id
large_stringlengths 9
16
| submitter
large_stringlengths 3
64
⌀ | authors
large_stringlengths 5
20.2k
| title
large_stringlengths 5
245
| comments
large_stringlengths 1
482
⌀ | journal-ref
large_stringlengths 4
382
⌀ | doi
large_stringlengths 9
151
⌀ | report-no
large_stringlengths 2
509
⌀ | categories
large_stringlengths 5
78
| license
large_stringclasses 9
values | abstract
large_stringlengths 83
3.55k
| versions
large listlengths 1
21
| update_date
timestamp[ms]date 2007-05-23 00:00:00
2025-05-30 00:00:00
| authors_parsed
large listlengths 1
1.35k
| prompt
large_stringlengths 166
3.64k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2212.02031 | Hui Zhang | Hui Zhang, Zuxuan Wu, Zheng Wang, Zhineng Chen, Yu-Gang Jiang | Prototypical Residual Networks for Anomaly Detection and Localization | Accepted by CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Anomaly detection and localization are widely used in industrial
manufacturing for its efficiency and effectiveness. Anomalies are rare and hard
to collect and supervised models easily over-fit to these seen anomalies with a
handful of abnormal samples, producing unsatisfactory performance. On the other
hand, anomalies are typically subtle, hard to discern, and of various
appearance, making it difficult to detect anomalies and let alone locate
anomalous regions. To address these issues, we propose a framework called
Prototypical Residual Network (PRN), which learns feature residuals of varying
scales and sizes between anomalous and normal patterns to accurately
reconstruct the segmentation maps of anomalous regions. PRN mainly consists of
two parts: multi-scale prototypes that explicitly represent the residual
features of anomalies to normal patterns; a multisize self-attention mechanism
that enables variable-sized anomalous feature learning. Besides, we present a
variety of anomaly generation strategies that consider both seen and unseen
appearance variance to enlarge and diversify anomalies. Extensive experiments
on the challenging and widely used MVTec AD benchmark show that PRN outperforms
current state-of-the-art unsupervised and supervised methods. We further report
SOTA results on three additional datasets to demonstrate the effectiveness and
generalizability of PRN.
| [
{
"version": "v1",
"created": "Mon, 5 Dec 2022 05:03:46 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 07:52:42 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zhang",
"Hui",
""
],
[
"Wu",
"Zuxuan",
""
],
[
"Wang",
"Zheng",
""
],
[
"Chen",
"Zhineng",
""
],
[
"Jiang",
"Yu-Gang",
""
]
] | TITLE: Prototypical Residual Networks for Anomaly Detection and Localization
ABSTRACT: Anomaly detection and localization are widely used in industrial
manufacturing for its efficiency and effectiveness. Anomalies are rare and hard
to collect and supervised models easily over-fit to these seen anomalies with a
handful of abnormal samples, producing unsatisfactory performance. On the other
hand, anomalies are typically subtle, hard to discern, and of various
appearance, making it difficult to detect anomalies and let alone locate
anomalous regions. To address these issues, we propose a framework called
Prototypical Residual Network (PRN), which learns feature residuals of varying
scales and sizes between anomalous and normal patterns to accurately
reconstruct the segmentation maps of anomalous regions. PRN mainly consists of
two parts: multi-scale prototypes that explicitly represent the residual
features of anomalies to normal patterns; a multisize self-attention mechanism
that enables variable-sized anomalous feature learning. Besides, we present a
variety of anomaly generation strategies that consider both seen and unseen
appearance variance to enlarge and diversify anomalies. Extensive experiments
on the challenging and widely used MVTec AD benchmark show that PRN outperforms
current state-of-the-art unsupervised and supervised methods. We further report
SOTA results on three additional datasets to demonstrate the effectiveness and
generalizability of PRN.
|
2212.07795 | Lukas Ortmann | Lukas Ortmann, Jean Maeght, Patrick Panciatici, Florian D\"orfler,
Saverio Bolognani | Online Feedback Optimization for Subtransmission Grid Control | null | null | null | null | eess.SY cs.SY | http://creativecommons.org/licenses/by/4.0/ | The increasing electric power consumption and the shift towards renewable
energy resources demand for new ways to operate transmission and
subtransmission grids. Online Feedback Optimization (OFO) is a feedback control
method that enables real-time, constrained, and optimal control of these grids.
Such controllers can minimize, e.g., curtailment and losses while satisfying
grid constraints like voltage and current limits. We tailor and extend the OFO
control method to handle discrete inputs and explain how to design an OFO
controller for the subtransmission grid. We present a novel and publicly
available benchmark which is of the real French subtransmission grid on which
we analyze the proposed controller in terms of robustness against model
mismatch, constraint satisfaction, and tracking performance. Overall, we show
that OFO controllers can help utilize the grid to its full extent, virtually
reinforce it, and operate it optimally and in real-time by using flexibility
offered by renewable generators connected to distribution grids.
| [
{
"version": "v1",
"created": "Thu, 15 Dec 2022 13:12:10 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 16:17:48 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Ortmann",
"Lukas",
""
],
[
"Maeght",
"Jean",
""
],
[
"Panciatici",
"Patrick",
""
],
[
"Dörfler",
"Florian",
""
],
[
"Bolognani",
"Saverio",
""
]
] | TITLE: Online Feedback Optimization for Subtransmission Grid Control
ABSTRACT: The increasing electric power consumption and the shift towards renewable
energy resources demand for new ways to operate transmission and
subtransmission grids. Online Feedback Optimization (OFO) is a feedback control
method that enables real-time, constrained, and optimal control of these grids.
Such controllers can minimize, e.g., curtailment and losses while satisfying
grid constraints like voltage and current limits. We tailor and extend the OFO
control method to handle discrete inputs and explain how to design an OFO
controller for the subtransmission grid. We present a novel and publicly
available benchmark which is of the real French subtransmission grid on which
we analyze the proposed controller in terms of robustness against model
mismatch, constraint satisfaction, and tracking performance. Overall, we show
that OFO controllers can help utilize the grid to its full extent, virtually
reinforce it, and operate it optimally and in real-time by using flexibility
offered by renewable generators connected to distribution grids.
