks. Just as with
the task of image segmentation, the degradation of haze is also
texture-related. So we propose to detect gray-scale coding, network channel
expansion, and pre-dehaze structures to integrate large-model prior knowledge
into any low-level dehazing network. We demonstrate the effectiveness and
applicability of large models in guiding low-level visual tasks through
different datasets and algorithms comparison experiments. Finally, we
demonstrate the effect of grayscale coding, network channel expansion, and
recurrent network structures through ablation experiments. Under the conditions
where additional data and training resources are not required, we successfully
prove that the integration of large-model prior knowledge will improve the
dehaze performance and save training time for low-level visual tasks.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 02:02:19 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Jin",
"Zheyan",
""
],
[
"Chen",
"Shiqi",
""
],
[
"Chen",
"Yueting",
""
],
[
"Xu",
"Zhihai",
""
],
[
"Feng",
"Huajun",
""
]
] | TITLE: Let Segment Anything Help Image Dehaze
ABSTRACT: The large language model and high-level vision model have achieved impressive
performance improvements with large datasets and model sizes. However,
low-level computer vision tasks, such as image dehaze and blur removal, still
rely on a small number of datasets and small-sized models, which generally
leads to overfitting and local optima. Therefore, we propose a framework to
integrate large-model prior into low-level computer vision tasks. Just as with
the task of image segmentation, the degradation of haze is also
texture-related. So we propose to detect gray-scale coding, network channel
expansion, and pre-dehaze structures to integrate large-model prior knowledge
into any low-level dehazing network. We demonstrate the effectiveness and
applicability of large models in guiding low-level visual tasks through
different datasets and algorithms comparison experiments. Finally, we
demonstrate the effect of grayscale coding, network channel expansion, and
recurrent network structures through ablation experiments. Under the conditions
where additional data and training resources are not required, we successfully
prove that the integration of large-model prior knowledge will improve the
dehaze performance and save training time for low-level visual tasks.
| no_new_dataset | 0.709787 |
2306.15876 | Bowen Shi | Bowen Shi, Xiaopeng Zhang, Yaoming Wang, Jin Li, Wenrui Dai, Junni
Zou, Hongkai Xiong, Qi Tian | Hybrid Distillation: Connecting Masked Autoencoders with Contrastive
Learners | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Representation learning has been evolving from traditional supervised
training to Contrastive Learning (CL) and Masked Image Modeling (MIM). Previous
works have demonstrated their pros and cons in specific scenarios, i.e., CL and
supervised pre-training excel at capturing longer-range global patterns and
enabling better feature discrimination, while MIM can introduce more local and
diverse attention across all transformer layers. In this paper, we explore how
to obtain a model that combines their strengths. We start by examining previous
feature distillation and mask feature reconstruction methods and identify their
limitations. We find that their increasing diversity mainly derives from the
asymmetric designs, but these designs may in turn compromise the discrimination
ability. In order to better obtain both discrimination and diversity, we
propose a simple but effective Hybrid Distillation strategy, which utilizes
both the supervised/CL teacher and the MIM teacher to jointly guide the student
model. Hybrid Distill imitates the token relations of the MIM teacher to
alleviate attention collapse, as well as distills the feature maps of the
supervised/CL teacher to enable discrimination. Furthermore, a progressive
redundant token masking strategy is also utilized to reduce the distilling
costs and avoid falling into local optima. Experiment results prove that Hybrid
Distill can achieve superior performance on different benchmarks.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 02:19:35 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Shi",
"Bowen",
""
],
[
"Zhang",
"Xiaopeng",
""
],
[
"Wang",
"Yaoming",
""
],
[
"Li",
"Jin",
""
],
[
"Dai",
"Wenrui",
""
],
[
"Zou",
"Junni",
""
],
[
"Xiong",
"Hongkai",
""
],
[
"Tian",
"Qi",
""
]
] | TITLE: Hybrid Distillation: Connecting Masked Autoencoders with Contrastive
Learners
ABSTRACT: Representation learning has been evolving from traditional supervised
training to Contrastive Learning (CL) and Masked Image Modeling (MIM). Previous
works have demonstrated their pros and cons in specific scenarios, i.e., CL and
supervised pre-training excel at capturing longer-range global patterns and
enabling better feature discrimination, while MIM can introduce more local and
diverse attention across all transformer layers. In this paper, we explore how
to obtain a model that combines their strengths. We start by examining previous
feature distillation and mask feature reconstruction methods and identify their
limitations. We find that their increasing diversity mainly derives from the
asymmetric designs, but these designs may in turn compromise the discrimination
ability. In order to better obtain both discrimination and diversity, we
propose a simple but effective Hybrid Distillation strategy, which utilizes
both the supervised/CL teacher and the MIM teacher to jointly guide the student
model. Hybrid Distill imitates the token relations of the MIM teacher to
alleviate attention collapse, as well as distills the feature maps of the
supervised/CL teacher to enable discrimination. Furthermore, a progressive
redundant token masking strategy is also utilized to reduce the distilling
costs and avoid falling into local optima. Experiment results prove that Hybrid
Distill can achieve superior performance on different benchmarks.
| no_new_dataset | 0.709372 |
2306.15884 | Zheyan Jin | Zheyan Jin, Shiqi Chen, Huajun Feng, Zhihai Xu, Yueting Chen | Toward Real Flare Removal: A Comprehensive Pipeline and A New Benchmark | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Photographing in the under-illuminated scenes, the presence of complex light
sources often leave strong flare artifacts in images, where the intensity, the
spectrum, the reflection, and the aberration altogether contribute the
deterioration. Besides the image quality, it also influence the performance of
down-stream visual applications. Thus, removing the lens flare and ghosts is a
challenge issue especially in low-light environment. However, existing methods
for flare removal mainly restricted to the problems of inadequate simulation
and real-world capture, where the categories of scattered flares are singular
and the reflected ghosts are unavailable. Therefore, a comprehensive
deterioration procedure is crucial for constructing the dataset of flare
removal. Based on the theoretical analysis and real-world evaluation, we
propose a well-developed methodology for generating the data-pairs with flare
deterioration. The procedure is comprehensive, where the similarity of
scattered flares and the symmetric effect of reflected ghosts are realized.
Moreover, we also construct a real-shot pipeline that respectively processes
the effects of scattering and reflective flares, aiming to directly generate
the data for end-to-end methods. Experimental results show that the proposed
methodology add diversity to the existing flare datasets and construct a
comprehensive mapping procedure for flare data pairs. And our method facilities
the data-driven model to realize better restoration in flare images and
proposes a better evaluation system based on real shots, resulting promote
progress in the area of real flare removal.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 02:57:25 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Jin",
"Zheyan",
""
],
[
"Chen",
"Shiqi",
""
],
[
"Feng",
"Huajun",
""
],
[
"Xu",
"Zhihai",
""
],
[
"Chen",
"Yueting",
""
]
] | TITLE: Toward Real Flare Removal: A Comprehensive Pipeline and A New Benchmark
ABSTRACT: Photographing in the under-illuminated scenes, the presence of complex light
sources often leave strong flare artifacts in images, where the intensity, the
spectrum, the reflection, and the aberration altogether contribute the
deterioration. Besides the image quality, it also influence the performance of
down-stream visual applications. Thus, removing the lens flare and ghosts is a
challenge issue especially in low-light environment. However, existing methods
for flare removal mainly restricted to the problems of inadequate simulation
and real-world capture, where the categories of scattered flares are singular
and the reflected ghosts are unavailable. Therefore, a comprehensive
deterioration procedure is crucial for constructing the dataset of flare
removal. Based on the theoretical analysis and real-world evaluation, we
propose a well-developed methodology for generating the data-pairs with flare
deterioration. The procedure is comprehensive, where the similarity of
scattered flares and the symmetric effect of reflected ghosts are realized.
Moreover, we also construct a real-shot pipeline that respectively processes
the effects of scattering and reflective flares, aiming to directly generate
the data for end-to-end methods. Experimental results show that the proposed
methodology add diversity to the existing flare datasets and construct a
comprehensive mapping procedure for flare data pairs. And our method facilities
the data-driven model to realize better restoration in flare images and
proposes a better evaluation system based on real shots, resulting promote
progress in the area of real flare removal.
| no_new_dataset | 0.707821 |
2306.15887 | Mohammad Mofrad | Salmonn Talebi, Elizabeth Tong and Mohammad R. K. Mofrad | Beyond the Hype: Assessing the Performance, Trustworthiness, and
Clinical Suitability of GPT3.5 | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by/4.0/ | The use of large language models (LLMs) in healthcare is gaining popularity,
but their practicality and safety in clinical settings have not been thoroughly
assessed. In high-stakes environments like medical settings, trust and safety
are critical issues for LLMs. To address these concerns, we present an approach
to evaluate the performance and trustworthiness of a GPT3.5 model for medical
image protocol assignment. We compare it with a fine-tuned BERT model and a
radiologist. In addition, we have a radiologist review the GPT3.5 output to
evaluate its decision-making process. Our evaluation dataset consists of 4,700
physician entries across 11 imaging protocol classes spanning the entire head.
Our findings suggest that the GPT3.5 performance falls behind BERT and a
radiologist. However, GPT3.5 outperforms BERT in its ability to explain its
decision, detect relevant word indicators, and model calibration. Furthermore,
by analyzing the explanations of GPT3.5 for misclassifications, we reveal
systematic errors that need to be resolved to enhance its safety and
suitability for clinical use.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 03:03:51 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Talebi",
"Salmonn",
""
],
[
"Tong",
"Elizabeth",
""
],
[
"Mofrad",
"Mohammad R. K.",
""
]
] | TITLE: Beyond the Hype: Assessing the Performance, Trustworthiness, and
Clinical Suitability of GPT3.5
ABSTRACT: The use of large language models (LLMs) in healthcare is gaining popularity,
but their practicality and safety in clinical settings have not been thoroughly
assessed. In high-stakes environments like medical settings, trust and safety
are critical issues for LLMs. To address these concerns, we present an approach
to evaluate the performance and trustworthiness of a GPT3.5 model for medical
image protocol assignment. We compare it with a fine-tuned BERT model and a
radiologist. In addition, we have a radiologist review the GPT3.5 output to
evaluate its decision-making process. Our evaluation dataset consists of 4,700
physician entries across 11 imaging protocol classes spanning the entire head.
Our findings suggest that the GPT3.5 performance falls behind BERT and a
radiologist. However, GPT3.5 outperforms BERT in its ability to explain its
decision, detect relevant word indicators, and model calibration. Furthermore,
by analyzing the explanations of GPT3.5 for misclassifications, we reveal
systematic errors that need to be resolved to enhance its safety and
suitability for clinical use.
| new_dataset | 0.71464 |
2306.15910 | Huu-Thanh Nguyen | Huu-Thanh Nguyen, Yu Cao, Chong-Wah Ngo, Wing-Kwong Chan | Incremental Learning on Food Instance Segmentation | null | null | null | null | cs.CV cs.MM | http://creativecommons.org/licenses/by/4.0/ | Food instance segmentation is essential to estimate the serving size of
dishes in a food image. The recent cutting-edge techniques for instance
segmentation are deep learning networks with impressive segmentation quality
and fast computation. Nonetheless, they are hungry for data and expensive for
annotation. This paper proposes an incremental learning framework to optimize
the model performance given a limited data labelling budget. The power of the
framework is a novel difficulty assessment model, which forecasts how
challenging an unlabelled sample is to the latest trained instance segmentation
model. The data collection procedure is divided into several stages, each in
which a new sample package is collected. The framework allocates the labelling
budget to the most difficult samples. The unlabelled samples that meet a
certain qualification from the assessment model are used to generate
pseudo-labels. Eventually, the manual labels and pseudo-labels are sent to the
training data to improve the instance segmentation model. On four large-scale
food datasets, our proposed framework outperforms current incremental learning
benchmarks and achieves competitive performance with the model trained on fully
annotated samples.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 04:17:16 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Nguyen",
"Huu-Thanh",
""
],
[
"Cao",
"Yu",
""
],
[
"Ngo",
"Chong-Wah",
""
],
[
"Chan",
"Wing-Kwong",
""
]
] | TITLE: Incremental Learning on Food Instance Segmentation
ABSTRACT: Food instance segmentation is essential to estimate the serving size of
dishes in a food image. The recent cutting-edge techniques for instance
segmentation are deep learning networks with impressive segmentation quality
and fast computation. Nonetheless, they are hungry for data and expensive for
annotation. This paper proposes an incremental learning framework to optimize
the model performance given a limited data labelling budget. The power of the
framework is a novel difficulty assessment model, which forecasts how
challenging an unlabelled sample is to the latest trained instance segmentation
model. The data collection procedure is divided into several stages, each in
which a new sample package is collected. The framework allocates the labelling
budget to the most difficult samples. The unlabelled samples that meet a
certain qualification from the assessment model are used to generate
pseudo-labels. Eventually, the manual labels and pseudo-labels are sent to the
training data to improve the instance segmentation model. On four large-scale
food datasets, our proposed framework outperforms current incremental learning
benchmarks and achieves competitive performance with the model trained on fully
annotated samples.
| no_new_dataset | 0.71 |
2306.15917 | Noah Bergam | William Yang, Noah Bergam, Arnav Jain, Nima Sheikhoslami | Confidence-Calibrated Ensemble Dense Phrase Retrieval | null | null | null | null | cs.CL cs.IR | http://creativecommons.org/licenses/by/4.0/ | In this paper, we consider the extent to which the transformer-based Dense
Passage Retrieval (DPR) algorithm, developed by (Karpukhin et. al. 2020), can
be optimized without further pre-training. Our method involves two particular
insights: we apply the DPR context encoder at various phrase lengths (e.g.
one-sentence versus five-sentence segments), and we take a
confidence-calibrated ensemble prediction over all of these different
segmentations. This somewhat exhaustive approach achieves start-of-the-art
results on benchmark datasets such as Google NQ and SQuAD. We also apply our
method to domain-specific datasets, and the results suggest how different
granularities are optimal for different domains
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 04:40:21 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Yang",
"William",
""
],
[
"Bergam",
"Noah",
""
],
[
"Jain",
"Arnav",
""
],
[
"Sheikhoslami",
"Nima",
""
]
] | TITLE: Confidence-Calibrated Ensemble Dense Phrase Retrieval
ABSTRACT: In this paper, we consider the extent to which the transformer-based Dense
Passage Retrieval (DPR) algorithm, developed by (Karpukhin et. al. 2020), can
be optimized without further pre-training. Our method involves two particular
insights: we apply the DPR context encoder at various phrase lengths (e.g.
one-sentence versus five-sentence segments), and we take a
confidence-calibrated ensemble prediction over all of these different
segmentations. This somewhat exhaustive approach achieves start-of-the-art
results on benchmark datasets such as Google NQ and SQuAD. We also apply our
method to domain-specific datasets, and the results suggest how different
granularities are optimal for different domains
| no_new_dataset | 0.711875 |
2306.15926 | Allen Roush | Allen Roush, Sanjay Basu, Akshay Moorthy, Dmitry Dubovoy | Most Language Models can be Poets too: An AI Writing Assistant and
Constrained Text Generation Studio | Published in the proceedings of the 2nd Workshop on When Creative AI
Meets Conversational AI (CAI2), COLING 2022, 6 pages, System Demonstration
Paper | null | null | null | cs.CL cs.AI cs.LG | http://creativecommons.org/licenses/by/4.0/ | Despite rapid advancement in the field of Constrained Natural Language
Generation, little time has been spent on exploring the potential of language
models which have had their vocabularies lexically, semantically, and/or
phonetically constrained. We find that most language models generate compelling
text even under significant constraints. We present a simple and universally
applicable technique for modifying the output of a language model by
compositionally applying filter functions to the language models vocabulary
before a unit of text is generated. This approach is plug-and-play and requires
no modification to the model. To showcase the value of this technique, we
present an easy to use AI writing assistant called Constrained Text Generation
Studio (CTGS). CTGS allows users to generate or choose from text with any
combination of a wide variety of constraints, such as banning a particular
letter, forcing the generated words to have a certain number of syllables,
and/or forcing the words to be partial anagrams of another word. We introduce a
novel dataset of prose that omits the letter e. We show that our method results
in strictly superior performance compared to fine-tuning alone on this dataset.