|
2301.03029 | Lifeng Han | Bernadeta Grici\=ut\.e and Lifeng Han and Goran Nenadic | Topic Modelling of Swedish Newspaper Articles about Coronavirus: a Case
Study using Latent Dirichlet Allocation Method | Accepted to International HealthNLP WS @ IEEE-ICHI2023
https://ieeeichi.github.io/ICHI2023/ | null | null | null | cs.CL cs.SI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Topic Modelling (TM) is from the research branches of natural language
understanding (NLU) and natural language processing (NLP) that is to facilitate
insightful analysis from large documents and datasets, such as a summarisation
of main topics and the topic changes. This kind of discovery is getting more
popular in real-life applications due to its impact on big data analytics. In
this study, from the social-media and healthcare domain, we apply popular
Latent Dirichlet Allocation (LDA) methods to model the topic changes in Swedish
newspaper articles about Coronavirus. We describe the corpus we created
including 6515 articles, methods applied, and statistics on topic changes over
approximately 1 year and two months period of time from 17th January 2020 to
13th March 2021. We hope this work can be an asset for grounding applications
of topic modelling and can be inspiring for similar case studies in an era with
pandemics, to support socio-economic impact research as well as clinical and
healthcare analytics. Our data and source code are openly available at
https://github. com/poethan/Swed_Covid_TM Keywords: Latent Dirichlet Allocation
(LDA); Topic Modelling; Coronavirus; Pandemics; Natural Language Understanding;
BERT-topic
| [
{
"version": "v1",
"created": "Sun, 8 Jan 2023 12:33:58 GMT"
},
{
"version": "v2",
"created": "Tue, 17 Jan 2023 22:00:52 GMT"
},
{
"version": "v3",
"created": "Sat, 4 Feb 2023 22:45:20 GMT"
},
{
"version": "v4",
"created": "Fri, 17 Feb 2023 16:47:27 GMT"
},
{
"version": "v5",
"created": "Fri, 10 Mar 2023 11:38:05 GMT"
},
{
"version": "v6",
"created": "Tue, 18 Apr 2023 16:43:11 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Griciūtė",
"Bernadeta",
""
],
[
"Han",
"Lifeng",
""
],
[
"Nenadic",
"Goran",
""
]
] | TITLE: Topic Modelling of Swedish Newspaper Articles about Coronavirus: a Case
Study using Latent Dirichlet Allocation Method
ABSTRACT: Topic Modelling (TM) is from the research branches of natural language
understanding (NLU) and natural language processing (NLP) that is to facilitate
insightful analysis from large documents and datasets, such as a summarisation
of main topics and the topic changes. This kind of discovery is getting more
popular in real-life applications due to its impact on big data analytics. In
this study, from the social-media and healthcare domain, we apply popular
Latent Dirichlet Allocation (LDA) methods to model the topic changes in Swedish
newspaper articles about Coronavirus. We describe the corpus we created
including 6515 articles, methods applied, and statistics on topic changes over
approximately 1 year and two months period of time from 17th January 2020 to
13th March 2021. We hope this work can be an asset for grounding applications
of topic modelling and can be inspiring for similar case studies in an era with
pandemics, to support socio-economic impact research as well as clinical and
healthcare analytics. Our data and source code are openly available at
https://github. com/poethan/Swed_Covid_TM Keywords: Latent Dirichlet Allocation
(LDA); Topic Modelling; Coronavirus; Pandemics; Natural Language Understanding;
BERT-topic
|
2301.07405 | Zongwei Wu | Zongwei Wu, Guillaume Allibert, Fabrice Meriaudeau, Chao Ma, and
C\'edric Demonceaux | HiDAnet: RGB-D Salient Object Detection via Hierarchical Depth Awareness | null | null | 10.1109/TIP.2023.3263111 | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | RGB-D saliency detection aims to fuse multi-modal cues to accurately localize
salient regions. Existing works often adopt attention modules for feature
modeling, with few methods explicitly leveraging fine-grained details to merge
with semantic cues. Thus, despite the auxiliary depth information, it is still
challenging for existing models to distinguish objects with similar appearances
but at distinct camera distances. In this paper, from a new perspective, we
propose a novel Hierarchical Depth Awareness network (HiDAnet) for RGB-D
saliency detection. Our motivation comes from the observation that the
multi-granularity properties of geometric priors correlate well with the neural
network hierarchies. To realize multi-modal and multi-level fusion, we first
use a granularity-based attention scheme to strengthen the discriminatory power
of RGB and depth features separately. Then we introduce a unified cross
dual-attention module for multi-modal and multi-level fusion in a
coarse-to-fine manner. The encoded multi-modal features are gradually
aggregated into a shared decoder. Further, we exploit a multi-scale loss to
take full advantage of the hierarchical information. Extensive experiments on
challenging benchmark datasets demonstrate that our HiDAnet performs favorably
over the state-of-the-art methods by large margins.
| [
{
"version": "v1",
"created": "Wed, 18 Jan 2023 10:00:59 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Wu",
"Zongwei",
""
],
[
"Allibert",
"Guillaume",
""
],
[
"Meriaudeau",
"Fabrice",
""
],
[
"Ma",
"Chao",
""
],
[
"Demonceaux",
"Cédric",
""
]
] | TITLE: HiDAnet: RGB-D Salient Object Detection via Hierarchical Depth Awareness
ABSTRACT: RGB-D saliency detection aims to fuse multi-modal cues to accurately localize
salient regions. Existing works often adopt attention modules for feature
modeling, with few methods explicitly leveraging fine-grained details to merge
with semantic cues. Thus, despite the auxiliary depth information, it is still
challenging for existing models to distinguish objects with similar appearances
but at distinct camera distances. In this paper, from a new perspective, we
propose a novel Hierarchical Depth Awareness network (HiDAnet) for RGB-D
saliency detection. Our motivation comes from the observation that the
multi-granularity properties of geometric priors correlate well with the neural
network hierarchies. To realize multi-modal and multi-level fusion, we first
use a granularity-based attention scheme to strengthen the discriminatory power
of RGB and depth features separately. Then we introduce a unified cross
dual-attention module for multi-modal and multi-level fusion in a
coarse-to-fine manner. The encoded multi-modal features are gradually
aggregated into a shared decoder. Further, we exploit a multi-scale loss to
take full advantage of the hierarchical information. Extensive experiments on
challenging benchmark datasets demonstrate that our HiDAnet performs favorably
over the state-of-the-art methods by large margins.
|
2301.11025 | Anatole Moureaux | Anatole Moureaux and Simon De Wergifosse and Chlo\'e Chopin and Jimmy
Weber and Flavio Abreu Araujo | Neuromorphic spintronics accelerated by an unconventional data-driven
Thiele equation approach | 11 pages, 8 figures | null | null | null | cs.ET eess.SP physics.app-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We design a neural network based on a single spin-torque vortex
nano-oscillator (STVO) multiplexed in time. The behavior of the STVO is
simulated with an improved ultra-fast and quantitative model based on the
Thiele equation approach. Different mathematical and numerical adaptations are
brought to the model in order to increase the accuracy and the speed of the
simulations. We demonstrate the high added value and adaptability of such a
neural network through the resolution of three standard machine learning tasks
in the framework of reservoir computing. The first one is a task of waveform
(sines and squares) classification. We show the ability of the system to
effectively classify waveforms with high accuracy and low root-mean-square
error thanks to the intrinsic short-term memory of the device. Given the high
throughput of the simulations, two innovative parametric studies on the
intensity of the input signal and the level of noise in the system are
performed to demonstrate the value of our new models. The efficiency of our
system is then tested during a speech recognition task on the TI-46 dataset and
shows the agreement between the new models and the corresponding experimental
measurements. Finally, we use our STVO-based neural network to perform image
recognition on the MNIST dataset. State-of-the-art performances are
demonstrated, and the interest of using the STVO dynamics as an activation
function is highlighted. These results support and facilitate the future
development of neuromorphic STVO-based hardware for energy-efficient machine
learning.
| [
{
"version": "v1",
"created": "Thu, 26 Jan 2023 10:30:31 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 15:18:36 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Moureaux",
"Anatole",
""
],
[
"De Wergifosse",
"Simon",
""
],
[
"Chopin",
"Chloé",
""
],
[
"Weber",
"Jimmy",
""
],
[
"Araujo",
"Flavio Abreu",
""
]
] | TITLE: Neuromorphic spintronics accelerated by an unconventional data-driven
Thiele equation approach
ABSTRACT: We design a neural network based on a single spin-torque vortex
nano-oscillator (STVO) multiplexed in time. The behavior of the STVO is
simulated with an improved ultra-fast and quantitative model based on the
Thiele equation approach. Different mathematical and numerical adaptations are
brought to the model in order to increase the accuracy and the speed of the
simulations. We demonstrate the high added value and adaptability of such a
neural network through the resolution of three standard machine learning tasks
in the framework of reservoir computing. The first one is a task of waveform
(sines and squares) classification. We show the ability of the system to
effectively classify waveforms with high accuracy and low root-mean-square
error thanks to the intrinsic short-term memory of the device. Given the high
throughput of the simulations, two innovative parametric studies on the
intensity of the input signal and the level of noise in the system are
performed to demonstrate the value of our new models. The efficiency of our
system is then tested during a speech recognition task on the TI-46 dataset and
shows the agreement between the new models and the corresponding experimental
measurements. Finally, we use our STVO-based neural network to perform image
recognition on the MNIST dataset. State-of-the-art performances are
demonstrated, and the interest of using the STVO dynamics as an activation
function is highlighted. These results support and facilitate the future
development of neuromorphic STVO-based hardware for energy-efficient machine
learning.
|
2302.02997 | Yftah Ziser | Shun Shao, Yftah Ziser and Shay Cohen | Erasure of Unaligned Attributes from Neural Representations | Accepted to Transactions of the Association for Computational
Linguistics, 22 pages (pre-MIT Press publication version) | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | We present the Assignment-Maximization Spectral Attribute removaL (AMSAL)
algorithm, which erases information from neural representations when the
information to be erased is implicit rather than directly being aligned to each
input example. Our algorithm works by alternating between two steps. In one, it
finds an assignment of the input representations to the information to be
erased, and in the other, it creates projections of both the input
representations and the information to be erased into a joint latent space. We
test our algorithm on an extensive array of datasets, including a Twitter
dataset with multiple guarded attributes, the BiasBios dataset and the
BiasBench benchmark. The last benchmark includes four datasets with various
types of protected attributes. Our results demonstrate that bias can often be
removed in our setup. We also discuss the limitations of our approach when
there is a strong entanglement between the main task and the information to be
erased.
| [
{
"version": "v1",
"created": "Mon, 6 Feb 2023 18:32:17 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 10:34:01 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Shao",
"Shun",
""
],
[
"Ziser",
"Yftah",
""
],
[
"Cohen",
"Shay",
""
]
] | TITLE: Erasure of Unaligned Attributes from Neural Representations
ABSTRACT: We present the Assignment-Maximization Spectral Attribute removaL (AMSAL)
algorithm, which erases information from neural representations when the
information to be erased is implicit rather than directly being aligned to each
input example. Our algorithm works by alternating between two steps. In one, it
finds an assignment of the input representations to the information to be
erased, and in the other, it creates projections of both the input
representations and the information to be erased into a joint latent space. We
test our algorithm on an extensive array of datasets, including a Twitter
dataset with multiple guarded attributes, the BiasBios dataset and the
BiasBench benchmark. The last benchmark includes four datasets with various
types of protected attributes. Our results demonstrate that bias can often be
removed in our setup. We also discuss the limitations of our approach when
there is a strong entanglement between the main task and the information to be
erased.
|
2302.10835 | Keith Mills Mr. | Fred X. Han, Keith G. Mills, Fabian Chudak, Parsa Riahi, Mohammad
Salameh, Jialin Zhang, Wei Lu, Shangling Jui, Di Niu | A General-Purpose Transferable Predictor for Neural Architecture Search | Accepted to SDM2023; version includes supplementary material; 12
Pages, 3 Figures, 6 Tables | null | 10.1137/1.9781611977653.ch81 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Understanding and modelling the performance of neural architectures is key to
Neural Architecture Search (NAS). Performance predictors have seen widespread
use in low-cost NAS and achieve high ranking correlations between predicted and
ground truth performance in several NAS benchmarks. However, existing
predictors are often designed based on network encodings specific to a
predefined search space and are therefore not generalizable to other search
spaces or new architecture families. In this paper, we propose a
general-purpose neural predictor for NAS that can transfer across search
spaces, by representing any given candidate Convolutional Neural Network (CNN)
with a Computation Graph (CG) that consists of primitive operators. We further
combine our CG network representation with Contrastive Learning (CL) and
propose a graph representation learning procedure that leverages the structural
information of unlabeled architectures from multiple families to train CG
embeddings for our performance predictor. Experimental results on
NAS-Bench-101, 201 and 301 demonstrate the efficacy of our scheme as we achieve
strong positive Spearman Rank Correlation Coefficient (SRCC) on every search
space, outperforming several Zero-Cost Proxies, including Synflow and Jacov,
which are also generalizable predictors across search spaces. Moreover, when
using our proposed general-purpose predictor in an evolutionary neural
architecture search algorithm, we can find high-performance architectures on
NAS-Bench-101 and find a MobileNetV3 architecture that attains 79.2% top-1
accuracy on ImageNet.
| [
{
"version": "v1",
"created": "Tue, 21 Feb 2023 17:28:05 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Han",
"Fred X.",
""
],
[
"Mills",
"Keith G.",
""
],
[
"Chudak",
"Fabian",
""
],
[
"Riahi",
"Parsa",
""
],
[
"Salameh",
"Mohammad",
""
],
[
"Zhang",
"Jialin",
""
],
[
"Lu",
"Wei",
""
],
[
"Jui",
"Shangling",
""
],
[
"Niu",
"Di",
""
]
] | TITLE: A General-Purpose Transferable Predictor for Neural Architecture Search
ABSTRACT: Understanding and modelling the performance of neural architectures is key to
Neural Architecture Search (NAS). Performance predictors have seen widespread
use in low-cost NAS and achieve high ranking correlations between predicted and
ground truth performance in several NAS benchmarks. However, existing
predictors are often designed based on network encodings specific to a
predefined search space and are therefore not generalizable to other search
spaces or new architecture families. In this paper, we propose a
general-purpose neural predictor for NAS that can transfer across search
spaces, by representing any given candidate Convolutional Neural Network (CNN)
with a Computation Graph (CG) that consists of primitive operators. We further
combine our CG network representation with Contrastive Learning (CL) and
propose a graph representation learning procedure that leverages the structural
information of unlabeled architectures from multiple families to train CG
embeddings for our performance predictor. Experimental results on
NAS-Bench-101, 201 and 301 demonstrate the efficacy of our scheme as we achieve
strong positive Spearman Rank Correlation Coefficient (SRCC) on every search
space, outperforming several Zero-Cost Proxies, including Synflow and Jacov,
which are also generalizable predictors across search spaces. Moreover, when
using our proposed general-purpose predictor in an evolutionary neural
architecture search algorithm, we can find high-performance architectures on
NAS-Bench-101 and find a MobileNetV3 architecture that attains 79.2% top-1
accuracy on ImageNet.
|
2303.04720 | Davide Moroni | A. Bruno, C.Caudai, G.R. Leone, M. Martinelli, D. Moroni, F. Crotti | Medical Waste Sorting: a computer vision approach for assisted primary
sorting | Accepted for inclusion in IWCIM 2023 @ ICASSP2023, June 5, 2023,
Rhodes, Greece | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Medical waste, i.e. waste produced during medical activities in hospitals,
clinics and laboratories, represents hazardous waste whose management involves
special care and high costs. However, this kind of waste contains a significant
fraction of highly valued materials that can enter a circular economy process.
To this end, in this paper, we propose a computer vision approach for assisting
in the primary sorting of medical waste. The feasibility of our approach is
demonstrated on a representative dataset we collected and made available to the
community, with which we have trained a model that achieves 100\% accuracy, and
a new dataset on which the trained model exhibits good generalization.
| [
{
"version": "v1",
"created": "Wed, 8 Mar 2023 17:07:02 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 12:29:54 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Bruno",
"A.",
""
],
[
"Caudai",
"C.",
""
],
[
"Leone",
"G. R.",
""
],
[
"Martinelli",
"M.",
""
],
[
"Moroni",
"D.",
""
],
[
"Crotti",
"F.",
""
]
] | TITLE: Medical Waste Sorting: a computer vision approach for assisted primary
sorting
ABSTRACT: Medical waste, i.e. waste produced during medical activities in hospitals,
clinics and laboratories, represents hazardous waste whose management involves
special care and high costs. However, this kind of waste contains a significant
fraction of highly valued materials that can enter a circular economy process.
To this end, in this paper, we propose a computer vision approach for assisting
in the primary sorting of medical waste. The feasibility of our approach is
demonstrated on a representative dataset we collected and made available to the
community, with which we have trained a model that achieves 100\% accuracy, and
a new dataset on which the trained model exhibits good generalization.
|
2303.07840 | Jun Wan | Jun Wan, Jun Liu, Jie Zhou, Zhihui Lai, Linlin Shen, Hang Sun, Ping
Xiong, Wenwen Min | Precise Facial Landmark Detection by Reference Heatmap Transformer | Accepted by IEEE Transactions on Image Processing, March 2023 | null | 10.1109/TIP.2023.3261749 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most facial landmark detection methods predict landmarks by mapping the input
facial appearance features to landmark heatmaps and have achieved promising
results. However, when the face image is suffering from large poses, heavy
occlusions and complicated illuminations, they cannot learn discriminative
feature representations and effective facial shape constraints, nor can they
accurately predict the value of each element in the landmark heatmap, limiting
their detection accuracy. To address this problem, we propose a novel Reference
Heatmap Transformer (RHT) by introducing reference heatmap information for more
precise facial landmark detection. The proposed RHT consists of a Soft
Transformation Module (STM) and a Hard Transformation Module (HTM), which can
cooperate with each other to encourage the accurate transformation of the
reference heatmap information and facial shape constraints. Then, a Multi-Scale
Feature Fusion Module (MSFFM) is proposed to fuse the transformed heatmap
features and the semantic features learned from the original face images to
enhance feature representations for producing more accurate target heatmaps. To
the best of our knowledge, this is the first study to explore how to enhance
facial landmark detection by transforming the reference heatmap information.