We also present a Huggingface space web-app presenting this technique called
Gadsby. The code is available to the public here:
https://github.com/Hellisotherpeople/Constrained-Text-Generation-Studio
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 05:10:51 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Roush",
"Allen",
""
],
[
"Basu",
"Sanjay",
""
],
[
"Moorthy",
"Akshay",
""
],
[
"Dubovoy",
"Dmitry",
""
]
] | TITLE: Most Language Models can be Poets too: An AI Writing Assistant and
Constrained Text Generation Studio
ABSTRACT: Despite rapid advancement in the field of Constrained Natural Language
Generation, little time has been spent on exploring the potential of language
models which have had their vocabularies lexically, semantically, and/or
phonetically constrained. We find that most language models generate compelling
text even under significant constraints. We present a simple and universally
applicable technique for modifying the output of a language model by
compositionally applying filter functions to the language models vocabulary
before a unit of text is generated. This approach is plug-and-play and requires
no modification to the model. To showcase the value of this technique, we
present an easy to use AI writing assistant called Constrained Text Generation
Studio (CTGS). CTGS allows users to generate or choose from text with any
combination of a wide variety of constraints, such as banning a particular
letter, forcing the generated words to have a certain number of syllables,
and/or forcing the words to be partial anagrams of another word. We introduce a
novel dataset of prose that omits the letter e. We show that our method results
in strictly superior performance compared to fine-tuning alone on this dataset.
We also present a Huggingface space web-app presenting this technique called
Gadsby. The code is available to the public here:
https://github.com/Hellisotherpeople/Constrained-Text-Generation-Studio
| new_dataset | 0.714205 |
2306.15934 | Isaac Kauvar | Isaac Kauvar, Chris Doyle, Linqi Zhou, Nick Haber | Curious Replay for Model-based Adaptation | Accepted at ICML 2023. Website at
https://sites.google.com/view/curious-replay | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Agents must be able to adapt quickly as an environment changes. We find that
existing model-based reinforcement learning agents are unable to do this well,
in part because of how they use past experiences to train their world model.
Here, we present Curious Replay -- a form of prioritized experience replay
tailored to model-based agents through use of a curiosity-based priority
signal. Agents using Curious Replay exhibit improved performance in an
exploration paradigm inspired by animal behavior and on the Crafter benchmark.
DreamerV3 with Curious Replay surpasses state-of-the-art performance on
Crafter, achieving a mean score of 19.4 that substantially improves on the
previous high score of 14.5 by DreamerV3 with uniform replay, while also
maintaining similar performance on the Deepmind Control Suite. Code for Curious
Replay is available at https://github.com/AutonomousAgentsLab/curiousreplay
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 05:34:53 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Kauvar",
"Isaac",
""
],
[
"Doyle",
"Chris",
""
],
[
"Zhou",
"Linqi",
""
],
[
"Haber",
"Nick",
""
]
] | TITLE: Curious Replay for Model-based Adaptation
ABSTRACT: Agents must be able to adapt quickly as an environment changes. We find that
existing model-based reinforcement learning agents are unable to do this well,
in part because of how they use past experiences to train their world model.
Here, we present Curious Replay -- a form of prioritized experience replay
tailored to model-based agents through use of a curiosity-based priority
signal. Agents using Curious Replay exhibit improved performance in an
exploration paradigm inspired by animal behavior and on the Crafter benchmark.
DreamerV3 with Curious Replay surpasses state-of-the-art performance on
Crafter, achieving a mean score of 19.4 that substantially improves on the
previous high score of 14.5 by DreamerV3 with uniform replay, while also
maintaining similar performance on the Deepmind Control Suite. Code for Curious
Replay is available at https://github.com/AutonomousAgentsLab/curiousreplay
| no_new_dataset | 0.708219 |
2306.15938 | Markus Lange-Hegermann | Amandeep Singh, Michael Weber, Markus Lange-Hegermann | Interpretable Anomaly Detection in Cellular Networks by Learning
Concepts in Variational Autoencoders | null | null | null | null | cs.LG cs.AI cs.NI stat.AP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper addresses the challenges of detecting anomalies in cellular
networks in an interpretable way and proposes a new approach using variational
autoencoders (VAEs) that learn interpretable representations of the latent
space for each Key Performance Indicator (KPI) in the dataset. This enables the
detection of anomalies based on reconstruction loss and Z-scores. We ensure the
interpretability of the anomalies via additional information centroids (c)
using the K-means algorithm to enhance representation learning. We evaluate the
performance of the model by analyzing patterns in the latent dimension for
specific KPIs and thereby demonstrate the interpretability and anomalies. The
proposed framework offers a faster and autonomous solution for detecting
anomalies in cellular networks and showcases the potential of deep
learning-based algorithms in handling big data.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 05:50:17 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Singh",
"Amandeep",
""
],
[
"Weber",
"Michael",
""
],
[
"Lange-Hegermann",
"Markus",
""
]
] | TITLE: Interpretable Anomaly Detection in Cellular Networks by Learning
Concepts in Variational Autoencoders
ABSTRACT: This paper addresses the challenges of detecting anomalies in cellular
networks in an interpretable way and proposes a new approach using variational
autoencoders (VAEs) that learn interpretable representations of the latent
space for each Key Performance Indicator (KPI) in the dataset. This enables the
detection of anomalies based on reconstruction loss and Z-scores. We ensure the
interpretability of the anomalies via additional information centroids (c)
using the K-means algorithm to enhance representation learning. We evaluate the
performance of the model by analyzing patterns in the latent dimension for
specific KPIs and thereby demonstrate the interpretability and anomalies. The
proposed framework offers a faster and autonomous solution for detecting
anomalies in cellular networks and showcases the potential of deep
learning-based algorithms in handling big data.
| no_new_dataset | 0.709221 |
2306.15961 | Yupu Guo | Yupu Guo, Fei Cai, Xin Zhanga, Jianming Zhenga, Honghui Chena | Disentangled Variational Auto-encoder Enhanced by Counterfactual Data
for Debiasing Recommendation | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recommender system always suffers from various recommendation biases,
seriously hindering its development. In this light, a series of debias methods
have been proposed in the recommender system, especially for two most common
biases, i.e., popularity bias and amplified subjective bias. However, exsisting
debias methods usually concentrate on correcting a single bias. Such
single-functionality debiases neglect the bias-coupling issue in which the
recommended items are collectively attributed to multiple biases. Besides,
previous work cannot tackle the lacking supervised signals brought by sparse
data, yet which has become a commonplace in the recommender system. In this
work, we introduce a disentangled debias variational auto-encoder
framework(DB-VAE) to address the single-functionality issue as well as a
counterfactual data enhancement method to mitigate the adverse effect due to
the data sparsity. In specific, DB-VAE first extracts two types of extreme
items only affected by a single bias based on the collier theory, which are
respectively employed to learn the latent representation of corresponding
biases, thereby realizing the bias decoupling. In this way, the exact unbiased
user representation can be learned by these decoupled bias representations.
Furthermore, the data generation module employs Pearl's framework to produce
massive counterfactual data, making up the lacking supervised signals due to
the sparse data. Extensive experiments on three real-world datasets demonstrate
the effectiveness of our proposed model. Besides, the counterfactual data can
further improve DB-VAE, especially on the dataset with low sparsity.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 06:50:11 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Guo",
"Yupu",
""
],
[
"Cai",
"Fei",
""
],
[
"Zhanga",
"Xin",
""
],
[
"Zhenga",
"Jianming",
""
],
[
"Chena",
"Honghui",
""
]
] | TITLE: Disentangled Variational Auto-encoder Enhanced by Counterfactual Data
for Debiasing Recommendation
ABSTRACT: Recommender system always suffers from various recommendation biases,
seriously hindering its development. In this light, a series of debias methods
have been proposed in the recommender system, especially for two most common
biases, i.e., popularity bias and amplified subjective bias. However, exsisting
debias methods usually concentrate on correcting a single bias. Such
single-functionality debiases neglect the bias-coupling issue in which the
recommended items are collectively attributed to multiple biases. Besides,
previous work cannot tackle the lacking supervised signals brought by sparse
data, yet which has become a commonplace in the recommender system. In this
work, we introduce a disentangled debias variational auto-encoder
framework(DB-VAE) to address the single-functionality issue as well as a
counterfactual data enhancement method to mitigate the adverse effect due to
the data sparsity. In specific, DB-VAE first extracts two types of extreme
items only affected by a single bias based on the collier theory, which are
respectively employed to learn the latent representation of corresponding
biases, thereby realizing the bias decoupling. In this way, the exact unbiased
user representation can be learned by these decoupled bias representations.
Furthermore, the data generation module employs Pearl's framework to produce
massive counterfactual data, making up the lacking supervised signals due to
the sparse data. Extensive experiments on three real-world datasets demonstrate
the effectiveness of our proposed model. Besides, the counterfactual data can
further improve DB-VAE, especially on the dataset with low sparsity.
| no_new_dataset | 0.708988 |
2306.15977 | Lingyu Si | Lingyu Si, Hongwei Dong, Wenwen Qiang, Junzhi Yu, Wenlong Zhai,
Changwen Zheng, Fanjiang Xu, Fuchun Sun | A Dimensional Structure based Knowledge Distillation Method for
Cross-Modal Learning | null | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to limitations in data quality, some essential visual tasks are difficult
to perform independently. Introducing previously unavailable information to
transfer informative dark knowledge has been a common way to solve such hard
tasks. However, research on why transferred knowledge works has not been
extensively explored. To address this issue, in this paper, we discover the
correlation between feature discriminability and dimensional structure (DS) by
analyzing and observing features extracted from simple and hard tasks. On this
basis, we express DS using deep channel-wise correlation and intermediate
spatial distribution, and propose a novel cross-modal knowledge distillation
(CMKD) method for better supervised cross-modal learning (CML) performance. The
proposed method enforces output features to be channel-wise independent and
intermediate ones to be uniformly distributed, thereby learning semantically
irrelevant features from the hard task to boost its accuracy. This is
especially useful in specific applications where the performance gap between
dual modalities is relatively large. Furthermore, we collect a real-world CML
dataset to promote community development. The dataset contains more than 10,000
paired optical and radar images and is continuously being updated. Experimental
results on real-world and benchmark datasets validate the effectiveness of the
proposed method.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 07:29:26 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Si",
"Lingyu",
""
],
[
"Dong",
"Hongwei",
""
],
[
"Qiang",
"Wenwen",
""
],
[
"Yu",
"Junzhi",
""
],
[
"Zhai",
"Wenlong",
""
],
[
"Zheng",
"Changwen",
""
],
[
"Xu",
"Fanjiang",
""
],
[
"Sun",
"Fuchun",
""
]
] | TITLE: A Dimensional Structure based Knowledge Distillation Method for
Cross-Modal Learning
ABSTRACT: Due to limitations in data quality, some essential visual tasks are difficult
to perform independently. Introducing previously unavailable information to
transfer informative dark knowledge has been a common way to solve such hard
tasks. However, research on why transferred knowledge works has not been
extensively explored. To address this issue, in this paper, we discover the
correlation between feature discriminability and dimensional structure (DS) by
analyzing and observing features extracted from simple and hard tasks. On this
basis, we express DS using deep channel-wise correlation and intermediate
spatial distribution, and propose a novel cross-modal knowledge distillation
(CMKD) method for better supervised cross-modal learning (CML) performance. The
proposed method enforces output features to be channel-wise independent and
intermediate ones to be uniformly distributed, thereby learning semantically
irrelevant features from the hard task to boost its accuracy. This is
especially useful in specific applications where the performance gap between
dual modalities is relatively large. Furthermore, we collect a real-world CML
dataset to promote community development. The dataset contains more than 10,000
paired optical and radar images and is continuously being updated. Experimental
results on real-world and benchmark datasets validate the effectiveness of the
proposed method.
| new_dataset | 0.718838 |
2306.15978 | Chengguang Gan | Chengguang Gan, Qinghao Zhang and Tatsunori Mori | Sentence-to-Label Generation Framework for Multi-task Learning of
Japanese Sentence Classification and Named Entity Recognition | Accept in NLDB2023 as Long Paper | null | 10.1007/978-3-031-35320-8_18 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Information extraction(IE) is a crucial subfield within natural language
processing. In this study, we introduce a Sentence Classification and Named
Entity Recognition Multi-task (SCNM) approach that combines Sentence
Classification (SC) and Named Entity Recognition (NER). We develop a
Sentence-to-Label Generation (SLG) framework for SCNM and construct a Wikipedia
dataset containing both SC and NER. Using a format converter, we unify input
formats and employ a generative model to generate SC-labels, NER-labels, and
associated text segments. We propose a Constraint Mechanism (CM) to improve
generated format accuracy. Our results show SC accuracy increased by 1.13
points and NER by 1.06 points in SCNM compared to standalone tasks, with CM
raising format accuracy from 63.61 to 100. The findings indicate mutual
reinforcement effects between SC and NER, and integration enhances both tasks'
performance.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 07:29:44 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Gan",
"Chengguang",
""
],
[
"Zhang",
"Qinghao",
""
],
[
"Mori",
"Tatsunori",
""
]
] | TITLE: Sentence-to-Label Generation Framework for Multi-task Learning of
Japanese Sentence Classification and Named Entity Recognition
ABSTRACT: Information extraction(IE) is a crucial subfield within natural language
processing. In this study, we introduce a Sentence Classification and Named
Entity Recognition Multi-task (SCNM) approach that combines Sentence
Classification (SC) and Named Entity Recognition (NER). We develop a
Sentence-to-Label Generation (SLG) framework for SCNM and construct a Wikipedia
dataset containing both SC and NER. Using a format converter, we unify input
formats and employ a generative model to generate SC-labels, NER-labels, and
associated text segments. We propose a Constraint Mechanism (CM) to improve
generated format accuracy. Our results show SC accuracy increased by 1.13
points and NER by 1.06 points in SCNM compared to standalone tasks, with CM
raising format accuracy from 63.61 to 100. The findings indicate mutual
reinforcement effects between SC and NER, and integration enhances both tasks'
performance.
| new_dataset | 0.714435 |
2306.15990 | Mattia Giovanni Campana | Mattia Giovanni Campana, Franca Delmastro | MyDigitalFootprint: an extensive context dataset for pervasive computing
applications at the edge | null | Pervasive and Mobile Computing, Volume 70, January 2021, 101309 | 10.1016/j.pmcj.2020.101309 | null | cs.LG cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The widespread diffusion of connected smart devices has contributed to the
rapid expansion and evolution of the Internet at its edge. Personal mobile
devices interact with other smart objects in their surroundings, adapting
behavior based on rapidly changing user context. The ability of mobile devices
to process this data locally is crucial for quick adaptation. This can be
achieved through a single elaboration process integrated into user applications
or a middleware platform for context processing. However, the lack of public
datasets considering user context complexity in the mobile environment hinders
research progress. We introduce MyDigitalFootprint, a large-scale dataset
comprising smartphone sensor data, physical proximity information, and Online
Social Networks interactions. This dataset supports multimodal context
recognition and social relationship modeling. It spans two months of
measurements from 31 volunteer users in their natural environment, allowing for
unrestricted behavior. Existing public datasets focus on limited context data
for specific applications, while ours offers comprehensive information on the
user context in the mobile environment. To demonstrate the dataset's
effectiveness, we present three context-aware applications utilizing various
machine learning tasks: (i) a social link prediction algorithm based on
physical proximity data, (ii) daily-life activity recognition using
smartphone-embedded sensors data, and (iii) a pervasive context-aware
recommender system. Our dataset, with its heterogeneity of information, serves
as a valuable resource to validate new research in mobile and edge computing.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 07:59:47 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Campana",
"Mattia Giovanni",
""
],
[
"Delmastro",
"Franca",
""
]
] | TITLE: MyDigitalFootprint: an extensive context dataset for pervasive computing
applications at the edge
ABSTRACT: The widespread diffusion of connected smart devices has contributed to the
rapid expansion and evolution of the Internet at its edge. Personal mobile
devices interact with other smart objects in their surroundings, adapting
behavior based on rapidly changing user context. The ability of mobile devices
to process this data locally is crucial for quick adaptation. This can be
achieved through a single elaboration process integrated into user applications
or a middleware platform for context processing. However, the lack of public
datasets considering user context complexity in the mobile environment hinders
research progress. We introduce MyDigitalFootprint, a large-scale dataset
comprising smartphone sensor data, physical proximity information, and Online
Social Networks interactions. This dataset supports multimodal context
recognition and social relationship modeling. It spans two months of
measurements from 31 volunteer users in their natural environment, allowing for
unrestricted behavior. Existing public datasets focus on limited context data
for specific applications, while ours offers comprehensive information on the
user context in the mobile environment. To demonstrate the dataset's
effectiveness, we present three context-aware applications utilizing various
machine learning tasks: (i) a social link prediction algorithm based on
physical proximity data, (ii) daily-life activity recognition using
smartphone-embedded sensors data, and (iii) a pervasive context-aware
recommender system. Our dataset, with its heterogeneity of information, serves
as a valuable resource to validate new research in mobile and edge computing.