The experimental results from challenging benchmark datasets demonstrate that
our proposed method outperforms the state-of-the-art methods in the literature.
| [
{
"version": "v1",
"created": "Tue, 14 Mar 2023 12:26:48 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Wan",
"Jun",
""
],
[
"Liu",
"Jun",
""
],
[
"Zhou",
"Jie",
""
],
[
"Lai",
"Zhihui",
""
],
[
"Shen",
"Linlin",
""
],
[
"Sun",
"Hang",
""
],
[
"Xiong",
"Ping",
""
],
[
"Min",
"Wenwen",
""
]
] | TITLE: Precise Facial Landmark Detection by Reference Heatmap Transformer
ABSTRACT: Most facial landmark detection methods predict landmarks by mapping the input
facial appearance features to landmark heatmaps and have achieved promising
results. However, when the face image is suffering from large poses, heavy
occlusions and complicated illuminations, they cannot learn discriminative
feature representations and effective facial shape constraints, nor can they
accurately predict the value of each element in the landmark heatmap, limiting
their detection accuracy. To address this problem, we propose a novel Reference
Heatmap Transformer (RHT) by introducing reference heatmap information for more
precise facial landmark detection. The proposed RHT consists of a Soft
Transformation Module (STM) and a Hard Transformation Module (HTM), which can
cooperate with each other to encourage the accurate transformation of the
reference heatmap information and facial shape constraints. Then, a Multi-Scale
Feature Fusion Module (MSFFM) is proposed to fuse the transformed heatmap
features and the semantic features learned from the original face images to
enhance feature representations for producing more accurate target heatmaps. To
the best of our knowledge, this is the first study to explore how to enhance
facial landmark detection by transforming the reference heatmap information.
The experimental results from challenging benchmark datasets demonstrate that
our proposed method outperforms the state-of-the-art methods in the literature.
|
2303.08818 | Daeseok Lee | Daeseok Lee, Jeunghyun Byun and Bonggun Shin | Boosting Convolutional Neural Networks' Protein Binding Site Prediction
Capacity Using SE(3)-invariant transformers, Transfer Learning and
Homology-based Augmentation | Updates in version 2: author order change (making it clear that
Bonggun Shin is the corresponding author) | null | null | null | q-bio.QM cs.LG cs.NE q-bio.BM | http://creativecommons.org/licenses/by/4.0/ | Figuring out small molecule binding sites in target proteins, in the
resolution of either pocket or residue, is critical in many virtual and real
drug-discovery scenarios. Since it is not always easy to find such binding
sites based on domain knowledge or traditional methods, different deep learning
methods that predict binding sites out of protein structures have been
developed in recent years. Here we present a new such deep learning algorithm,
that significantly outperformed all state-of-the-art baselines in terms of the
both resolutions$\unicode{x2013}$pocket and residue. This good performance was
also demonstrated in a case study involving the protein human serum albumin and
its binding sites. Our algorithm included new ideas both in the model
architecture and in the training method. For the model architecture, it
incorporated SE(3)-invariant geometric self-attention layers that operate on
top of residue-level CNN outputs. This residue-level processing of the model
allowed a transfer learning between the two resolutions, which turned out to
significantly improve the binding pocket prediction. Moreover, we developed
novel augmentation method based on protein homology, which prevented our model
from over-fitting. Overall, we believe that our contribution to the literature
is twofold. First, we provided a new computational method for binding site
prediction that is relevant to real-world applications, as shown by the good
performance on different benchmarks and case study. Second, the novel ideas in
our method$\unicode{x2013}$the model architecture, transfer learning and the
homology augmentation$\unicode{x2013}$would serve as useful components in
future works.
| [
{
"version": "v1",
"created": "Mon, 20 Feb 2023 05:02:40 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 05:05:00 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Lee",
"Daeseok",
""
],
[
"Byun",
"Jeunghyun",
""
],
[
"Shin",
"Bonggun",
""
]
] | TITLE: Boosting Convolutional Neural Networks' Protein Binding Site Prediction
Capacity Using SE(3)-invariant transformers, Transfer Learning and
Homology-based Augmentation
ABSTRACT: Figuring out small molecule binding sites in target proteins, in the
resolution of either pocket or residue, is critical in many virtual and real
drug-discovery scenarios. Since it is not always easy to find such binding
sites based on domain knowledge or traditional methods, different deep learning
methods that predict binding sites out of protein structures have been
developed in recent years. Here we present a new such deep learning algorithm,
that significantly outperformed all state-of-the-art baselines in terms of the
both resolutions$\unicode{x2013}$pocket and residue. This good performance was
also demonstrated in a case study involving the protein human serum albumin and
its binding sites. Our algorithm included new ideas both in the model
architecture and in the training method. For the model architecture, it
incorporated SE(3)-invariant geometric self-attention layers that operate on
top of residue-level CNN outputs. This residue-level processing of the model
allowed a transfer learning between the two resolutions, which turned out to
significantly improve the binding pocket prediction. Moreover, we developed
novel augmentation method based on protein homology, which prevented our model
from over-fitting. Overall, we believe that our contribution to the literature
is twofold. First, we provided a new computational method for binding site
prediction that is relevant to real-world applications, as shown by the good
performance on different benchmarks and case study. Second, the novel ideas in
our method$\unicode{x2013}$the model architecture, transfer learning and the
homology augmentation$\unicode{x2013}$would serve as useful components in
future works.
|
2303.14175 | Ye Zhu | Ye Zhu, Jie Yang, Si-Qi Liu and Ruimao Zhang | Inherent Consistent Learning for Accurate Semi-supervised Medical Image
Segmentation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Semi-supervised medical image segmentation has attracted much attention in
recent years because of the high cost of medical image annotations. In this
paper, we propose a novel Inherent Consistent Learning (ICL) method, aims to
learn robust semantic category representations through the semantic consistency
guidance of labeled and unlabeled data to help segmentation. In practice, we
introduce two external modules, namely Supervised Semantic Proxy Adaptor (SSPA)
and Unsupervised Semantic Consistent Learner (USCL) that is based on the
attention mechanism to align the semantic category representations of labeled
and unlabeled data, as well as update the global semantic representations over
the entire training set. The proposed ICL is a plug-and-play scheme for various
network architectures, and the two modules are not involved in the testing
stage. Experimental results on three public benchmarks show that the proposed
method can outperform the state-of-the-art, especially when the number of
annotated data is extremely limited. Code is available at:
https://github.com/zhuye98/ICL.git.