| new_dataset | 0.71815 |
2306.15992 | Shuli Lv | Shuli Lv, Yan Gao, Jiaxing Che, Quan Quan | Autonomous Drone Racing: Time-Optimal Spatial Iterative Learning Control
within a Virtual Tube | null | null | null | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is often necessary for drones to complete delivery, photography, and
rescue in the shortest time to increase efficiency. Many autonomous drone races
provide platforms to pursue algorithms to finish races as quickly as possible
for the above purpose. Unfortunately, existing methods often fail to keep
training and racing time short in drone racing competitions. This motivates us
to develop a high-efficient learning method by imitating the training
experience of top racing drivers. Unlike traditional iterative learning control
methods for accurate tracking, the proposed approach iteratively learns a
trajectory online to finish the race as quickly as possible. Simulations and
experiments using different models show that the proposed approach is
model-free and is able to achieve the optimal result with low computation
requirements. Furthermore, this approach surpasses some state-of-the-art
methods in racing time on a benchmark drone racing platform. An experiment on a
real quadcopter is also performed to demonstrate its effectiveness.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 08:03:59 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Lv",
"Shuli",
""
],
[
"Gao",
"Yan",
""
],
[
"Che",
"Jiaxing",
""
],
[
"Quan",
"Quan",
""
]
] | TITLE: Autonomous Drone Racing: Time-Optimal Spatial Iterative Learning Control
within a Virtual Tube
ABSTRACT: It is often necessary for drones to complete delivery, photography, and
rescue in the shortest time to increase efficiency. Many autonomous drone races
provide platforms to pursue algorithms to finish races as quickly as possible
for the above purpose. Unfortunately, existing methods often fail to keep
training and racing time short in drone racing competitions. This motivates us
to develop a high-efficient learning method by imitating the training
experience of top racing drivers. Unlike traditional iterative learning control
methods for accurate tracking, the proposed approach iteratively learns a
trajectory online to finish the race as quickly as possible. Simulations and
experiments using different models show that the proposed approach is
model-free and is able to achieve the optimal result with low computation
requirements. Furthermore, this approach surpasses some state-of-the-art
methods in racing time on a benchmark drone racing platform. An experiment on a
real quadcopter is also performed to demonstrate its effectiveness.
| no_new_dataset | 0.709233 |
2306.15994 | In\^es Oliveira E Silva | I. Oliveira e Silva, C. Soares, I. Sousa, R. Ghani | Systematic analysis of the impact of label noise correction on ML
Fairness | null | null | null | null | cs.LG cs.CY | http://creativecommons.org/licenses/by/4.0/ | Arbitrary, inconsistent, or faulty decision-making raises serious concerns,
and preventing unfair models is an increasingly important challenge in Machine
Learning. Data often reflect past discriminatory behavior, and models trained
on such data may reflect bias on sensitive attributes, such as gender, race, or
age. One approach to developing fair models is to preprocess the training data
to remove the underlying biases while preserving the relevant information, for
example, by correcting biased labels. While multiple label noise correction
methods are available, the information about their behavior in identifying
discrimination is very limited. In this work, we develop an empirical
methodology to systematically evaluate the effectiveness of label noise
correction techniques in ensuring the fairness of models trained on biased
datasets. Our methodology involves manipulating the amount of label noise and
can be used with fairness benchmarks but also with standard ML datasets. We
apply the methodology to analyze six label noise correction methods according
to several fairness metrics on standard OpenML datasets. Our results suggest
that the Hybrid Label Noise Correction method achieves the best trade-off
between predictive performance and fairness. Clustering-Based Correction can
reduce discrimination the most, however, at the cost of lower predictive
performance.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 08:08:14 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Silva",
"I. Oliveira e",
""
],
[
"Soares",
"C.",
""
],
[
"Sousa",
"I.",
""
],
[
"Ghani",
"R.",
""
]
] | TITLE: Systematic analysis of the impact of label noise correction on ML
Fairness
ABSTRACT: Arbitrary, inconsistent, or faulty decision-making raises serious concerns,
and preventing unfair models is an increasingly important challenge in Machine
Learning. Data often reflect past discriminatory behavior, and models trained
on such data may reflect bias on sensitive attributes, such as gender, race, or
age. One approach to developing fair models is to preprocess the training data
to remove the underlying biases while preserving the relevant information, for
example, by correcting biased labels. While multiple label noise correction
methods are available, the information about their behavior in identifying
discrimination is very limited. In this work, we develop an empirical
methodology to systematically evaluate the effectiveness of label noise
correction techniques in ensuring the fairness of models trained on biased
datasets. Our methodology involves manipulating the amount of label noise and
can be used with fairness benchmarks but also with standard ML datasets. We
apply the methodology to analyze six label noise correction methods according
to several fairness metrics on standard OpenML datasets. Our results suggest
that the Hybrid Label Noise Correction method achieves the best trade-off
between predictive performance and fairness. Clustering-Based Correction can
reduce discrimination the most, however, at the cost of lower predictive
performance.
| no_new_dataset | 0.709598 |
2306.16007 | Yu Wu | Yuang Li, Yu Wu, Jinyu Li, Shujie Liu | Prompting Large Language Models for Zero-Shot Domain Adaptation in
Speech Recognition | null | null | null | null | cs.CL eess.AS eess.SP | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The integration of Language Models (LMs) has proven to be an effective way to
address domain shifts in speech recognition. However, these approaches usually
require a significant amount of target domain text data for the training of
LMs. Different from these methods, in this work, with only a domain-specific
text prompt, we propose two zero-shot ASR domain adaptation methods using
LLaMA, a 7-billion-parameter large language model (LLM). LLM is used in two
ways: 1) second-pass rescoring: reranking N-best hypotheses of a given ASR
system with LLaMA; 2) deep LLM-fusion: incorporating LLM into the decoder of an
encoder-decoder based ASR system. Experiments show that, with only one domain
prompt, both methods can effectively reduce word error rates (WER) on
out-of-domain TedLium-2 and SPGISpeech datasets. Especially, the deep
LLM-fusion has the advantage of better recall of entity and out-of-vocabulary
words.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 08:29:00 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Li",
"Yuang",
""
],
[
"Wu",
"Yu",
""
],
[
"Li",
"Jinyu",
""
],
[
"Liu",
"Shujie",
""
]
] | TITLE: Prompting Large Language Models for Zero-Shot Domain Adaptation in
Speech Recognition
ABSTRACT: The integration of Language Models (LMs) has proven to be an effective way to
address domain shifts in speech recognition. However, these approaches usually
require a significant amount of target domain text data for the training of
LMs. Different from these methods, in this work, with only a domain-specific
text prompt, we propose two zero-shot ASR domain adaptation methods using
LLaMA, a 7-billion-parameter large language model (LLM). LLM is used in two
ways: 1) second-pass rescoring: reranking N-best hypotheses of a given ASR
system with LLaMA; 2) deep LLM-fusion: incorporating LLM into the decoder of an
encoder-decoder based ASR system. Experiments show that, with only one domain
prompt, both methods can effectively reduce word error rates (WER) on
out-of-domain TedLium-2 and SPGISpeech datasets. Especially, the deep
LLM-fusion has the advantage of better recall of entity and out-of-vocabulary
words.
| no_new_dataset | 0.710396 |
2306.16020 | Sebastian Krapf | Sebastian Krapf, Kevin Mayer, Martin Fischer | Points for Energy Renovation (PointER): A LiDAR-Derived Point Cloud
Dataset of One Million English Buildings Linked to Energy Characteristics | The PointER dataset can be downloaded from
https://doi.org/10.14459/2023mp1713501. The code used for generating building
point clouds is available at https://github.com/kdmayer/PointER | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Rapid renovation of Europe's inefficient buildings is required to reduce
climate change. However, analyzing and evaluating buildings at scale is
challenging because every building is unique. In current practice, the energy
performance of buildings is assessed during on-site visits, which are slow,
costly, and local. This paper presents a building point cloud dataset that
promotes a data-driven, large-scale understanding of the 3D representation of
buildings and their energy characteristics. We generate building point clouds
by intersecting building footprints with geo-referenced LiDAR data and link
them with attributes from UK's energy performance database via the Unique
Property Reference Number (UPRN). To achieve a representative sample, we select
one million buildings from a range of rural and urban regions across England,
of which half a million are linked to energy characteristics. Building point
clouds in new regions can be generated with the open-source code published
alongside the paper. The dataset enables novel research in building energy
modeling and can be easily expanded to other research fields by adding building
features via the UPRN or geo-location.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 08:48:22 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Krapf",
"Sebastian",
""
],
[
"Mayer",
"Kevin",
""
],
[
"Fischer",
"Martin",
""
]
] | TITLE: Points for Energy Renovation (PointER): A LiDAR-Derived Point Cloud
Dataset of One Million English Buildings Linked to Energy Characteristics
ABSTRACT: Rapid renovation of Europe's inefficient buildings is required to reduce
climate change. However, analyzing and evaluating buildings at scale is
challenging because every building is unique. In current practice, the energy
performance of buildings is assessed during on-site visits, which are slow,
costly, and local. This paper presents a building point cloud dataset that
promotes a data-driven, large-scale understanding of the 3D representation of
buildings and their energy characteristics. We generate building point clouds
by intersecting building footprints with geo-referenced LiDAR data and link
them with attributes from UK's energy performance database via the Unique
Property Reference Number (UPRN). To achieve a representative sample, we select
one million buildings from a range of rural and urban regions across England,
of which half a million are linked to energy characteristics. Building point
clouds in new regions can be generated with the open-source code published
alongside the paper. The dataset enables novel research in building energy
modeling and can be easily expanded to other research fields by adding building
features via the UPRN or geo-location.
| new_dataset | 0.715424 |
2306.16029 | Mattia Giovanni Campana | Mattia Giovanni Campana, Dimitris Chatzopoulos, Franca Delmastro, Pan
Hui | Lightweight Modeling of User Context Combining Physical and Virtual
Sensor Data | null | null | 10.1145/3267305.3274178 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The multitude of data generated by sensors available on users' mobile
devices, combined with advances in machine learning techniques, support
context-aware services in recognizing the current situation of a user (i.e.,
physical context) and optimizing the system's personalization features.
However, context-awareness performances mainly depend on the accuracy of the
context inference process, which is strictly tied to the availability of
large-scale and labeled datasets. In this work, we present a framework
developed to collect datasets containing heterogeneous sensing data derived
from personal mobile devices. The framework has been used by 3 voluntary users
for two weeks, generating a dataset with more than 36K samples and 1331
features. We also propose a lightweight approach to model the user context able
to efficiently perform the entire reasoning process on the user mobile device.
To this aim, we used six dimensionality reduction techniques in order to
optimize the context classification. Experimental results on the generated
dataset show that we achieve a 10x speed up and a feature reduction of more
than 90% while keeping the accuracy loss less than 3%.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 08:57:01 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Campana",
"Mattia Giovanni",
""
],
[
"Chatzopoulos",
"Dimitris",
""
],
[
"Delmastro",
"Franca",
""
],
[
"Hui",
"Pan",
""
]
] | TITLE: Lightweight Modeling of User Context Combining Physical and Virtual
Sensor Data
ABSTRACT: The multitude of data generated by sensors available on users' mobile
devices, combined with advances in machine learning techniques, support
context-aware services in recognizing the current situation of a user (i.e.,
physical context) and optimizing the system's personalization features.
However, context-awareness performances mainly depend on the accuracy of the
context inference process, which is strictly tied to the availability of
large-scale and labeled datasets. In this work, we present a framework
developed to collect datasets containing heterogeneous sensing data derived
from personal mobile devices. The framework has been used by 3 voluntary users
for two weeks, generating a dataset with more than 36K samples and 1331
features. We also propose a lightweight approach to model the user context able
to efficiently perform the entire reasoning process on the user mobile device.
To this aim, we used six dimensionality reduction techniques in order to
optimize the context classification. Experimental results on the generated
dataset show that we achieve a 10x speed up and a feature reduction of more
than 90% while keeping the accuracy loss less than 3%.
| new_dataset | 0.715405 |
2306.16036 | Fakai Wang | Fakai Wang, Chi-Tung Cheng, Chien-Wei Peng, Ke Yan, Min Wu, Le Lu,
Chien-Hung Liao, and Ling Zhang | A Cascaded Approach for ultraly High Performance Lesion Detection and
False Positive Removal in Liver CT Scans | null | null | null | null | eess.IV cs.CV | http://creativecommons.org/licenses/by/4.0/ | Liver cancer has high morbidity and mortality rates in the world. Multi-phase
CT is a main medical imaging modality for detecting/identifying and diagnosing
liver tumors. Automatically detecting and classifying liver lesions in CT
images have the potential to improve the clinical workflow. This task remains
challenging due to liver lesions' large variations in size, appearance, image
contrast, and the complexities of tumor types or subtypes. In this work, we
customize a multi-object labeling tool for multi-phase CT images, which is used
to curate a large-scale dataset containing 1,631 patients with four-phase CT
images, multi-organ masks, and multi-lesion (six major types of liver lesions
confirmed by pathology) masks. We develop a two-stage liver lesion detection
pipeline, where the high-sensitivity detecting algorithms in the first stage
discover as many lesion proposals as possible, and the lesion-reclassification
algorithms in the second stage remove as many false alarms as possible. The
multi-sensitivity lesion detection algorithm maximizes the information
utilization of the individual probability maps of segmentation, and the
lesion-shuffle augmentation effectively explores the texture contrast between
lesions and the liver. Independently tested on 331 patient cases, the proposed
model achieves high sensitivity and specificity for malignancy classification
in the multi-phase contrast-enhanced CT (99.2%, 97.1%, diagnosis setting) and
in the noncontrast CT (97.3%, 95.7%, screening setting).