| [
{
"version": "v1",
"created": "Fri, 24 Mar 2023 17:38:03 GMT"
},
{
"version": "v2",
"created": "Tue, 28 Mar 2023 10:52:58 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Apr 2023 18:15:49 GMT"
},
{
"version": "v4",
"created": "Tue, 18 Apr 2023 09:50:03 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zhu",
"Ye",
""
],
[
"Yang",
"Jie",
""
],
[
"Liu",
"Si-Qi",
""
],
[
"Zhang",
"Ruimao",
""
]
] | TITLE: Inherent Consistent Learning for Accurate Semi-supervised Medical Image
Segmentation
ABSTRACT: Semi-supervised medical image segmentation has attracted much attention in
recent years because of the high cost of medical image annotations. In this
paper, we propose a novel Inherent Consistent Learning (ICL) method, aims to
learn robust semantic category representations through the semantic consistency
guidance of labeled and unlabeled data to help segmentation. In practice, we
introduce two external modules, namely Supervised Semantic Proxy Adaptor (SSPA)
and Unsupervised Semantic Consistent Learner (USCL) that is based on the
attention mechanism to align the semantic category representations of labeled
and unlabeled data, as well as update the global semantic representations over
the entire training set. The proposed ICL is a plug-and-play scheme for various
network architectures, and the two modules are not involved in the testing
stage. Experimental results on three public benchmarks show that the proposed
method can outperform the state-of-the-art, especially when the number of
annotated data is extremely limited. Code is available at:
https://github.com/zhuye98/ICL.git.
|
2303.14357 | Xiaoxiao He | Xiaoxiao He, Chaowei Tan, Bo Liu, Liping Si, Weiwu Yao, Liang Zhao, Di
Liu, Qilong Zhangli, Qi Chang, Kang Li and Dimitris N. Metaxas | Dealing With Heterogeneous 3D MR Knee Images: A Federated Few-Shot
Learning Method With Dual Knowledge Distillation | null | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Federated Learning has gained popularity among medical institutions since it
enables collaborative training between clients (e.g., hospitals) without
aggregating data. However, due to the high cost associated with creating
annotations, especially for large 3D image datasets, clinical institutions do
not have enough supervised data for training locally. Thus, the performance of
the collaborative model is subpar under limited supervision. On the other hand,
large institutions have the resources to compile data repositories with
high-resolution images and labels. Therefore, individual clients can utilize
the knowledge acquired in the public data repositories to mitigate the shortage
of private annotated images. In this paper, we propose a federated few-shot
learning method with dual knowledge distillation. This method allows joint
training with limited annotations across clients without jeopardizing privacy.
The supervised learning of the proposed method extracts features from limited
labeled data in each client, while the unsupervised data is used to distill
both feature and response-based knowledge from a national data repository to
further improve the accuracy of the collaborative model and reduce the
communication cost. Extensive evaluations are conducted on 3D magnetic
resonance knee images from a private clinical dataset. Our proposed method
shows superior performance and less training time than other semi-supervised
federated learning methods. Codes and additional visualization results are
available at https://github.com/hexiaoxiao-cs/fedml-knee.
| [
{
"version": "v1",
"created": "Sat, 25 Mar 2023 04:46:25 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 02:37:56 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"He",
"Xiaoxiao",
""
],
[
"Tan",
"Chaowei",
""
],
[
"Liu",
"Bo",
""
],
[
"Si",
"Liping",
""
],
[
"Yao",
"Weiwu",
""
],
[
"Zhao",
"Liang",
""
],
[
"Liu",
"Di",
""
],
[
"Zhangli",
"Qilong",
""
],
[
"Chang",
"Qi",
""
],
[
"Li",
"Kang",
""
],
[
"Metaxas",
"Dimitris N.",
""
]
] | TITLE: Dealing With Heterogeneous 3D MR Knee Images: A Federated Few-Shot
Learning Method With Dual Knowledge Distillation
ABSTRACT: Federated Learning has gained popularity among medical institutions since it
enables collaborative training between clients (e.g., hospitals) without
aggregating data. However, due to the high cost associated with creating
annotations, especially for large 3D image datasets, clinical institutions do
not have enough supervised data for training locally. Thus, the performance of
the collaborative model is subpar under limited supervision. On the other hand,
large institutions have the resources to compile data repositories with
high-resolution images and labels. Therefore, individual clients can utilize
the knowledge acquired in the public data repositories to mitigate the shortage
of private annotated images. In this paper, we propose a federated few-shot
learning method with dual knowledge distillation. This method allows joint
training with limited annotations across clients without jeopardizing privacy.
The supervised learning of the proposed method extracts features from limited
labeled data in each client, while the unsupervised data is used to distill
both feature and response-based knowledge from a national data repository to
further improve the accuracy of the collaborative model and reduce the
communication cost. Extensive evaluations are conducted on 3D magnetic
resonance knee images from a private clinical dataset. Our proposed method
shows superior performance and less training time than other semi-supervised
federated learning methods. Codes and additional visualization results are
available at https://github.com/hexiaoxiao-cs/fedml-knee.
|
2303.16117 | Thomas Wong | Thomas Wong, Mauricio Barahona | Feature Engineering Methods on Multivariate Time-Series Data for
Financial Data Science Competitions | arXiv admin note: substantial text overlap with arXiv:2303.07925 | null | null | null | q-fin.ST cs.LG | http://creativecommons.org/licenses/by/4.0/ | This paper is a work in progress. We are looking for collaborators to provide
us financial datasets in Equity/Futures market to conduct more bench-marking
studies. The authors have papers employing similar methods applied on the
Numerai dataset, which is freely available but obfuscated.
We apply different feature engineering methods for time-series to US market
price data. The predictive power of models are tested against Numerai-Signals
targets.
| [
{
"version": "v1",
"created": "Sun, 26 Mar 2023 00:57:35 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 17:27:16 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Wong",
"Thomas",
""
],
[
"Barahona",
"Mauricio",
""
]
] | TITLE: Feature Engineering Methods on Multivariate Time-Series Data for
Financial Data Science Competitions
ABSTRACT: This paper is a work in progress. We are looking for collaborators to provide
us financial datasets in Equity/Futures market to conduct more bench-marking
studies. The authors have papers employing similar methods applied on the
Numerai dataset, which is freely available but obfuscated.
We apply different feature engineering methods for time-series to US market
price data. The predictive power of models are tested against Numerai-Signals
targets.
|
2303.16727 | Limin Wang | Limin Wang, Bingkun Huang, Zhiyu Zhao, Zhan Tong, Yinan He, Yi Wang,
Yali Wang, Yu Qiao | VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking | CVPR 2023 camera-ready version | null | null | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Scale is the primary factor for building a powerful foundation model that
could well generalize to a variety of downstream tasks. However, it is still
challenging to train video foundation models with billions of parameters. This
paper shows that video masked autoencoder (VideoMAE) is a scalable and general
self-supervised pre-trainer for building video foundation models. We scale the
VideoMAE in both model and data with a core design. Specifically, we present a
dual masking strategy for efficient pre-training, with an encoder operating on
a subset of video tokens and a decoder processing another subset of video
tokens. Although VideoMAE is very efficient due to high masking ratio in
encoder, masking decoder can still further reduce the overall computational
cost. This enables the efficient pre-training of billion-level models in video.
We also use a progressive training paradigm that involves an initial
pre-training on a diverse multi-sourced unlabeled dataset, followed by a
post-pre-training on a mixed labeled dataset. Finally, we successfully train a
video ViT model with a billion parameters, which achieves a new
state-of-the-art performance on the datasets of Kinetics (90.0% on K400 and
89.9% on K600) and Something-Something (68.7% on V1 and 77.0% on V2). In
addition, we extensively verify the pre-trained video ViT models on a variety
of downstream tasks, demonstrating its effectiveness as a general video
representation learner. The code and model is available at
\url{https://github.com/OpenGVLab/VideoMAEv2}.
| [
{
"version": "v1",
"created": "Wed, 29 Mar 2023 14:28:41 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 11:46:41 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Wang",
"Limin",
""
],
[
"Huang",
"Bingkun",
""
],
[
"Zhao",
"Zhiyu",
""
],
[
"Tong",
"Zhan",
""
],
[
"He",
"Yinan",
""
],
[
"Wang",
"Yi",
""
],
[
"Wang",
"Yali",
""
],
[
"Qiao",
"Yu",
""
]
] | TITLE: VideoMAE V2: Scaling Video Masked Autoencoders with Dual Masking
ABSTRACT: Scale is the primary factor for building a powerful foundation model that
could well generalize to a variety of downstream tasks. However, it is still
challenging to train video foundation models with billions of parameters. This
paper shows that video masked autoencoder (VideoMAE) is a scalable and general
self-supervised pre-trainer for building video foundation models. We scale the
VideoMAE in both model and data with a core design. Specifically, we present a
dual masking strategy for efficient pre-training, with an encoder operating on
a subset of video tokens and a decoder processing another subset of video
tokens. Although VideoMAE is very efficient due to high masking ratio in
encoder, masking decoder can still further reduce the overall computational
cost. This enables the efficient pre-training of billion-level models in video.