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 09:11:34 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Wang",
"Fakai",
""
],
[
"Cheng",
"Chi-Tung",
""
],
[
"Peng",
"Chien-Wei",
""
],
[
"Yan",
"Ke",
""
],
[
"Wu",
"Min",
""
],
[
"Lu",
"Le",
""
],
[
"Liao",
"Chien-Hung",
""
],
[
"Zhang",
"Ling",
""
]
] | TITLE: A Cascaded Approach for ultraly High Performance Lesion Detection and
False Positive Removal in Liver CT Scans
ABSTRACT: Liver cancer has high morbidity and mortality rates in the world. Multi-phase
CT is a main medical imaging modality for detecting/identifying and diagnosing
liver tumors. Automatically detecting and classifying liver lesions in CT
images have the potential to improve the clinical workflow. This task remains
challenging due to liver lesions' large variations in size, appearance, image
contrast, and the complexities of tumor types or subtypes. In this work, we
customize a multi-object labeling tool for multi-phase CT images, which is used
to curate a large-scale dataset containing 1,631 patients with four-phase CT
images, multi-organ masks, and multi-lesion (six major types of liver lesions
confirmed by pathology) masks. We develop a two-stage liver lesion detection
pipeline, where the high-sensitivity detecting algorithms in the first stage
discover as many lesion proposals as possible, and the lesion-reclassification
algorithms in the second stage remove as many false alarms as possible. The
multi-sensitivity lesion detection algorithm maximizes the information
utilization of the individual probability maps of segmentation, and the
lesion-shuffle augmentation effectively explores the texture contrast between
lesions and the liver. Independently tested on 331 patient cases, the proposed
model achieves high sensitivity and specificity for malignancy classification
in the multi-phase contrast-enhanced CT (99.2%, 97.1%, diagnosis setting) and
in the noncontrast CT (97.3%, 95.7%, screening setting).
| new_dataset | 0.710446 |
2306.16049 | Mohammad Belal | James She, Kamilla Swart-Arries, Mohammad Belal and Simon Wong | What Sentiment and Fun Facts We Learnt Before FIFA World Cup Qatar 2022
Using Twitter and AI | null | null | null | null | cs.CL cs.SI | http://creativecommons.org/licenses/by/4.0/ | Twitter is a social media platform bridging most countries and allows
real-time news discovery. Since the tweets on Twitter are usually short and
express public feelings, thus provide a source for opinion mining and sentiment
analysis for global events. This paper proposed an effective solution, in
providing a sentiment on tweets related to the FIFA World Cup. At least 130k
tweets, as the first in the community, are collected and implemented as a
dataset to evaluate the performance of the proposed machine learning solution.
These tweets are collected with the related hashtags and keywords of the Qatar
World Cup 2022. The Vader algorithm is used in this paper for sentiment
analysis. Through the machine learning method and collected Twitter tweets, we
discovered the sentiments and fun facts of several aspects important to the
period before the World Cup. The result shows people are positive to the
opening of the World Cup.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 09:29:23 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"She",
"James",
""
],
[
"Swart-Arries",
"Kamilla",
""
],
[
"Belal",
"Mohammad",
""
],
[
"Wong",
"Simon",
""
]
] | TITLE: What Sentiment and Fun Facts We Learnt Before FIFA World Cup Qatar 2022
Using Twitter and AI
ABSTRACT: Twitter is a social media platform bridging most countries and allows
real-time news discovery. Since the tweets on Twitter are usually short and
express public feelings, thus provide a source for opinion mining and sentiment
analysis for global events. This paper proposed an effective solution, in
providing a sentiment on tweets related to the FIFA World Cup. At least 130k
tweets, as the first in the community, are collected and implemented as a
dataset to evaluate the performance of the proposed machine learning solution.
These tweets are collected with the related hashtags and keywords of the Qatar
World Cup 2022. The Vader algorithm is used in this paper for sentiment
analysis. Through the machine learning method and collected Twitter tweets, we
discovered the sentiments and fun facts of several aspects important to the
period before the World Cup. The result shows people are positive to the
opening of the World Cup.
| new_dataset | 0.711205 |
2306.16053 | Yi Wu | Ruiyuan Li, Zheng Li, Yi Wu, Chao Chen, Songtao Guo, Ming Zhang, Yu
Zheng | Erasing-based lossless compression method for streaming floating-point
time series | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | There are a prohibitively large number of floating-point time series data
generated at an unprecedentedly high rate. An efficient, compact and lossless
compression for time series data is of great importance for a wide range of
scenarios. Most existing lossless floating-point compression methods are based
on the XOR operation, but they do not fully exploit the trailing zeros, which
usually results in an unsatisfactory compression ratio. This paper proposes an
Erasing-based Lossless Floating-point compression algorithm, i.e., Elf. The
main idea of Elf is to erase the last few bits (i.e., set them to zero) of
floating-point values, so the XORed values are supposed to contain many
trailing zeros. The challenges of the erasing-based method are three-fold.
First, how to quickly determine the erased bits? Second, how to losslessly
recover the original data from the erased ones? Third, how to compactly encode
the erased data? Through rigorous mathematical analysis, Elf can directly
determine the erased bits and restore the original values without losing any
precision. To further improve the compression ratio, we propose a novel
encoding strategy for the XORed values with many trailing zeros. Furthermore,
observing the values in a time series usually have similar significand counts,
we propose an upgraded version of Elf named Elf+ by optimizing the significand
count encoding strategy, which improves the compression ratio and reduces the
running time further. Both Elf and Elf+ work in a streaming fashion. They take
only O(N) (where N is the length of a time series) in time and O(1) in space,
and achieve a notable compression ratio with a theoretical guarantee. Extensive
experiments using 22 datasets show the powerful performance of Elf and Elf+
compared with 9 advanced competitors for both double-precision and
single-precision floating-point values.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 09:33:31 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Li",
"Ruiyuan",
""
],
[
"Li",
"Zheng",
""
],
[
"Wu",
"Yi",
""
],
[
"Chen",
"Chao",
""
],
[
"Guo",
"Songtao",
""
],
[
"Zhang",
"Ming",
""
],
[
"Zheng",
"Yu",
""
]
] | TITLE: Erasing-based lossless compression method for streaming floating-point
time series
ABSTRACT: There are a prohibitively large number of floating-point time series data
generated at an unprecedentedly high rate. An efficient, compact and lossless
compression for time series data is of great importance for a wide range of
scenarios. Most existing lossless floating-point compression methods are based
on the XOR operation, but they do not fully exploit the trailing zeros, which
usually results in an unsatisfactory compression ratio. This paper proposes an
Erasing-based Lossless Floating-point compression algorithm, i.e., Elf. The
main idea of Elf is to erase the last few bits (i.e., set them to zero) of
floating-point values, so the XORed values are supposed to contain many
trailing zeros. The challenges of the erasing-based method are three-fold.
First, how to quickly determine the erased bits? Second, how to losslessly
recover the original data from the erased ones? Third, how to compactly encode
the erased data? Through rigorous mathematical analysis, Elf can directly
determine the erased bits and restore the original values without losing any
precision. To further improve the compression ratio, we propose a novel
encoding strategy for the XORed values with many trailing zeros. Furthermore,
observing the values in a time series usually have similar significand counts,
we propose an upgraded version of Elf named Elf+ by optimizing the significand
count encoding strategy, which improves the compression ratio and reduces the
running time further. Both Elf and Elf+ work in a streaming fashion. They take
only O(N) (where N is the length of a time series) in time and O(1) in space,
and achieve a notable compression ratio with a theoretical guarantee. Extensive
experiments using 22 datasets show the powerful performance of Elf and Elf+
compared with 9 advanced competitors for both double-precision and
single-precision floating-point values.
| no_new_dataset | 0.710879 |
2306.16054 | Michael K\"olle | Michael K\"olle, Steffen Illium, Maximilian Zorn, Jonas N\"u{\ss}lein,
Patrick Suchostawski and Claudia Linnhoff-Popien | Improving Primate Sounds Classification using Binary Presorting for Deep
Learning | DeLTA | null | null | null | cs.SD cs.CV cs.LG eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the field of wildlife observation and conservation, approaches involving
machine learning on audio recordings are becoming increasingly popular.
Unfortunately, available datasets from this field of research are often not
optimal learning material; Samples can be weakly labeled, of different lengths
or come with a poor signal-to-noise ratio. In this work, we introduce a
generalized approach that first relabels subsegments of MEL spectrogram
representations, to achieve higher performances on the actual multi-class
classification tasks. For both the binary pre-sorting and the classification,
we make use of convolutional neural networks (CNN) and various
data-augmentation techniques. We showcase the results of this approach on the
challenging \textit{ComparE 2021} dataset, with the task of classifying between
different primate species sounds, and report significantly higher Accuracy and
UAR scores in contrast to comparatively equipped model baselines.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 09:35:09 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Kölle",
"Michael",
""
],
[
"Illium",
"Steffen",
""
],
[
"Zorn",
"Maximilian",
""
],
[
"Nüßlein",
"Jonas",
""
],
[
"Suchostawski",
"Patrick",
""
],
[
"Linnhoff-Popien",
"Claudia",
""
]
] | TITLE: Improving Primate Sounds Classification using Binary Presorting for Deep
Learning
ABSTRACT: In the field of wildlife observation and conservation, approaches involving
machine learning on audio recordings are becoming increasingly popular.
Unfortunately, available datasets from this field of research are often not
optimal learning material; Samples can be weakly labeled, of different lengths
or come with a poor signal-to-noise ratio. In this work, we introduce a
generalized approach that first relabels subsegments of MEL spectrogram
representations, to achieve higher performances on the actual multi-class
classification tasks. For both the binary pre-sorting and the classification,
we make use of convolutional neural networks (CNN) and various
data-augmentation techniques. We showcase the results of this approach on the
challenging \textit{ComparE 2021} dataset, with the task of classifying between
different primate species sounds, and report significantly higher Accuracy and
UAR scores in contrast to comparatively equipped model baselines.
| no_new_dataset | 0.710848 |
2306.16133 | LUCAS MEYER | Lucas Meyer (EDF R\&D, SINCLAIR AI Lab, DATAMOVE ), Marc Schouler
(DATAMOVE ), Robert Alexander Caulk (DATAMOVE ), Alejandro Rib\'es (SINCLAIR
AI Lab, EDF R\&D), Bruno Raffin (DATAMOVE ) | Training Deep Surrogate Models with Large Scale Online Learning | null | null | null | null | cs.AI cs.DC physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The spatiotemporal resolution of Partial Differential Equations (PDEs) plays
important roles in the mathematical description of the world's physical
phenomena. In general, scientists and engineers solve PDEs numerically by the
use of computationally demanding solvers. Recently, deep learning algorithms
have emerged as a viable alternative for obtaining fast solutions for PDEs.
Models are usually trained on synthetic data generated by solvers, stored on
disk and read back for training. This paper advocates that relying on a
traditional static dataset to train these models does not allow the full
benefit of the solver to be used as a data generator. It proposes an open
source online training framework for deep surrogate models. The framework
implements several levels of parallelism focused on simultaneously generating
numerical simulations and training deep neural networks. This approach
suppresses the I/O and storage bottleneck associated with disk-loaded datasets,
and opens the way to training on significantly larger datasets. Experiments
compare the offline and online training of four surrogate models, including
state-of-the-art architectures. Results indicate that exposing deep surrogate
models to more dataset diversity, up to hundreds of GB, can increase model
generalization capabilities. Fully connected neural networks, Fourier Neural
Operator (FNO), and Message Passing PDE Solver prediction accuracy is improved
by 68%, 16% and 7%, respectively.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 12:02:27 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Meyer",
"Lucas",
"",
"EDF R\\&D, SINCLAIR AI Lab, DATAMOVE"
],
[
"Schouler",
"Marc",
"",
"DATAMOVE"
],
[
"Caulk",
"Robert Alexander",
"",
"DATAMOVE"
],
[
"Ribés",
"Alejandro",
"",
"SINCLAIR\n AI Lab, EDF R\\&D"
],
[
"Raffin",
"Bruno",
"",
"DATAMOVE"
]
] | TITLE: Training Deep Surrogate Models with Large Scale Online Learning
ABSTRACT: The spatiotemporal resolution of Partial Differential Equations (PDEs) plays
important roles in the mathematical description of the world's physical
phenomena. In general, scientists and engineers solve PDEs numerically by the
use of computationally demanding solvers. Recently, deep learning algorithms
have emerged as a viable alternative for obtaining fast solutions for PDEs.
Models are usually trained on synthetic data generated by solvers, stored on
disk and read back for training. This paper advocates that relying on a
traditional static dataset to train these models does not allow the full
benefit of the solver to be used as a data generator. It proposes an open
source online training framework for deep surrogate models. The framework
implements several levels of parallelism focused on simultaneously generating
numerical simulations and training deep neural networks. This approach
suppresses the I/O and storage bottleneck associated with disk-loaded datasets,
and opens the way to training on significantly larger datasets. Experiments
compare the offline and online training of four surrogate models, including
state-of-the-art architectures. Results indicate that exposing deep surrogate
models to more dataset diversity, up to hundreds of GB, can increase model
generalization capabilities. Fully connected neural networks, Fourier Neural
Operator (FNO), and Message Passing PDE Solver prediction accuracy is improved
by 68%, 16% and 7%, respectively.
| no_new_dataset | 0.712226 |
2306.16171 | Morteza Zakeri-Nasrabadi | Morteza Zakeri-Nasrabadi and Saeed Parsa and Mohammad Ramezani and
Chanchal Roy and Masoud Ekhtiarzadeh | A systematic literature review on source code similarity measurement and
clone detection: techniques, applications, and challenges | 49 pages, 10 figures, 6 tables | null | null | null | cs.SE cs.AI cs.PL | http://creativecommons.org/licenses/by/4.0/ | Measuring and evaluating source code similarity is a fundamental software
engineering activity that embraces a broad range of applications, including but
not limited to code recommendation, duplicate code, plagiarism, malware, and
smell detection. This paper proposes a systematic literature review and
meta-analysis on code similarity measurement and evaluation techniques to shed
light on the existing approaches and their characteristics in different
applications. We initially found over 10000 articles by querying four digital
libraries and ended up with 136 primary studies in the field. The studies were
classified according to their methodology, programming languages, datasets,
tools, and applications. A deep investigation reveals 80 software tools,
working with eight different techniques on five application domains. Nearly 49%
of the tools work on Java programs and 37% support C and C++, while there is no
support for many programming languages. A noteworthy point was the existence of
12 datasets related to source code similarity measurement and duplicate codes,
of which only eight datasets were publicly accessible. The lack of reliable
datasets, empirical evaluations, hybrid methods, and focuses on multi-paradigm
languages are the main challenges in the field. Emerging applications of code
similarity measurement concentrate on the development phase in addition to the
maintenance.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 12:49:22 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Zakeri-Nasrabadi",
"Morteza",
""
],
[
"Parsa",
"Saeed",
""
],
[
"Ramezani",
"Mohammad",
""
],
[
"Roy",
"Chanchal",
""
],
[
"Ekhtiarzadeh",
"Masoud",
""
]
] | TITLE: A systematic literature review on source code similarity measurement and
clone detection: techniques, applications, and challenges
ABSTRACT: Measuring and evaluating source code similarity is a fundamental software
engineering activity that embraces a broad range of applications, including but
not limited to code recommendation, duplicate code, plagiarism, malware, and
smell detection. This paper proposes a systematic literature review and
meta-analysis on code similarity measurement and evaluation techniques to shed
light on the existing approaches and their characteristics in different
applications. We initially found over 10000 articles by querying four digital
libraries and ended up with 136 primary studies in the field. The studies were
classified according to their methodology, programming languages, datasets,
tools, and applications. A deep investigation reveals 80 software tools,
working with eight different techniques on five application domains. Nearly 49%
of the tools work on Java programs and 37% support C and C++, while there is no
support for many programming languages. A noteworthy point was the existence of
12 datasets related to source code similarity measurement and duplicate codes,
of which only eight datasets were publicly accessible. The lack of reliable
datasets, empirical evaluations, hybrid methods, and focuses on multi-paradigm
languages are the main challenges in the field. Emerging applications of code
similarity measurement concentrate on the development phase in addition to the
maintenance.
| no_new_dataset | 0.708011 |
2306.16176 | Zhangyin Feng | Zhangyin Feng, Yong Dai, Fan Zhang, Duyu Tang, Xiaocheng Feng,
Shuangzhi Wu, Bing Qin, Yunbo Cao and Shuming Shi | SkillNet-X: A Multilingual Multitask Model with Sparsely Activated
Skills | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Traditional multitask learning methods basically can only exploit common
knowledge in task- or language-wise, which lose either cross-language or
cross-task knowledge. This paper proposes a general multilingual multitask
model, named SkillNet-X, which enables a single model to tackle many different
tasks from different languages. To this end, we define several
language-specific skills and task-specific skills, each of which corresponds to
a skill module. SkillNet-X sparsely activates parts of the skill modules which
are relevant either to the target task or the target language. Acting as
knowledge transit hubs, skill modules are capable of absorbing task-related
knowledge and language-related knowledge consecutively. Based on Transformer,
we modify the multi-head attention layer and the feed forward network layer to
accommodate skill modules. We evaluate SkillNet-X on eleven natural language
understanding datasets in four languages. Results show that SkillNet-X performs
better than task-specific baselines and two multitask learning baselines (i.e.,
dense joint model and Mixture-of-Experts model). Furthermore, skill
pre-training further improves the performance of SkillNet-X on almost all
datasets. To investigate the generalization of our model, we conduct
experiments on two new tasks and find that SkillNet-X significantly outperforms
baselines.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 12:53:30 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Feng",
"Zhangyin",
""
],
[
"Dai",
"Yong",
""
],
[
"Zhang",
"Fan",
""
],
[
"Tang",
"Duyu",
""
],
[
"Feng",
"Xiaocheng",
""
],
[
"Wu",
"Shuangzhi",
""
],
[
"Qin",
"Bing",
""
],
[
"Cao",
"Yunbo",
""
],
[
"Shi",
"Shuming",
""
]
] | TITLE: SkillNet-X: A Multilingual Multitask Model with Sparsely Activated
Skills
ABSTRACT: Traditional multitask learning methods basically can only exploit common
knowledge in task- or language-wise, which lose either cross-language or
cross-task knowledge. This paper proposes a general multilingual multitask
model, named SkillNet-X, which enables a single model to tackle many different
tasks from different languages. To this end, we define several
language-specific skills and task-specific skills, each of which corresponds to
a skill module. SkillNet-X sparsely activates parts of the skill modules which
are relevant either to the target task or the target language. Acting as
knowledge transit hubs, skill modules are capable of absorbing task-related
knowledge and language-related knowledge consecutively. Based on Transformer,
we modify the multi-head attention layer and the feed forward network layer to
accommodate skill modules. We evaluate SkillNet-X on eleven natural language
understanding datasets in four languages. Results show that SkillNet-X performs
better than task-specific baselines and two multitask learning baselines (i.e.,
dense joint model and Mixture-of-Experts model). Furthermore, skill
pre-training further improves the performance of SkillNet-X on almost all
datasets. To investigate the generalization of our model, we conduct
experiments on two new tasks and find that SkillNet-X significantly outperforms
baselines.
| no_new_dataset | 0.70676 |
2306.16186 | Zhewei Chen | Zhewei Chen, Wai Keung Wong, Zuofeng Zhong, Jinpiao Liao, Ying Qu | Effective Transfer of Pretrained Large Visual Model for Fabric Defect
Segmentation via Specifc Knowledge Injection | 13 pages,4 figures, 3 tables | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Fabric defect segmentation is integral to textile quality control. Despite
this, the scarcity of high-quality annotated data and the diversity of fabric
defects present significant challenges to the application of deep learning in
this field. These factors limit the generalization and segmentation performance
of existing models, impeding their ability to handle the complexity of diverse
fabric types and defects. To overcome these obstacles, this study introduces an
innovative method to infuse specialized knowledge of fabric defects into the
Segment Anything Model (SAM), a large-scale visual model. By introducing and
training a unique set of fabric defect-related parameters, this approach
seamlessly integrates domain-specific knowledge into SAM without the need for
extensive modifications to the pre-existing model parameters. The revamped SAM
model leverages generalized image understanding learned from large-scale
natural image datasets while incorporating fabric defect-specific knowledge,
ensuring its proficiency in fabric defect segmentation tasks. The experimental
results reveal a significant improvement in the model's segmentation
performance, attributable to this novel amalgamation of generic and
fabric-specific knowledge. When benchmarking against popular existing
segmentation models across three datasets, our proposed model demonstrates a
substantial leap in performance. Its impressive results in cross-dataset
comparisons and few-shot learning experiments further demonstrate its potential
for practical applications in textile quality control.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 13:08:08 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Chen",
"Zhewei",
""
],
[
"Wong",
"Wai Keung",
""
],
[
"Zhong",
"Zuofeng",
""
],
[
"Liao",
"Jinpiao",
""
],
[
"Qu",
"Ying",
""
]
] | TITLE: Effective Transfer of Pretrained Large Visual Model for Fabric Defect
Segmentation via Specifc Knowledge Injection
ABSTRACT: Fabric defect segmentation is integral to textile quality control. Despite
this, the scarcity of high-quality annotated data and the diversity of fabric
defects present significant challenges to the application of deep learning in
this field. These factors limit the generalization and segmentation performance
of existing models, impeding their ability to handle the complexity of diverse
fabric types and defects. To overcome these obstacles, this study introduces an
innovative method to infuse specialized knowledge of fabric defects into the
Segment Anything Model (SAM), a large-scale visual model. By introducing and
training a unique set of fabric defect-related parameters, this approach
seamlessly integrates domain-specific knowledge into SAM without the need for
extensive modifications to the pre-existing model parameters. The revamped SAM
model leverages generalized image understanding learned from large-scale
natural image datasets while incorporating fabric defect-specific knowledge,
ensuring its proficiency in fabric defect segmentation tasks. The experimental
results reveal a significant improvement in the model's segmentation
performance, attributable to this novel amalgamation of generic and
fabric-specific knowledge. When benchmarking against popular existing
segmentation models across three datasets, our proposed model demonstrates a
substantial leap in performance. Its impressive results in cross-dataset
comparisons and few-shot learning experiments further demonstrate its potential
for practical applications in textile quality control.
| no_new_dataset | 0.712282 |
2306.16201 | Guandu Liu | Guandu Liu, Fangyuan Zhang, Tianxiang Pan, Bin Wang | Low-Confidence Samples Mining for Semi-supervised Object Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reliable pseudo-labels from unlabeled data play a key role in semi-supervised
object detection (SSOD). However, the state-of-the-art SSOD methods all rely on
pseudo-labels with high confidence, which ignore valuable pseudo-labels with
lower confidence. Additionally, the insufficient excavation for unlabeled data
results in an excessively low recall rate thus hurting the network training. In
this paper, we propose a novel Low-confidence Samples Mining (LSM) method to
utilize low-confidence pseudo-labels efficiently. Specifically, we develop an
additional pseudo information mining (PIM) branch on account of low-resolution
feature maps to extract reliable large-area instances, the IoUs of which are
higher than small-area ones. Owing to the complementary predictions between PIM
and the main branch, we further design self-distillation (SD) to compensate for
both in a mutually-learning manner. Meanwhile, the extensibility of the above
approaches enables our LSM to apply to Faster-RCNN and Deformable-DETR
respectively. On the MS-COCO benchmark, our method achieves 3.54% mAP
improvement over state-of-the-art methods under 5% labeling ratios.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 13:29:06 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Liu",
"Guandu",
""
],
[
"Zhang",
"Fangyuan",
""
],
[
"Pan",
"Tianxiang",
""
],
[
"Wang",
"Bin",
""
]
] | TITLE: Low-Confidence Samples Mining for Semi-supervised Object Detection
ABSTRACT: Reliable pseudo-labels from unlabeled data play a key role in semi-supervised
object detection (SSOD). However, the state-of-the-art SSOD methods all rely on
pseudo-labels with high confidence, which ignore valuable pseudo-labels with
lower confidence. Additionally, the insufficient excavation for unlabeled data
results in an excessively low recall rate thus hurting the network training. In
this paper, we propose a novel Low-confidence Samples Mining (LSM) method to
utilize low-confidence pseudo-labels efficiently. Specifically, we develop an
additional pseudo information mining (PIM) branch on account of low-resolution
feature maps to extract reliable large-area instances, the IoUs of which are
higher than small-area ones. Owing to the complementary predictions between PIM
and the main branch, we further design self-distillation (SD) to compensate for
both in a mutually-learning manner. Meanwhile, the extensibility of the above
approaches enables our LSM to apply to Faster-RCNN and Deformable-DETR
respectively. On the MS-COCO benchmark, our method achieves 3.54% mAP
improvement over state-of-the-art methods under 5% labeling ratios.
| no_new_dataset | 0.710867 |
2306.16244 | Yufei Huang | Yufei Huang and Deyi Xiong | CBBQ: A Chinese Bias Benchmark Dataset Curated with Human-AI
Collaboration for Large Language Models | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Holistically measuring societal biases of large language models is crucial
for detecting and reducing ethical risks in highly capable AI models. In this
work, we present a Chinese Bias Benchmark dataset that consists of over 100K
questions jointly constructed by human experts and generative language models,
covering stereotypes and societal biases in 14 social dimensions related to
Chinese culture and values. The curation process contains 4 essential steps:
bias identification via extensive literature review, ambiguous context
generation, AI-assisted disambiguous context generation, snd manual review \&
recomposition. The testing instances in the dataset are automatically derived
from 3K+ high-quality templates manually authored with stringent quality
control. The dataset exhibits wide coverage and high diversity. Extensive
experiments demonstrate the effectiveness of the dataset in detecting model
bias, with all 10 publicly available Chinese large language models exhibiting
strong bias in certain categories. Additionally, we observe from our
experiments that fine-tuned models could, to a certain extent, heed
instructions and avoid generating outputs that are morally harmful in some
types, in the way of "moral self-correction". Our dataset and results are
publicly available at
\href{https://github.com/YFHuangxxxx/CBBQ}{https://github.com/YFHuangxxxx/CBBQ},
offering debiasing research opportunities to a widened community.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 14:14:44 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Huang",
"Yufei",
""
],
[
"Xiong",
"Deyi",
""
]
] | TITLE: CBBQ: A Chinese Bias Benchmark Dataset Curated with Human-AI
Collaboration for Large Language Models
ABSTRACT: Holistically measuring societal biases of large language models is crucial
for detecting and reducing ethical risks in highly capable AI models. In this
work, we present a Chinese Bias Benchmark dataset that consists of over 100K
questions jointly constructed by human experts and generative language models,
covering stereotypes and societal biases in 14 social dimensions related to
Chinese culture and values. The curation process contains 4 essential steps:
bias identification via extensive literature review, ambiguous context
generation, AI-assisted disambiguous context generation, snd manual review \&
recomposition. The testing instances in the dataset are automatically derived
from 3K+ high-quality templates manually authored with stringent quality
control. The dataset exhibits wide coverage and high diversity. Extensive
experiments demonstrate the effectiveness of the dataset in detecting model
bias, with all 10 publicly available Chinese large language models exhibiting
strong bias in certain categories. Additionally, we observe from our
experiments that fine-tuned models could, to a certain extent, heed
instructions and avoid generating outputs that are morally harmful in some
types, in the way of "moral self-correction". Our dataset and results are
publicly available at
\href{https://github.com/YFHuangxxxx/CBBQ}{https://github.com/YFHuangxxxx/CBBQ},
offering debiasing research opportunities to a widened community.
| new_dataset | 0.718175 |
2306.16250 | Jun Chen | Jun Chen, Wei Rao, Zilin Wang, Jiuxin Lin, Yukai Ju, Shulin He, Yannan
Wang, Zhiyong Wu | MC-SpEx: Towards Effective Speaker Extraction with Multi-Scale
Interfusion and Conditional Speaker Modulation | Accepted by InterSpeech 2023 | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by-sa/4.0/ | The previous SpEx+ has yielded outstanding performance in speaker extraction
and attracted much attention. However, it still encounters inadequate
utilization of multi-scale information and speaker embedding. To this end, this
paper proposes a new effective speaker extraction system with multi-scale
interfusion and conditional speaker modulation (ConSM), which is called
MC-SpEx. First of all, we design the weight-share multi-scale fusers
(ScaleFusers) for efficiently leveraging multi-scale information as well as
ensuring consistency of the model's feature space. Then, to consider different
scale information while generating masks, the multi-scale interactive mask
generator (ScaleInterMG) is presented. Moreover, we introduce ConSM module to
fully exploit speaker embedding in the speech extractor. Experimental results
on the Libri2Mix dataset demonstrate the effectiveness of our improvements and
the state-of-the-art performance of our proposed MC-SpEx.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 14:21:06 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Chen",
"Jun",
""
],
[
"Rao",
"Wei",
""
],
[
"Wang",
"Zilin",
""
],
[
"Lin",
"Jiuxin",
""
],
[
"Ju",
"Yukai",
""
],
[
"He",
"Shulin",
""
],
[
"Wang",
"Yannan",
""
],
[
"Wu",
"Zhiyong",
""
]
] | TITLE: MC-SpEx: Towards Effective Speaker Extraction with Multi-Scale
Interfusion and Conditional Speaker Modulation
ABSTRACT: The previous SpEx+ has yielded outstanding performance in speaker extraction
and attracted much attention. However, it still encounters inadequate
utilization of multi-scale information and speaker embedding. To this end, this
paper proposes a new effective speaker extraction system with multi-scale
interfusion and conditional speaker modulation (ConSM), which is called
MC-SpEx. First of all, we design the weight-share multi-scale fusers
(ScaleFusers) for efficiently leveraging multi-scale information as well as
ensuring consistency of the model's feature space. Then, to consider different
scale information while generating masks, the multi-scale interactive mask
generator (ScaleInterMG) is presented. Moreover, we introduce ConSM module to
fully exploit speaker embedding in the speech extractor. Experimental results
on the Libri2Mix dataset demonstrate the effectiveness of our improvements and
the state-of-the-art performance of our proposed MC-SpEx.
| no_new_dataset | 0.708452 |
2306.16268 | Mohammad Ali Hussiny | Mohammad Ali Hussiny, Lilja {\O}vrelid | Emotion Analysis of Tweets Banning Education in Afghanistan | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by/4.0/ | This paper introduces the first emotion annotated dataset for the Dari
variant of Persian spoken in Afghanistan. The LetHerLearn dataset contains
7,600 tweets posted in reaction to the Taliban ban of women rights to education
in 2022 and has been manually annotated according to Ekman emotion categories.
We here detail the data collection and annotation process, present relevant
dataset statistics as well as initial experiments on the resulting dataset,
benchmarking a number of different neural architectures for the task of Dari
emotion classification.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 14:50:49 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Hussiny",
"Mohammad Ali",
""
],
[
"Øvrelid",
"Lilja",
""
]
] | TITLE: Emotion Analysis of Tweets Banning Education in Afghanistan
ABSTRACT: This paper introduces the first emotion annotated dataset for the Dari
variant of Persian spoken in Afghanistan. The LetHerLearn dataset contains
7,600 tweets posted in reaction to the Taliban ban of women rights to education
in 2022 and has been manually annotated according to Ekman emotion categories.