We also use a progressive training paradigm that involves an initial
pre-training on a diverse multi-sourced unlabeled dataset, followed by a
post-pre-training on a mixed labeled dataset. Finally, we successfully train a
video ViT model with a billion parameters, which achieves a new
state-of-the-art performance on the datasets of Kinetics (90.0% on K400 and
89.9% on K600) and Something-Something (68.7% on V1 and 77.0% on V2). In
addition, we extensively verify the pre-trained video ViT models on a variety
of downstream tasks, demonstrating its effectiveness as a general video
representation learner. The code and model is available at
\url{https://github.com/OpenGVLab/VideoMAEv2}.
|
2303.18164 | Dr. Suryansh Kumar | Ce Liu, Suryansh Kumar, Shuhang Gu, Radu Timofte, Luc Van Gool | Single Image Depth Prediction Made Better: A Multivariate Gaussian Take | Accepted to IEEE/CVF CVPR 2023. Draft info: 17 pages, 13 Figures, 9
Tables | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Neural-network-based single image depth prediction (SIDP) is a challenging
task where the goal is to predict the scene's per-pixel depth at test time.
Since the problem, by definition, is ill-posed, the fundamental goal is to come
up with an approach that can reliably model the scene depth from a set of
training examples. In the pursuit of perfect depth estimation, most existing
state-of-the-art learning techniques predict a single scalar depth value
per-pixel. Yet, it is well-known that the trained model has accuracy limits and
can predict imprecise depth. Therefore, an SIDP approach must be mindful of the
expected depth variations in the model's prediction at test time. Accordingly,
we introduce an approach that performs continuous modeling of per-pixel depth,
where we can predict and reason about the per-pixel depth and its distribution.
To this end, we model per-pixel scene depth using a multivariate Gaussian
distribution. Moreover, contrary to the existing uncertainty modeling methods
-- in the same spirit, where per-pixel depth is assumed to be independent, we
introduce per-pixel covariance modeling that encodes its depth dependency w.r.t
all the scene points. Unfortunately, per-pixel depth covariance modeling leads
to a computationally expensive continuous loss function, which we solve
efficiently using the learned low-rank approximation of the overall covariance
matrix. Notably, when tested on benchmark datasets such as KITTI, NYU, and
SUN-RGB-D, the SIDP model obtained by optimizing our loss function shows
state-of-the-art results. Our method's accuracy (named MG) is among the top on
the KITTI depth-prediction benchmark leaderboard.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 16:01:03 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 08:52:18 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Liu",
"Ce",
""
],
[
"Kumar",
"Suryansh",
""
],
[
"Gu",
"Shuhang",
""
],
[
"Timofte",
"Radu",
""
],
[
"Van Gool",
"Luc",
""
]
] | TITLE: Single Image Depth Prediction Made Better: A Multivariate Gaussian Take
ABSTRACT: Neural-network-based single image depth prediction (SIDP) is a challenging
task where the goal is to predict the scene's per-pixel depth at test time.
Since the problem, by definition, is ill-posed, the fundamental goal is to come
up with an approach that can reliably model the scene depth from a set of
training examples. In the pursuit of perfect depth estimation, most existing
state-of-the-art learning techniques predict a single scalar depth value
per-pixel. Yet, it is well-known that the trained model has accuracy limits and
can predict imprecise depth. Therefore, an SIDP approach must be mindful of the
expected depth variations in the model's prediction at test time. Accordingly,
we introduce an approach that performs continuous modeling of per-pixel depth,
where we can predict and reason about the per-pixel depth and its distribution.
To this end, we model per-pixel scene depth using a multivariate Gaussian
distribution. Moreover, contrary to the existing uncertainty modeling methods
-- in the same spirit, where per-pixel depth is assumed to be independent, we
introduce per-pixel covariance modeling that encodes its depth dependency w.r.t
all the scene points. Unfortunately, per-pixel depth covariance modeling leads
to a computationally expensive continuous loss function, which we solve
efficiently using the learned low-rank approximation of the overall covariance
matrix. Notably, when tested on benchmark datasets such as KITTI, NYU, and
SUN-RGB-D, the SIDP model obtained by optimizing our loss function shows
state-of-the-art results. Our method's accuracy (named MG) is among the top on
the KITTI depth-prediction benchmark leaderboard.
|
2304.03262 | Jiuhai Chen | Jiuhai Chen, Lichang Chen, Heng Huang, Tianyi Zhou | When do you need Chain-of-Thought Prompting for ChatGPT? | null | null | null | null | cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Chain-of-Thought (CoT) prompting can effectively elicit complex multi-step
reasoning from Large Language Models~(LLMs). For example, by simply adding CoT
instruction ``Let's think step-by-step'' to each input query of MultiArith
dataset, GPT-3's accuracy can be improved from 17.7\% to 78.7\%. However, it is
not clear whether CoT is still effective on more recent instruction finetuned
(IFT) LLMs such as ChatGPT. Surprisingly, on ChatGPT, CoT is no longer
effective for certain tasks such as arithmetic reasoning while still keeping
effective on other reasoning tasks. Moreover, on the former tasks, ChatGPT
usually achieves the best performance and can generate CoT even without being
instructed to do so. Hence, it is plausible that ChatGPT has already been
trained on these tasks with CoT and thus memorized the instruction so it
implicitly follows such an instruction when applied to the same queries, even
without CoT. Our analysis reflects a potential risk of overfitting/bias toward
instructions introduced in IFT, which becomes more common in training LLMs. In
addition, it indicates possible leakage of the pretraining recipe, e.g., one
can verify whether a dataset and instruction were used in training ChatGPT. Our
experiments report new baseline results of ChatGPT on a variety of reasoning
tasks and shed novel insights into LLM's profiling, instruction memorization,
and pretraining dataset leakage.
| [
{
"version": "v1",
"created": "Thu, 6 Apr 2023 17:47:29 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 14:45:18 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Chen",
"Jiuhai",
""
],
[
"Chen",
"Lichang",
""
],
[
"Huang",
"Heng",
""
],
[
"Zhou",
"Tianyi",
""
]
] | TITLE: When do you need Chain-of-Thought Prompting for ChatGPT?
ABSTRACT: Chain-of-Thought (CoT) prompting can effectively elicit complex multi-step
reasoning from Large Language Models~(LLMs). For example, by simply adding CoT
instruction ``Let's think step-by-step'' to each input query of MultiArith
dataset, GPT-3's accuracy can be improved from 17.7\% to 78.7\%. However, it is
not clear whether CoT is still effective on more recent instruction finetuned
(IFT) LLMs such as ChatGPT. Surprisingly, on ChatGPT, CoT is no longer
effective for certain tasks such as arithmetic reasoning while still keeping
effective on other reasoning tasks. Moreover, on the former tasks, ChatGPT
usually achieves the best performance and can generate CoT even without being
instructed to do so. Hence, it is plausible that ChatGPT has already been
trained on these tasks with CoT and thus memorized the instruction so it
implicitly follows such an instruction when applied to the same queries, even
without CoT. Our analysis reflects a potential risk of overfitting/bias toward
instructions introduced in IFT, which becomes more common in training LLMs. In
addition, it indicates possible leakage of the pretraining recipe, e.g., one
can verify whether a dataset and instruction were used in training ChatGPT. Our
experiments report new baseline results of ChatGPT on a variety of reasoning
tasks and shed novel insights into LLM's profiling, instruction memorization,
and pretraining dataset leakage.