We here detail the data collection and annotation process, present relevant
dataset statistics as well as initial experiments on the resulting dataset,
benchmarking a number of different neural architectures for the task of Dari
emotion classification.
| new_dataset | 0.715418 |
2306.16270 | Ke Liu | Ke Liu and Kaifan Yang and Jiahong Zhang and Renjun Xu | S2SNet: A Pretrained Neural Network for Superconductivity Discovery | Accepted to IJCAI 2022 | null | null | null | cond-mat.supr-con cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Superconductivity allows electrical current to flow without any energy loss,
and thus making solids superconducting is a grand goal of physics, material
science, and electrical engineering. More than 16 Nobel Laureates have been
awarded for their contribution to superconductivity research. Superconductors
are valuable for sustainable development goals (SDGs), such as climate change
mitigation, affordable and clean energy, industry, innovation and
infrastructure, and so on. However, a unified physics theory explaining all
superconductivity mechanism is still unknown. It is believed that
superconductivity is microscopically due to not only molecular compositions but
also the geometric crystal structure. Hence a new dataset, S2S, containing both
crystal structures and superconducting critical temperature, is built upon
SuperCon and Material Project. Based on this new dataset, we propose a novel
model, S2SNet, which utilizes the attention mechanism for superconductivity
prediction. To overcome the shortage of data, S2SNet is pre-trained on the
whole Material Project dataset with Masked-Language Modeling (MLM). S2SNet
makes a new state-of-the-art, with out-of-sample accuracy of 92% and Area Under
Curve (AUC) of 0.92. To the best of our knowledge, S2SNet is the first work to
predict superconductivity with only information of crystal structures. This
work is beneficial to superconductivity discovery and further SDGs. Code and
datasets are available in https://github.com/zjuKeLiu/S2SNet
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 14:52:50 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Liu",
"Ke",
""
],
[
"Yang",
"Kaifan",
""
],
[
"Zhang",
"Jiahong",
""
],
[
"Xu",
"Renjun",
""
]
] | TITLE: S2SNet: A Pretrained Neural Network for Superconductivity Discovery
ABSTRACT: Superconductivity allows electrical current to flow without any energy loss,
and thus making solids superconducting is a grand goal of physics, material
science, and electrical engineering. More than 16 Nobel Laureates have been
awarded for their contribution to superconductivity research. Superconductors
are valuable for sustainable development goals (SDGs), such as climate change
mitigation, affordable and clean energy, industry, innovation and
infrastructure, and so on. However, a unified physics theory explaining all
superconductivity mechanism is still unknown. It is believed that
superconductivity is microscopically due to not only molecular compositions but
also the geometric crystal structure. Hence a new dataset, S2S, containing both
crystal structures and superconducting critical temperature, is built upon
SuperCon and Material Project. Based on this new dataset, we propose a novel
model, S2SNet, which utilizes the attention mechanism for superconductivity
prediction. To overcome the shortage of data, S2SNet is pre-trained on the
whole Material Project dataset with Masked-Language Modeling (MLM). S2SNet
makes a new state-of-the-art, with out-of-sample accuracy of 92% and Area Under
Curve (AUC) of 0.92. To the best of our knowledge, S2SNet is the first work to
predict superconductivity with only information of crystal structures. This
work is beneficial to superconductivity discovery and further SDGs. Code and
datasets are available in https://github.com/zjuKeLiu/S2SNet
| new_dataset | 0.716572 |
2306.16285 | An Wang | An Wang, Mobarakol Islam, Mengya Xu, Hongliang Ren | Generalizing Surgical Instruments Segmentation to Unseen Domains with
One-to-Many Synthesis | First two authors contributed equally. Accepted by IROS2023 | null | null | null | eess.IV cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite their impressive performance in various surgical scene understanding
tasks, deep learning-based methods are frequently hindered from deploying to
real-world surgical applications for various causes. Particularly, data
collection, annotation, and domain shift in-between sites and patients are the
most common obstacles. In this work, we mitigate data-related issues by
efficiently leveraging minimal source images to generate synthetic surgical
instrument segmentation datasets and achieve outstanding generalization
performance on unseen real domains. Specifically, in our framework, only one
background tissue image and at most three images of each foreground instrument
are taken as the seed images. These source images are extensively transformed
and employed to build up the foreground and background image pools, from which
randomly sampled tissue and instrument images are composed with multiple
blending techniques to generate new surgical scene images. Besides, we
introduce hybrid training-time augmentations to diversify the training data
further. Extensive evaluation on three real-world datasets, i.e., Endo2017,
Endo2018, and RoboTool, demonstrates that our one-to-many synthetic surgical
instruments datasets generation and segmentation framework can achieve
encouraging performance compared with training with real data. Notably, on the
RoboTool dataset, where a more significant domain gap exists, our framework
shows its superiority of generalization by a considerable margin. We expect
that our inspiring results will attract research attention to improving model
generalization with data synthesizing.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 15:06:44 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Wang",
"An",
""
],
[
"Islam",
"Mobarakol",
""
],
[
"Xu",
"Mengya",
""
],
[
"Ren",
"Hongliang",
""
]
] | TITLE: Generalizing Surgical Instruments Segmentation to Unseen Domains with
One-to-Many Synthesis
ABSTRACT: Despite their impressive performance in various surgical scene understanding
tasks, deep learning-based methods are frequently hindered from deploying to
real-world surgical applications for various causes. Particularly, data
collection, annotation, and domain shift in-between sites and patients are the
most common obstacles. In this work, we mitigate data-related issues by
efficiently leveraging minimal source images to generate synthetic surgical
instrument segmentation datasets and achieve outstanding generalization
performance on unseen real domains. Specifically, in our framework, only one
background tissue image and at most three images of each foreground instrument
are taken as the seed images. These source images are extensively transformed
and employed to build up the foreground and background image pools, from which
randomly sampled tissue and instrument images are composed with multiple
blending techniques to generate new surgical scene images. Besides, we
introduce hybrid training-time augmentations to diversify the training data
further. Extensive evaluation on three real-world datasets, i.e., Endo2017,
Endo2018, and RoboTool, demonstrates that our one-to-many synthetic surgical
instruments datasets generation and segmentation framework can achieve
encouraging performance compared with training with real data. Notably, on the
RoboTool dataset, where a more significant domain gap exists, our framework
shows its superiority of generalization by a considerable margin. We expect
that our inspiring results will attract research attention to improving model
generalization with data synthesizing.
| no_new_dataset | 0.699896 |
2306.16313 | Fanyu Wang | Fanyu Wang and Zhenping Xie | An Adversarial Multi-Task Learning Method for Chinese Text Correction
with Semantic Detection | Published on 31st International Conference on Artificial Neural
Network | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Text correction, especially the semantic correction of more widely used
scenes, is strongly required to improve, for the fluency and writing efficiency
of the text. An adversarial multi-task learning method is proposed to enhance
the modeling and detection ability of character polysemy in Chinese sentence
context. Wherein, two models, the masked language model and scoring language
model, are introduced as a pair of not only coupled but also adversarial
learning tasks. Moreover, the Monte Carlo tree search strategy and a policy
network are introduced to accomplish the efficient Chinese text correction task
with semantic detection. The experiments are executed on three datasets and
five comparable methods, and the experimental results show that our method can
obtain good performance in Chinese text correction task for better semantic
rationality.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 15:46:00 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Wang",
"Fanyu",
""
],
[
"Xie",
"Zhenping",
""
]
] | TITLE: An Adversarial Multi-Task Learning Method for Chinese Text Correction
with Semantic Detection
ABSTRACT: Text correction, especially the semantic correction of more widely used
scenes, is strongly required to improve, for the fluency and writing efficiency
of the text. An adversarial multi-task learning method is proposed to enhance
the modeling and detection ability of character polysemy in Chinese sentence
context. Wherein, two models, the masked language model and scoring language
model, are introduced as a pair of not only coupled but also adversarial
learning tasks. Moreover, the Monte Carlo tree search strategy and a policy
network are introduced to accomplish the efficient Chinese text correction task
with semantic detection. The experiments are executed on three datasets and
five comparable methods, and the experimental results show that our method can
obtain good performance in Chinese text correction task for better semantic
rationality.
| no_new_dataset | 0.710829 |
2306.16322 | Zaid Alyafeai Mr | Zaid Alyafeai and Maged S. Alshaibani and Badr AlKhamissi and Hamzah
Luqman and Ebrahim Alareqi and Ali Fadel | Taqyim: Evaluating Arabic NLP Tasks Using ChatGPT Models | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Large language models (LLMs) have demonstrated impressive performance on
various downstream tasks without requiring fine-tuning, including ChatGPT, a
chat-based model built on top of LLMs such as GPT-3.5 and GPT-4. Despite having
a lower training proportion compared to English, these models also exhibit
remarkable capabilities in other languages. In this study, we assess the
performance of GPT-3.5 and GPT-4 models on seven distinct Arabic NLP tasks:
sentiment analysis, translation, transliteration, paraphrasing, part of speech
tagging, summarization, and diacritization. Our findings reveal that GPT-4
outperforms GPT-3.5 on five out of the seven tasks. Furthermore, we conduct an
extensive analysis of the sentiment analysis task, providing insights into how
LLMs achieve exceptional results on a challenging dialectal dataset.
Additionally, we introduce a new Python interface
https://github.com/ARBML/Taqyim that facilitates the evaluation of these tasks
effortlessly.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 15:54:29 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Alyafeai",
"Zaid",
""
],
[
"Alshaibani",
"Maged S.",
""
],
[
"AlKhamissi",
"Badr",
""
],
[
"Luqman",
"Hamzah",
""
],
[
"Alareqi",
"Ebrahim",
""
],
[
"Fadel",
"Ali",
""
]
] | TITLE: Taqyim: Evaluating Arabic NLP Tasks Using ChatGPT Models
ABSTRACT: Large language models (LLMs) have demonstrated impressive performance on
various downstream tasks without requiring fine-tuning, including ChatGPT, a
chat-based model built on top of LLMs such as GPT-3.5 and GPT-4. Despite having
a lower training proportion compared to English, these models also exhibit
remarkable capabilities in other languages. In this study, we assess the
performance of GPT-3.5 and GPT-4 models on seven distinct Arabic NLP tasks:
sentiment analysis, translation, transliteration, paraphrasing, part of speech
tagging, summarization, and diacritization. Our findings reveal that GPT-4
outperforms GPT-3.5 on five out of the seven tasks. Furthermore, we conduct an
extensive analysis of the sentiment analysis task, providing insights into how
LLMs achieve exceptional results on a challenging dialectal dataset.
Additionally, we introduce a new Python interface
https://github.com/ARBML/Taqyim that facilitates the evaluation of these tasks
effortlessly.
| no_new_dataset | 0.707771 |
2306.16326 | Oren Barkan | Oren Barkan, Avi Caciularu, Idan Rejwan, Ori Katz, Jonathan Weill,
Itzik Malkiel, Noam Koenigstein | Representation Learning via Variational Bayesian Networks | null | null | null | null | cs.LG cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present Variational Bayesian Network (VBN) - a novel Bayesian entity
representation learning model that utilizes hierarchical and relational side
information and is particularly useful for modeling entities in the
``long-tail'', where the data is scarce. VBN provides better modeling for
long-tail entities via two complementary mechanisms: First, VBN employs
informative hierarchical priors that enable information propagation between
entities sharing common ancestors. Additionally, VBN models explicit relations
between entities that enforce complementary structure and consistency, guiding
the learned representations towards a more meaningful arrangement in space.
Second, VBN represents entities by densities (rather than vectors), hence
modeling uncertainty that plays a complementary role in coping with data
scarcity. Finally, we propose a scalable Variational Bayes optimization
algorithm that enables fast approximate Bayesian inference. We evaluate the
effectiveness of VBN on linguistic, recommendations, and medical inference
tasks. Our findings show that VBN outperforms other existing methods across
multiple datasets, and especially in the long-tail.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 16:00:45 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Barkan",
"Oren",
""
],
[
"Caciularu",
"Avi",
""
],
[
"Rejwan",
"Idan",
""
],
[
"Katz",
"Ori",
""
],
[
"Weill",
"Jonathan",
""
],
[
"Malkiel",
"Itzik",
""
],
[
"Koenigstein",
"Noam",
""
]
] | TITLE: Representation Learning via Variational Bayesian Networks
ABSTRACT: We present Variational Bayesian Network (VBN) - a novel Bayesian entity
representation learning model that utilizes hierarchical and relational side
information and is particularly useful for modeling entities in the
``long-tail'', where the data is scarce. VBN provides better modeling for
long-tail entities via two complementary mechanisms: First, VBN employs
informative hierarchical priors that enable information propagation between
entities sharing common ancestors. Additionally, VBN models explicit relations
between entities that enforce complementary structure and consistency, guiding
the learned representations towards a more meaningful arrangement in space.
Second, VBN represents entities by densities (rather than vectors), hence
modeling uncertainty that plays a complementary role in coping with data
scarcity. Finally, we propose a scalable Variational Bayes optimization
algorithm that enables fast approximate Bayesian inference. We evaluate the
effectiveness of VBN on linguistic, recommendations, and medical inference
tasks. Our findings show that VBN outperforms other existing methods across
multiple datasets, and especially in the long-tail.
| no_new_dataset | 0.709831 |
2306.16329 | Ruihang Chu | Ruihang Chu, Enze Xie, Shentong Mo, Zhenguo Li, Matthias Nie{\ss}ner,
Chi-Wing Fu, Jiaya Jia | DiffComplete: Diffusion-based Generative 3D Shape Completion | Project Page: https://ruihangchu.com/diffcomplete.html | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce a new diffusion-based approach for shape completion on 3D range
scans. Compared with prior deterministic and probabilistic methods, we strike a
balance between realism, multi-modality, and high fidelity. We propose
DiffComplete by casting shape completion as a generative task conditioned on
the incomplete shape. Our key designs are two-fold. First, we devise a
hierarchical feature aggregation mechanism to inject conditional features in a
spatially-consistent manner. So, we can capture both local details and broader
contexts of the conditional inputs to control the shape completion. Second, we
propose an occupancy-aware fusion strategy in our model to enable the
completion of multiple partial shapes and introduce higher flexibility on the
input conditions. DiffComplete sets a new SOTA performance (e.g., 40% decrease
on l_1 error) on two large-scale 3D shape completion benchmarks. Our completed
shapes not only have a realistic outlook compared with the deterministic
methods but also exhibit high similarity to the ground truths compared with the
probabilistic alternatives. Further, DiffComplete has strong generalizability
on objects of entirely unseen classes for both synthetic and real data,
eliminating the need for model re-training in various applications.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 16:07:36 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Chu",
"Ruihang",
""
],
[
"Xie",
"Enze",
""
],
[
"Mo",
"Shentong",
""
],
[
"Li",
"Zhenguo",
""
],
[
"Nießner",
"Matthias",
""
],
[
"Fu",
"Chi-Wing",
""
],
[
"Jia",
"Jiaya",
""
]
] | TITLE: DiffComplete: Diffusion-based Generative 3D Shape Completion
ABSTRACT: We introduce a new diffusion-based approach for shape completion on 3D range
scans. Compared with prior deterministic and probabilistic methods, we strike a
balance between realism, multi-modality, and high fidelity. We propose
DiffComplete by casting shape completion as a generative task conditioned on
the incomplete shape. Our key designs are two-fold. First, we devise a
hierarchical feature aggregation mechanism to inject conditional features in a
spatially-consistent manner. So, we can capture both local details and broader
contexts of the conditional inputs to control the shape completion. Second, we
propose an occupancy-aware fusion strategy in our model to enable the
completion of multiple partial shapes and introduce higher flexibility on the
input conditions. DiffComplete sets a new SOTA performance (e.g., 40% decrease
on l_1 error) on two large-scale 3D shape completion benchmarks. Our completed
shapes not only have a realistic outlook compared with the deterministic
methods but also exhibit high similarity to the ground truths compared with the
probabilistic alternatives. Further, DiffComplete has strong generalizability
on objects of entirely unseen classes for both synthetic and real data,
eliminating the need for model re-training in various applications.
| no_new_dataset | 0.709372 |
2306.16357 | Slimane Larabi | Leyla Benhamida, Slimane Larabi | Theater Aid System for the Visually Impaired Through Transfer Learning
of Spatio-Temporal Graph Convolution Networks | 28 pages, 18 figures, research paper not published | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The aim of this research is to recognize human actions performed on stage to
aid visually impaired and blind individuals. To achieve this, we have created a
theatre human action recognition system that uses skeleton data captured by
depth image as input. We collected new samples of human actions in a theatre
environment, and then tested the transfer learning technique with three
pre-trained Spatio-Temporal Graph Convolution Networks for skeleton-based human
action recognition: the spatio-temporal graph convolution network, the
two-stream adaptive graph convolution network, and the multi-scale disentangled
unified graph convolution network. We selected the NTU-RGBD human action
benchmark as the source domain and used our collected dataset as the target
domain. We analyzed the transferability of the pre-trained models and proposed
two configurations to apply and adapt the transfer learning technique to the
diversity between the source and target domains. The use of transfer learning
helped to improve the performance of the human action system within the context
of theatre. The results indicate that Spatio-Temporal Graph Convolution
Networks is positively transferred, and there was an improvement in performance
compared to the baseline without transfer learning.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 16:40:52 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Benhamida",
"Leyla",
""
],
[
"Larabi",
"Slimane",
""
]
] | TITLE: Theater Aid System for the Visually Impaired Through Transfer Learning
of Spatio-Temporal Graph Convolution Networks
ABSTRACT: The aim of this research is to recognize human actions performed on stage to
aid visually impaired and blind individuals. To achieve this, we have created a
theatre human action recognition system that uses skeleton data captured by
depth image as input. We collected new samples of human actions in a theatre
environment, and then tested the transfer learning technique with three
pre-trained Spatio-Temporal Graph Convolution Networks for skeleton-based human
action recognition: the spatio-temporal graph convolution network, the
two-stream adaptive graph convolution network, and the multi-scale disentangled
unified graph convolution network. We selected the NTU-RGBD human action
benchmark as the source domain and used our collected dataset as the target
domain. We analyzed the transferability of the pre-trained models and proposed
two configurations to apply and adapt the transfer learning technique to the
diversity between the source and target domains. The use of transfer learning
helped to improve the performance of the human action system within the context
of theatre. The results indicate that Spatio-Temporal Graph Convolution
Networks is positively transferred, and there was an improvement in performance
compared to the baseline without transfer learning.
| new_dataset | 0.608536 |
2306.16398 | Richard Rose | Richard Rose, Oscar Chang, Olivier Siohan | Cascaded encoders for fine-tuning ASR models on overlapped speech | null | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by/4.0/ | Multi-talker speech recognition (MT-ASR) has been shown to improve ASR
performance on speech containing overlapping utterances from more than one
speaker. Multi-talker models have typically been trained from scratch using
simulated or actual overlapping speech datasets. On the other hand, the trend
in ASR has been to train foundation models using massive datasets collected
from a wide variety of task domains. Given the scale of these models and their
ability to generalize well across a variety of domains, it makes sense to
consider scenarios where a foundation model is augmented with multi-talker
capability. This paper presents an MT-ASR model formed by combining a
well-trained foundation model with a multi-talker mask model in a cascaded
RNN-T encoder configuration. Experimental results show that the cascade
configuration provides improved WER on overlapping speech utterances with
respect to a baseline multi-talker model without sacrificing performance
achievable by the foundation model on non-overlapping utterances.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 17:44:30 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Rose",
"Richard",
""
],
[
"Chang",
"Oscar",
""
],
[
"Siohan",
"Olivier",
""
]
] | TITLE: Cascaded encoders for fine-tuning ASR models on overlapped speech
ABSTRACT: Multi-talker speech recognition (MT-ASR) has been shown to improve ASR
performance on speech containing overlapping utterances from more than one
speaker. Multi-talker models have typically been trained from scratch using
simulated or actual overlapping speech datasets. On the other hand, the trend
in ASR has been to train foundation models using massive datasets collected
from a wide variety of task domains. Given the scale of these models and their
ability to generalize well across a variety of domains, it makes sense to
consider scenarios where a foundation model is augmented with multi-talker
capability. This paper presents an MT-ASR model formed by combining a
well-trained foundation model with a multi-talker mask model in a cascaded
RNN-T encoder configuration. Experimental results show that the cascade
configuration provides improved WER on overlapping speech utterances with
respect to a baseline multi-talker model without sacrificing performance
achievable by the foundation model on non-overlapping utterances.
| no_new_dataset | 0.710848 |
2306.16413 | Paul Pu Liang | Paul Pu Liang, Yiwei Lyu, Xiang Fan, Arav Agarwal, Yun Cheng,
Louis-Philippe Morency, Ruslan Salakhutdinov | MultiZoo & MultiBench: A Standardized Toolkit for Multimodal Deep
Learning | JMLR Open Source Software 2023, Code available at
https://github.com/pliang279/MultiBench | null | null | null | cs.LG cs.AI cs.CL cs.CV cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Learning multimodal representations involves integrating information from
multiple heterogeneous sources of data. In order to accelerate progress towards
understudied modalities and tasks while ensuring real-world robustness, we
release MultiZoo, a public toolkit consisting of standardized implementations
of > 20 core multimodal algorithms and MultiBench, a large-scale benchmark
spanning 15 datasets, 10 modalities, 20 prediction tasks, and 6 research areas.
Together, these provide an automated end-to-end machine learning pipeline that
simplifies and standardizes data loading, experimental setup, and model
evaluation. To enable holistic evaluation, we offer a comprehensive methodology
to assess (1) generalization, (2) time and space complexity, and (3) modality
robustness. MultiBench paves the way towards a better understanding of the
capabilities and limitations of multimodal models, while ensuring ease of use,
accessibility, and reproducibility. Our toolkits are publicly available, will
be regularly updated, and welcome inputs from the community.
| [
{
"version": "v1",
"created": "Wed, 28 Jun 2023 17:59:10 GMT"
}
] | 2023-06-29T00:00:00 | [
[
"Liang",
"Paul Pu",
""
],
[
"Lyu",
"Yiwei",
""
],
[
"Fan",
"Xiang",
""
],
[
"Agarwal",
"Arav",
""
],
[
"Cheng",
"Yun",
""
],
[
"Morency",
"Louis-Philippe",
""
],
[
"Salakhutdinov",
"Ruslan",
""
]
] | TITLE: MultiZoo & MultiBench: A Standardized Toolkit for Multimodal Deep
Learning
ABSTRACT: Learning multimodal representations involves integrating information from
multiple heterogeneous sources of data. In order to accelerate progress towards
understudied modalities and tasks while ensuring real-world robustness, we
release MultiZoo, a public toolkit consisting of standardized implementations
of > 20 core multimodal algorithms and MultiBench, a large-scale benchmark
spanning 15 datasets, 10 modalities, 20 prediction tasks, and 6 research areas.
Together, these provide an automated end-to-end machine learning pipeline that
simplifies and standardizes data loading, experimental setup, and model
evaluation. To enable holistic evaluation, we offer a comprehensive methodology
to assess (1) generalization, (2) time and space complexity, and (3) modality
robustness. MultiBench paves the way towards a better understanding of the
capabilities and limitations of multimodal models, while ensuring ease of use,
accessibility, and reproducibility. Our toolkits are publicly available, will
be regularly updated, and welcome inputs from the community.
| no_new_dataset | 0.696771 |
2010.07154 | Liyuan Xu | Liyuan Xu, Yutian Chen, Siddarth Srinivasan, Nando de Freitas, Arnaud
Doucet, Arthur Gretton | Learning Deep Features in Instrumental Variable Regression | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Instrumental variable (IV) regression is a standard strategy for learning
causal relationships between confounded treatment and outcome variables from
observational data by utilizing an instrumental variable, which affects the
outcome only through the treatment. In classical IV regression, learning
proceeds in two stages: stage 1 performs linear regression from the instrument
to the treatment; and stage 2 performs linear regression from the treatment to
the outcome, conditioned on the instrument. We propose a novel method, deep
feature instrumental variable regression (DFIV), to address the case where
relations between instruments, treatments, and outcomes may be nonlinear. In
this case, deep neural nets are trained to define informative nonlinear
features on the instruments and treatments. We propose an alternating training
regime for these features to ensure good end-to-end performance when composing
stages 1 and 2, thus obtaining highly flexible feature maps in a
computationally efficient manner. DFIV outperforms recent state-of-the-art
methods on challenging IV benchmarks, including settings involving high
dimensional image data. DFIV also exhibits competitive performance in
off-policy policy evaluation for reinforcement learning, which can be
understood as an IV regression task.
| [
{
"version": "v1",
"created": "Wed, 14 Oct 2020 15:14:49 GMT"
},
{
"version": "v2",
"created": "Wed, 28 Oct 2020 10:29:20 GMT"
},
{
"version": "v3",
"created": "Sun, 1 Nov 2020 15:36:04 GMT"
},
{
"version": "v4",
"created": "Tue, 27 Jun 2023 10:20:45 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Xu",
"Liyuan",
""
],
[
"Chen",
"Yutian",
""
],
[
"Srinivasan",
"Siddarth",
""
],
[
"de Freitas",
"Nando",
""
],
[
"Doucet",
"Arnaud",
""
],
[
"Gretton",
"Arthur",
""
]
] | TITLE: Learning Deep Features in Instrumental Variable Regression
ABSTRACT: Instrumental variable (IV) regression is a standard strategy for learning
causal relationships between confounded treatment and outcome variables from
observational data by utilizing an instrumental variable, which affects the
outcome only through the treatment. In classical IV regression, learning
proceeds in two stages: stage 1 performs linear regression from the instrument
to the treatment; and stage 2 performs linear regression from the treatment to
the outcome, conditioned on the instrument. We propose a novel method, deep
feature instrumental variable regression (DFIV), to address the case where
relations between instruments, treatments, and outcomes may be nonlinear. In
this case, deep neural nets are trained to define informative nonlinear
features on the instruments and treatments. We propose an alternating training
regime for these features to ensure good end-to-end performance when composing
stages 1 and 2, thus obtaining highly flexible feature maps in a
computationally efficient manner. DFIV outperforms recent state-of-the-art
methods on challenging IV benchmarks, including settings involving high
dimensional image data. DFIV also exhibits competitive performance in
off-policy policy evaluation for reinforcement learning, which can be
understood as an IV regression task.
| no_new_dataset | 0.707815 |
2101.05201 | Ka Man Yim | Ka Man Yim, Jacob Leygonie | Optimisation of Spectral Wavelets for Persistence-based Graph
Classification | null | null | 10.3389/fams.2021.651467 | null | eess.SP cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | A graph's spectral wavelet signature determines a filtration, and
consequently an associated set of extended persistence diagrams. We propose a
framework that optimises the choice of wavelet for a dataset of graphs, such
that their associated persistence diagrams capture features of the graphs that
are best suited to a given data science problem. Since the spectral wavelet
signature of a graph is derived from its Laplacian, our framework encodes
geometric properties of graphs in their associated persistence diagrams and can
be applied to graphs without a priori node attributes. We apply our framework
to graph classification problems and obtain performances competitive with other
persistence-based architectures. To provide the underlying theoretical
foundations, we extend the differentiability result for ordinary persistent
homology to extended persistent homology.
| [
{
"version": "v1",
"created": "Sun, 10 Jan 2021 18:34:27 GMT"
},
{
"version": "v2",
"created": "Mon, 1 Mar 2021 16:43:12 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Yim",
"Ka Man",
""
],
[
"Leygonie",
"Jacob",
""
]
] | TITLE: Optimisation of Spectral Wavelets for Persistence-based Graph
Classification
ABSTRACT: A graph's spectral wavelet signature determines a filtration, and
consequently an associated set of extended persistence diagrams. We propose a
framework that optimises the choice of wavelet for a dataset of graphs, such
that their associated persistence diagrams capture features of the graphs that
are best suited to a given data science problem. Since the spectral wavelet
signature of a graph is derived from its Laplacian, our framework encodes
geometric properties of graphs in their associated persistence diagrams and can
be applied to graphs without a priori node attributes. We apply our framework
to graph classification problems and obtain performances competitive with other
persistence-based architectures. To provide the underlying theoretical
foundations, we extend the differentiability result for ordinary persistent
homology to extended persistent homology.
| no_new_dataset | 0.710641 |
2103.11853 | Markus Reiter-Haas | Markus Reiter-Haas, Simone Kopeinik, Elisabeth Lex | Studying Moral-based Differences in the Framing of Political Tweets | Accepted for publication in ICWSM-2021 - link to published version
will be added | Proceedings of the International AAAI Conference on Web and Social
Media Vol. 15 (2021) 1085-1089 | 10.1609/icwsm.v15i1.18135 | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we study the moral framing of political content on Twitter.
Specifically, we examine differences in moral framing in two datasets: (i)
tweets from US-based politicians annotated with political affiliation and (ii)
COVID-19 related tweets in German from followers of the leaders of the five
major Austrian political parties. Our research is based on recent work that
introduces an unsupervised approach to extract framing bias and intensity in
news using a dictionary of moral virtues and vices. In this paper, we use a
more extensive dictionary and adapt it to German-language tweets. Overall, in
both datasets, we observe a moral framing that is congruent with the public
perception of the political parties. In the US dataset, democrats have a
tendency to frame tweets in terms of care, while loyalty is a characteristic
frame for republicans. In the Austrian dataset, we find that the followers of
the governing conservative party emphasize care, which is a key message and
moral frame in the party's COVID-19 campaign slogan. Our work complements
existing studies on moral framing in social media. Also, our empirical findings
provide novel insights into moral-based framing on COVID-19 in Austria.
| [
{
"version": "v1",
"created": "Mon, 22 Mar 2021 13:48:21 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Reiter-Haas",
"Markus",
""
],
[
"Kopeinik",
"Simone",
""
],
[
"Lex",
"Elisabeth",
""
]
] | TITLE: Studying Moral-based Differences in the Framing of Political Tweets
ABSTRACT: In this paper, we study the moral framing of political content on Twitter.
Specifically, we examine differences in moral framing in two datasets: (i)
tweets from US-based politicians annotated with political affiliation and (ii)
COVID-19 related tweets in German from followers of the leaders of the five
major Austrian political parties. Our research is based on recent work that
introduces an unsupervised approach to extract framing bias and intensity in
news using a dictionary of moral virtues and vices. In this paper, we use a
more extensive dictionary and adapt it to German-language tweets. Overall, in
both datasets, we observe a moral framing that is congruent with the public
perception of the political parties. In the US dataset, democrats have a
tendency to frame tweets in terms of care, while loyalty is a characteristic
frame for republicans. In the Austrian dataset, we find that the followers of
the governing conservative party emphasize care, which is a key message and
moral frame in the party's COVID-19 campaign slogan. Our work complements
existing studies on moral framing in social media. Also, our empirical findings
provide novel insights into moral-based framing on COVID-19 in Austria.
| no_new_dataset | 0.709598 |
2108.02138 | Markus Reiter-Haas | Markus Reiter-Haas, Emilia Parada-Cabaleiro, Markus Schedl, Elham
Motamedi, Marko Tkalcic, and Elisabeth Lex | Predicting Music Relistening Behavior Using the ACT-R Framework | Accepted for publication in RecSys'21 late-breaking results | RecSys '21: Fifteenth ACM Conference on Recommender Systems,
Amsterdam, Netherlands, September 2021 | 10.1145/3460231.3478846 | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Providing suitable recommendations is of vital importance to improve the user
satisfaction of music recommender systems. Here, users often listen to the same
track repeatedly and appreciate recommendations of the same song multiple
times. Thus, accounting for users' relistening behavior is critical for music
recommender systems. In this paper, we describe a psychology-informed approach
to model and predict music relistening behavior that is inspired by studies in
music psychology, which relate music preferences to human memory. We adopt a
well-established psychological theory of human cognition that models the
operations of human memory, i.e., Adaptive Control of Thought-Rational (ACT-R).