|
2304.04133 | Federico Semeraro | Federico Semeraro, Yi Zhang, Wenying Wu, Patrick Carroll | NeRF applied to satellite imagery for surface reconstruction | null | null | null | null | cs.CV cs.AI cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present Surf-NeRF, a modified implementation of the recently introduced
Shadow Neural Radiance Field (S-NeRF) model. This method is able to synthesize
novel views from a sparse set of satellite images of a scene, while accounting
for the variation in lighting present in the pictures. The trained model can
also be used to accurately estimate the surface elevation of the scene, which
is often a desirable quantity for satellite observation applications. S-NeRF
improves on the standard Neural Radiance Field (NeRF) method by considering the
radiance as a function of the albedo and the irradiance. Both these quantities
are output by fully connected neural network branches of the model, and the
latter is considered as a function of the direct light from the sun and the
diffuse color from the sky. The implementations were run on a dataset of
satellite images, augmented using a zoom-and-crop technique. A hyperparameter
study for NeRF was carried out, leading to intriguing observations on the
model's convergence. Finally, both NeRF and S-NeRF were run until 100k epochs
in order to fully fit the data and produce their best possible predictions. The
code related to this article can be found at
https://github.com/fsemerar/surfnerf.
| [
{
"version": "v1",
"created": "Sun, 9 Apr 2023 01:37:13 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Apr 2023 16:38:35 GMT"
},
{
"version": "v3",
"created": "Wed, 12 Apr 2023 01:11:10 GMT"
},
{
"version": "v4",
"created": "Tue, 18 Apr 2023 15:32:23 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Semeraro",
"Federico",
""
],
[
"Zhang",
"Yi",
""
],
[
"Wu",
"Wenying",
""
],
[
"Carroll",
"Patrick",
""
]
] | TITLE: NeRF applied to satellite imagery for surface reconstruction
ABSTRACT: We present Surf-NeRF, a modified implementation of the recently introduced
Shadow Neural Radiance Field (S-NeRF) model. This method is able to synthesize
novel views from a sparse set of satellite images of a scene, while accounting
for the variation in lighting present in the pictures. The trained model can
also be used to accurately estimate the surface elevation of the scene, which
is often a desirable quantity for satellite observation applications. S-NeRF
improves on the standard Neural Radiance Field (NeRF) method by considering the
radiance as a function of the albedo and the irradiance. Both these quantities
are output by fully connected neural network branches of the model, and the
latter is considered as a function of the direct light from the sun and the
diffuse color from the sky. The implementations were run on a dataset of
satellite images, augmented using a zoom-and-crop technique. A hyperparameter
study for NeRF was carried out, leading to intriguing observations on the
model's convergence. Finally, both NeRF and S-NeRF were run until 100k epochs
in order to fully fit the data and produce their best possible predictions. The
code related to this article can be found at
https://github.com/fsemerar/surfnerf.
|
2304.06907 | Mohammad Mehdi Hosseini | Seyed Mahdi Roostaiyan, Mohammad Mehdi Hosseini, Mahya Mohammadi
Kashani, S. Hamid Amiri | Toward Real-Time Image Annotation Using Marginalized Coupled Dictionary
Learning | @article{roostaiyan2022toward, title={Toward real-time image
annotation using marginalized coupled dictionary learning},
author={Roostaiyan, Seyed Mahdi and Hosseini, Mohammad Mehdi and Kashani,
Mahya Mohammadi and Amiri, S Hamid}, journal={Journal of Real-Time Image
Processing}, volume={19}, number={3}, pages={623--638}, year={2022},
publisher={Springer} } | Journal of Real-Time Image Processing. 2022 Jun;19(3):623-38 | 10.1007/s11554-022-01210-6 | null | cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | In most image retrieval systems, images include various high-level semantics,
called tags or annotations. Virtually all the state-of-the-art image annotation
methods that handle imbalanced labeling are search-based techniques which are
time-consuming. In this paper, a novel coupled dictionary learning approach is
proposed to learn a limited number of visual prototypes and their corresponding
semantics simultaneously. This approach leads to a real-time image annotation
procedure. Another contribution of this paper is that utilizes a marginalized
loss function instead of the squared loss function that is inappropriate for
image annotation with imbalanced labels. We have employed a marginalized loss
function in our method to leverage a simple and effective method of prototype
updating. Meanwhile, we have introduced ${\ell}_1$ regularization on semantic
prototypes to preserve the sparse and imbalanced nature of labels in learned
semantic prototypes. Finally, comprehensive experimental results on various
datasets demonstrate the efficiency of the proposed method for image annotation
tasks in terms of accuracy and time. The reference implementation is publicly
available on https://github.com/hamid-amiri/MCDL-Image-Annotation.
| [
{
"version": "v1",
"created": "Fri, 14 Apr 2023 03:01:44 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Apr 2023 20:00:44 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Roostaiyan",
"Seyed Mahdi",
""
],
[
"Hosseini",
"Mohammad Mehdi",
""
],
[
"Kashani",
"Mahya Mohammadi",
""
],
[
"Amiri",
"S. Hamid",
""
]
] | TITLE: Toward Real-Time Image Annotation Using Marginalized Coupled Dictionary
Learning
ABSTRACT: In most image retrieval systems, images include various high-level semantics,
called tags or annotations. Virtually all the state-of-the-art image annotation
methods that handle imbalanced labeling are search-based techniques which are
time-consuming. In this paper, a novel coupled dictionary learning approach is
proposed to learn a limited number of visual prototypes and their corresponding
semantics simultaneously. This approach leads to a real-time image annotation
procedure. Another contribution of this paper is that utilizes a marginalized
loss function instead of the squared loss function that is inappropriate for
image annotation with imbalanced labels. We have employed a marginalized loss
function in our method to leverage a simple and effective method of prototype
updating. Meanwhile, we have introduced ${\ell}_1$ regularization on semantic
prototypes to preserve the sparse and imbalanced nature of labels in learned
semantic prototypes. Finally, comprehensive experimental results on various
datasets demonstrate the efficiency of the proposed method for image annotation
tasks in terms of accuracy and time. The reference implementation is publicly
available on https://github.com/hamid-amiri/MCDL-Image-Annotation.
|
2304.07874 | Huan Liu | Yangyi Liu, Huan Liu, Liangyan Li, Zijun Wu and Jun Chen | A Data-Centric Solution to NonHomogeneous Dehazing via Vision
Transformer | Accepted by CVPRW 2023 | null | null | null | cs.CV cs.AI cs.GR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent years have witnessed an increased interest in image dehazing. Many
deep learning methods have been proposed to tackle this challenge, and have
made significant accomplishments dealing with homogeneous haze. However, these
solutions cannot maintain comparable performance when they are applied to
images with non-homogeneous haze, e.g., NH-HAZE23 dataset introduced by NTIRE
challenges. One of the reasons for such failures is that non-homogeneous haze
does not obey one of the assumptions that is required for modeling homogeneous
haze. In addition, a large number of pairs of non-homogeneous hazy image and
the clean counterpart is required using traditional end-to-end training
approaches, while NH-HAZE23 dataset is of limited quantities. Although it is
possible to augment the NH-HAZE23 dataset by leveraging other non-homogeneous
dehazing datasets, we observe that it is necessary to design a proper
data-preprocessing approach that reduces the distribution gaps between the
target dataset and the augmented one. This finding indeed aligns with the
essence of data-centric AI. With a novel network architecture and a principled
data-preprocessing approach that systematically enhances data quality, we
present an innovative dehazing method. Specifically, we apply RGB-channel-wise
transformations on the augmented datasets, and incorporate the state-of-the-art
transformers as the backbone in the two-branch framework. We conduct extensive
experiments and ablation study to demonstrate the effectiveness of our proposed
method.