In contrast to prior work, which uses only the base-level component of ACT-R,
we utilize five components of ACT-R, i.e., base-level, spreading, partial
matching, valuation, and noise, to investigate the effect of five factors on
music relistening behavior: (i) recency and frequency of prior exposure to
tracks, (ii) co-occurrence of tracks, (iii) the similarity between tracks, (iv)
familiarity with tracks, and (v) randomness in behavior. On a dataset of 1.7
million listening events from Last.fm, we evaluate the performance of our
approach by sequentially predicting the next track(s) in user sessions. We find
that recency and frequency of prior exposure to tracks is an effective
predictor of relistening behavior. Besides, considering the co-occurrence of
tracks and familiarity with tracks further improves performance in terms of
R-precision. We hope that our work inspires future research on the merits of
considering cognitive aspects of memory retrieval to model and predict complex
user behavior.
| [
{
"version": "v1",
"created": "Wed, 4 Aug 2021 16:11:41 GMT"
},
{
"version": "v2",
"created": "Thu, 5 Aug 2021 15:58:53 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Reiter-Haas",
"Markus",
""
],
[
"Parada-Cabaleiro",
"Emilia",
""
],
[
"Schedl",
"Markus",
""
],
[
"Motamedi",
"Elham",
""
],
[
"Tkalcic",
"Marko",
""
],
[
"Lex",
"Elisabeth",
""
]
] | TITLE: Predicting Music Relistening Behavior Using the ACT-R Framework
ABSTRACT: Providing suitable recommendations is of vital importance to improve the user
satisfaction of music recommender systems. Here, users often listen to the same
track repeatedly and appreciate recommendations of the same song multiple
times. Thus, accounting for users' relistening behavior is critical for music
recommender systems. In this paper, we describe a psychology-informed approach
to model and predict music relistening behavior that is inspired by studies in
music psychology, which relate music preferences to human memory. We adopt a
well-established psychological theory of human cognition that models the
operations of human memory, i.e., Adaptive Control of Thought-Rational (ACT-R).
In contrast to prior work, which uses only the base-level component of ACT-R,
we utilize five components of ACT-R, i.e., base-level, spreading, partial
matching, valuation, and noise, to investigate the effect of five factors on
music relistening behavior: (i) recency and frequency of prior exposure to
tracks, (ii) co-occurrence of tracks, (iii) the similarity between tracks, (iv)
familiarity with tracks, and (v) randomness in behavior. On a dataset of 1.7
million listening events from Last.fm, we evaluate the performance of our
approach by sequentially predicting the next track(s) in user sessions. We find
that recency and frequency of prior exposure to tracks is an effective
predictor of relistening behavior. Besides, considering the co-occurrence of
tracks and familiarity with tracks further improves performance in terms of
R-precision. We hope that our work inspires future research on the merits of
considering cognitive aspects of memory retrieval to model and predict complex
user behavior.
| no_new_dataset | 0.710201 |
2109.07446 | Patrick Fernandes | Patrick Fernandes, Kayo Yin, Emmy Liu, Andr\'e F. T. Martins, Graham
Neubig | When Does Translation Require Context? A Data-driven, Multilingual
Exploration | Accepted at ACL2023 | null | null | null | cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Although proper handling of discourse significantly contributes to the
quality of machine translation (MT), these improvements are not adequately
measured in common translation quality metrics. Recent works in context-aware
MT attempt to target a small set of discourse phenomena during evaluation,
however not in a fully systematic way. In this paper, we develop the
Multilingual Discourse-Aware (MuDA) benchmark, a series of taggers that
identify and evaluate model performance on discourse phenomena in any given
dataset. The choice of phenomena is inspired by a novel methodology to
systematically identify translations requiring context. We confirm the
difficulty of previously studied phenomena while uncovering others that were
previously unaddressed. We find that common context-aware MT models make only
marginal improvements over context-agnostic models, which suggests these models
do not handle these ambiguities effectively. We release code and data for 14
language pairs to encourage the MT community to focus on accurately capturing
discourse phenomena.
| [
{
"version": "v1",
"created": "Wed, 15 Sep 2021 17:29:30 GMT"
},
{
"version": "v2",
"created": "Tue, 27 Jun 2023 17:10:50 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Fernandes",
"Patrick",
""
],
[
"Yin",
"Kayo",
""
],
[
"Liu",
"Emmy",
""
],
[
"Martins",
"André F. T.",
""
],
[
"Neubig",
"Graham",
""
]
] | TITLE: When Does Translation Require Context? A Data-driven, Multilingual
Exploration
ABSTRACT: Although proper handling of discourse significantly contributes to the
quality of machine translation (MT), these improvements are not adequately
measured in common translation quality metrics. Recent works in context-aware
MT attempt to target a small set of discourse phenomena during evaluation,
however not in a fully systematic way. In this paper, we develop the
Multilingual Discourse-Aware (MuDA) benchmark, a series of taggers that
identify and evaluate model performance on discourse phenomena in any given
dataset. The choice of phenomena is inspired by a novel methodology to
systematically identify translations requiring context. We confirm the
difficulty of previously studied phenomena while uncovering others that were
previously unaddressed. We find that common context-aware MT models make only
marginal improvements over context-agnostic models, which suggests these models
do not handle these ambiguities effectively. We release code and data for 14
language pairs to encourage the MT community to focus on accurately capturing
discourse phenomena.
| new_dataset | 0.720583 |
2110.12646 | Ta-Chung Chi | Ta-Chung Chi and Alexander I. Rudnicky | Zero-Shot Dialogue Disentanglement by Self-Supervised Entangled Response
Selection | 6 pages, accepted by EMNLP 2021. Update Acknowledgment | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Dialogue disentanglement aims to group utterances in a long and
multi-participant dialogue into threads. This is useful for discourse analysis
and downstream applications such as dialogue response selection, where it can
be the first step to construct a clean context/response set. Unfortunately,
labeling all~\emph{reply-to} links takes quadratic effort w.r.t the number of
utterances: an annotator must check all preceding utterances to identify the
one to which the current utterance is a reply. In this paper, we are the first
to propose a~\textbf{zero-shot} dialogue disentanglement solution. Firstly, we
train a model on a multi-participant response selection dataset harvested from
the web which is not annotated; we then apply the trained model to perform
zero-shot dialogue disentanglement. Without any labeled data, our model can
achieve a cluster F1 score of 25. We also fine-tune the model using various
amounts of labeled data. Experiments show that with only 10\% of the data, we
achieve nearly the same performance of using the full dataset\footnote{Code is
released at
\url{https://github.com/chijames/zero_shot_dialogue_disentanglement}}.
| [
{
"version": "v1",
"created": "Mon, 25 Oct 2021 05:15:01 GMT"
},
{
"version": "v2",
"created": "Mon, 26 Jun 2023 22:44:04 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Chi",
"Ta-Chung",
""
],
[
"Rudnicky",
"Alexander I.",
""
]
] | TITLE: Zero-Shot Dialogue Disentanglement by Self-Supervised Entangled Response
Selection
ABSTRACT: Dialogue disentanglement aims to group utterances in a long and
multi-participant dialogue into threads. This is useful for discourse analysis
and downstream applications such as dialogue response selection, where it can
be the first step to construct a clean context/response set. Unfortunately,
labeling all~\emph{reply-to} links takes quadratic effort w.r.t the number of
utterances: an annotator must check all preceding utterances to identify the
one to which the current utterance is a reply. In this paper, we are the first
to propose a~\textbf{zero-shot} dialogue disentanglement solution. Firstly, we
train a model on a multi-participant response selection dataset harvested from
the web which is not annotated; we then apply the trained model to perform
zero-shot dialogue disentanglement. Without any labeled data, our model can
achieve a cluster F1 score of 25. We also fine-tune the model using various
amounts of labeled data. Experiments show that with only 10\% of the data, we
achieve nearly the same performance of using the full dataset\footnote{Code is
released at
\url{https://github.com/chijames/zero_shot_dialogue_disentanglement}}.
| no_new_dataset | 0.70332 |
2111.08851 | Sebastian Raschka | Xintong Shi, Wenzhi Cao, Sebastian Raschka | Deep Neural Networks for Rank-Consistent Ordinal Regression Based On
Conditional Probabilities | Accepted for publication in Pattern Analysis and Applications | Pattern Analysis and Applications 2023 | 10.1007/s10044-023-01181-9 | null | cs.LG cs.CV stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In recent times, deep neural networks achieved outstanding predictive
performance on various classification and pattern recognition tasks. However,
many real-world prediction problems have ordinal response variables, and this
ordering information is ignored by conventional classification losses such as
the multi-category cross-entropy. Ordinal regression methods for deep neural
networks address this. One such method is the CORAL method, which is based on
an earlier binary label extension framework and achieves rank consistency among
its output layer tasks by imposing a weight-sharing constraint. However, while
earlier experiments showed that CORAL's rank consistency is beneficial for
performance, it is limited by a weight-sharing constraint in a neural network's
fully connected output layer, which may restrict the expressiveness and
capacity of a network trained using CORAL. We propose a new method for
rank-consistent ordinal regression without this limitation. Our rank-consistent
ordinal regression framework (CORN) achieves rank consistency by a novel
training scheme. This training scheme uses conditional training sets to obtain
the unconditional rank probabilities through applying the chain rule for
conditional probability distributions. Experiments on various datasets
demonstrate the efficacy of the proposed method to utilize the ordinal target
information, and the absence of the weight-sharing restriction improves the
performance substantially compared to the CORAL reference approach.
Additionally, the suggested CORN method is not tied to any specific
architecture and can be utilized with any deep neural network classifier to
train it for ordinal regression tasks.
| [
{
"version": "v1",
"created": "Wed, 17 Nov 2021 01:10:23 GMT"
},
{
"version": "v2",
"created": "Fri, 28 Jan 2022 03:42:10 GMT"
},
{
"version": "v3",
"created": "Thu, 7 Jul 2022 18:21:58 GMT"
},
{
"version": "v4",
"created": "Thu, 29 Sep 2022 16:40:54 GMT"
},
{
"version": "v5",
"created": "Thu, 1 Jun 2023 00:40:22 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Shi",
"Xintong",
""
],
[
"Cao",
"Wenzhi",
""
],
[
"Raschka",
"Sebastian",
""
]
] | TITLE: Deep Neural Networks for Rank-Consistent Ordinal Regression Based On
Conditional Probabilities
ABSTRACT: In recent times, deep neural networks achieved outstanding predictive
performance on various classification and pattern recognition tasks. However,
many real-world prediction problems have ordinal response variables, and this
ordering information is ignored by conventional classification losses such as
the multi-category cross-entropy. Ordinal regression methods for deep neural
networks address this. One such method is the CORAL method, which is based on
an earlier binary label extension framework and achieves rank consistency among
its output layer tasks by imposing a weight-sharing constraint. However, while
earlier experiments showed that CORAL's rank consistency is beneficial for
performance, it is limited by a weight-sharing constraint in a neural network's
fully connected output layer, which may restrict the expressiveness and
capacity of a network trained using CORAL. We propose a new method for
rank-consistent ordinal regression without this limitation. Our rank-consistent
ordinal regression framework (CORN) achieves rank consistency by a novel
training scheme. This training scheme uses conditional training sets to obtain
the unconditional rank probabilities through applying the chain rule for
conditional probability distributions. Experiments on various datasets
demonstrate the efficacy of the proposed method to utilize the ordinal target
information, and the absence of the weight-sharing restriction improves the
performance substantially compared to the CORAL reference approach.
Additionally, the suggested CORN method is not tied to any specific
architecture and can be utilized with any deep neural network classifier to
train it for ordinal regression tasks.
| no_new_dataset | 0.710208 |
2201.06268 | Lukas Hedegaard | Lukas Hedegaard and Arian Bakhtiarnia and Alexandros Iosifidis | Continual Transformers: Redundancy-Free Attention for Online Inference | 16 pages, 6 figures, 7 tables | International Conference on Learning Representations, 2023 | null | null | cs.AI cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Transformers in their common form are inherently limited to operate on whole
token sequences rather than on one token at a time. Consequently, their use
during online inference on time-series data entails considerable redundancy due
to the overlap in successive token sequences. In this work, we propose novel
formulations of the Scaled Dot-Product Attention, which enable Transformers to
perform efficient online token-by-token inference on a continual input stream.
Importantly, our modifications are purely to the order of computations, while
the outputs and learned weights are identical to those of the original
Transformer Encoder. We validate our Continual Transformer Encoder with
experiments on the THUMOS14, TVSeries and GTZAN datasets with remarkable
results: Our Continual one- and two-block architectures reduce the floating
point operations per prediction by up to 63x and 2.6x, respectively, while
retaining predictive performance.
| [
{
"version": "v1",
"created": "Mon, 17 Jan 2022 08:20:09 GMT"
},
{
"version": "v2",
"created": "Mon, 7 Nov 2022 07:56:35 GMT"
},
{
"version": "v3",
"created": "Tue, 24 Jan 2023 07:42:08 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Hedegaard",
"Lukas",
""
],
[
"Bakhtiarnia",
"Arian",
""
],
[
"Iosifidis",
"Alexandros",
""
]
] | TITLE: Continual Transformers: Redundancy-Free Attention for Online Inference
ABSTRACT: Transformers in their common form are inherently limited to operate on whole
token sequences rather than on one token at a time. Consequently, their use
during online inference on time-series data entails considerable redundancy due
to the overlap in successive token sequences. In this work, we propose novel
formulations of the Scaled Dot-Product Attention, which enable Transformers to
perform efficient online token-by-token inference on a continual input stream.
Importantly, our modifications are purely to the order of computations, while
the outputs and learned weights are identical to those of the original
Transformer Encoder. We validate our Continual Transformer Encoder with
experiments on the THUMOS14, TVSeries and GTZAN datasets with remarkable
results: Our Continual one- and two-block architectures reduce the floating
point operations per prediction by up to 63x and 2.6x, respectively, while
retaining predictive performance.
| no_new_dataset | 0.71123 |
2201.06843 | Rohitash Chandra | Rohitash Chandra, Yash Vardhan Sharma | Surrogate-assisted distributed swarm optimisation for computationally
expensive geoscientific models | null | Computational Geosciences, 2023 | 10.1007/s10596-023-10223-4 | null | cs.DC cs.AI | http://creativecommons.org/licenses/by/4.0/ | Evolutionary algorithms provide gradient-free optimisation which is
beneficial for models that have difficulty in obtaining gradients; for
instance, geoscientific landscape evolution models. However, such models are at
times computationally expensive and even distributed swarm-based optimisation
with parallel computing struggles. We can incorporate efficient strategies such
as surrogate-assisted optimisation to address the challenges; however,
implementing inter-process communication for surrogate-based model training is
difficult. In this paper, we implement surrogate-based estimation of fitness
evaluation in distributed swarm optimisation over a parallel computing
architecture. We first test the framework on a set of benchmark optimisation
problems and then apply it to a geoscientific model that features a landscape
evolution model. Our results demonstrate very promising results for benchmark
functions and the Badlands landscape evolution model. We obtain a reduction in
computational time while retaining optimisation solution accuracy through the
use of surrogates in a parallel computing environment. The major contribution
of the paper is in the application of surrogate-based optimisation for
geoscientific models which can in the future help in a better understanding of
paleoclimate and geomorphology.
| [
{
"version": "v1",
"created": "Tue, 18 Jan 2022 09:37:14 GMT"
},
{
"version": "v2",
"created": "Sat, 3 Jun 2023 21:45:47 GMT"
},
{
"version": "v3",
"created": "Mon, 26 Jun 2023 18:39:40 GMT"
}
] | 2023-06-28T00:00:00 | [
[
"Chandra",
"Rohitash",
""
],
[
"Sharma",
"Yash Vardhan",
""
]
] | TITLE: Surrogate-assisted distributed swarm optimisation for computationally
expensive geoscientific models
ABSTRACT: Evolutionary algorithms provide gradient-free optimisation which is
beneficial for models that have difficulty in obtaining gradients; for
instance, geoscientific landscape evolution models. However, such models are at
times computationally expensive and even distributed swarm-based optimisation
with parallel computing struggles. We can incorporate efficient strategies such
as surrogate-assisted optimisation to address the challenges; however,
implementing inter-process communication for surrogate-based model training is
difficult. In this paper, we implement surrogate-based estimation of fitness
evaluation in distributed swarm optimisation over a parallel computing
architecture. We first test the framework on a set of benchmark optimisation
problems and then apply it to a geoscientific model that features a landscape
evolution model. Our results demonstrate very promising results for benchmark
functions and the Badlands landscape evolution model. We obtain a reduction in
computational time while retaining optimisation solution accuracy through the
use of surrogates in a parallel computing environment. The major contribution
of the paper is in the application of surrogate-based optimisation for
geoscientific models which can in the future help in a better understanding of
paleoclimate and geomorphology.
| no_new_dataset | 0.709849 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.