| [
{
"version": "v1",
"created": "Sun, 16 Apr 2023 19:44:25 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 17:54:43 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Liu",
"Yangyi",
""
],
[
"Liu",
"Huan",
""
],
[
"Li",
"Liangyan",
""
],
[
"Wu",
"Zijun",
""
],
[
"Chen",
"Jun",
""
]
] | TITLE: A Data-Centric Solution to NonHomogeneous Dehazing via Vision
Transformer
ABSTRACT: Recent years have witnessed an increased interest in image dehazing. Many
deep learning methods have been proposed to tackle this challenge, and have
made significant accomplishments dealing with homogeneous haze. However, these
solutions cannot maintain comparable performance when they are applied to
images with non-homogeneous haze, e.g., NH-HAZE23 dataset introduced by NTIRE
challenges. One of the reasons for such failures is that non-homogeneous haze
does not obey one of the assumptions that is required for modeling homogeneous
haze. In addition, a large number of pairs of non-homogeneous hazy image and
the clean counterpart is required using traditional end-to-end training
approaches, while NH-HAZE23 dataset is of limited quantities. Although it is
possible to augment the NH-HAZE23 dataset by leveraging other non-homogeneous
dehazing datasets, we observe that it is necessary to design a proper
data-preprocessing approach that reduces the distribution gaps between the
target dataset and the augmented one. This finding indeed aligns with the
essence of data-centric AI. With a novel network architecture and a principled
data-preprocessing approach that systematically enhances data quality, we
present an innovative dehazing method. Specifically, we apply RGB-channel-wise
transformations on the augmented datasets, and incorporate the state-of-the-art
transformers as the backbone in the two-branch framework. We conduct extensive
experiments and ablation study to demonstrate the effectiveness of our proposed
method.
|
2304.08109 | Baochang Ma | Xianghui Sun, Yunjie Ji, Baochang Ma, Xiangang Li | A Comparative Study between Full-Parameter and LoRA-based Fine-Tuning on
Chinese Instruction Data for Instruction Following Large Language Model | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, the instruction-tuning of large language models is a crucial area
of research in the field of natural language processing. Due to resource and
cost limitations, several researchers have employed parameter-efficient tuning
techniques, such as LoRA, for instruction tuning, and have obtained encouraging
results In comparison to full-parameter fine-tuning, LoRA-based tuning
demonstrates salient benefits in terms of training costs. In this study, we
undertook experimental comparisons between full-parameter fine-tuning and
LoRA-based tuning methods, utilizing LLaMA as the base model. The experimental
results show that the selection of the foundational model, training dataset
scale, learnable parameter quantity, and model training cost are all important
factors. We hope that the experimental conclusions of this paper can provide
inspiration for training large language models, especially in the field of
Chinese, and help researchers find a better trade-off strategy between training
cost and model performance. To facilitate the reproduction of the paper's
results, the dataset, model and code will be released.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 09:36:36 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Apr 2023 03:08:18 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Sun",
"Xianghui",
""
],
[
"Ji",
"Yunjie",
""
],
[
"Ma",
"Baochang",
""
],
[
"Li",
"Xiangang",
""
]
] | TITLE: A Comparative Study between Full-Parameter and LoRA-based Fine-Tuning on
Chinese Instruction Data for Instruction Following Large Language Model
ABSTRACT: Recently, the instruction-tuning of large language models is a crucial area
of research in the field of natural language processing. Due to resource and
cost limitations, several researchers have employed parameter-efficient tuning
techniques, such as LoRA, for instruction tuning, and have obtained encouraging
results In comparison to full-parameter fine-tuning, LoRA-based tuning
demonstrates salient benefits in terms of training costs. In this study, we
undertook experimental comparisons between full-parameter fine-tuning and
LoRA-based tuning methods, utilizing LLaMA as the base model. The experimental
results show that the selection of the foundational model, training dataset
scale, learnable parameter quantity, and model training cost are all important
factors. We hope that the experimental conclusions of this paper can provide
inspiration for training large language models, especially in the field of
Chinese, and help researchers find a better trade-off strategy between training
cost and model performance. To facilitate the reproduction of the paper's
results, the dataset, model and code will be released.
|
2304.08498 | Hamid Tizhoosh | Pooria Mazaheri, Azam Asilian Bidgoli, Shahryar Rahnamayan, H.R.
Tizhoosh | Ranking Loss and Sequestering Learning for Reducing Image Search Bias in
Histopathology | Under Review for publication | null | null | null | eess.IV cs.AI cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Recently, deep learning has started to play an essential role in healthcare
applications, including image search in digital pathology. Despite the recent
progress in computer vision, significant issues remain for image searching in
histopathology archives. A well-known problem is AI bias and lack of
generalization. A more particular shortcoming of deep models is the ignorance
toward search functionality. The former affects every model, the latter only
search and matching. Due to the lack of ranking-based learning, researchers
must train models based on the classification error and then use the resultant
embedding for image search purposes. Moreover, deep models appear to be prone
to internal bias even if using a large image repository of various hospitals.
This paper proposes two novel ideas to improve image search performance. First,
we use a ranking loss function to guide feature extraction toward the
matching-oriented nature of the search. By forcing the model to learn the
ranking of matched outputs, the representation learning is customized toward
image search instead of learning a class label. Second, we introduce the
concept of sequestering learning to enhance the generalization of feature
extraction. By excluding the images of the input hospital from the matched
outputs, i.e., sequestering the input domain, the institutional bias is
reduced. The proposed ideas are implemented and validated through the largest
public dataset of whole slide images. The experiments demonstrate superior
results compare to the-state-of-art.
| [
{
"version": "v1",
"created": "Sat, 15 Apr 2023 03:38:09 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Mazaheri",
"Pooria",
""
],
[
"Bidgoli",
"Azam Asilian",
""
],
[
"Rahnamayan",
"Shahryar",
""
],
[
"Tizhoosh",
"H. R.",
""
]
] | TITLE: Ranking Loss and Sequestering Learning for Reducing Image Search Bias in
Histopathology
ABSTRACT: Recently, deep learning has started to play an essential role in healthcare
applications, including image search in digital pathology. Despite the recent
progress in computer vision, significant issues remain for image searching in
histopathology archives. A well-known problem is AI bias and lack of
generalization. A more particular shortcoming of deep models is the ignorance
toward search functionality. The former affects every model, the latter only
search and matching. Due to the lack of ranking-based learning, researchers
must train models based on the classification error and then use the resultant
embedding for image search purposes. Moreover, deep models appear to be prone
to internal bias even if using a large image repository of various hospitals.
This paper proposes two novel ideas to improve image search performance. First,
we use a ranking loss function to guide feature extraction toward the
matching-oriented nature of the search. By forcing the model to learn the
ranking of matched outputs, the representation learning is customized toward
image search instead of learning a class label. Second, we introduce the
concept of sequestering learning to enhance the generalization of feature
extraction. By excluding the images of the input hospital from the matched
outputs, i.e., sequestering the input domain, the institutional bias is
reduced. The proposed ideas are implemented and validated through the largest
public dataset of whole slide images. The experiments demonstrate superior
results compare to the-state-of-art.
|
2304.08499 | Shreyan Mitra | Shreyan Mitra and Leilani Gilpin | The XAISuite framework and the implications of explanatory system
dissonance | 41 pages, 23 figures | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Explanatory systems make machine learning models more transparent. However,
they are often inconsistent. In order to quantify and isolate possible
scenarios leading to this discrepancy, this paper compares two explanatory
systems, SHAP and LIME, based on the correlation of their respective importance
scores using 14 machine learning models (7 regression and 7 classification) and
4 tabular datasets (2 regression and 2 classification). We make two novel
findings. Firstly, the magnitude of importance is not significant in
explanation consistency. The correlations between SHAP and LIME importance
scores for the most important features may or may not be more variable than the
correlation between SHAP and LIME importance scores averaged across all
features. Secondly, the similarity between SHAP and LIME importance scores
cannot predict model accuracy. In the process of our research, we construct an
open-source library, XAISuite, that unifies the process of training and
explaining models. Finally, this paper contributes a generalized framework to
better explain machine learning models and optimize their performance.
| [
{
"version": "v1",
"created": "Sat, 15 Apr 2023 04:40:03 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Mitra",
"Shreyan",
""
],
[
"Gilpin",
"Leilani",
""
]
] | TITLE: The XAISuite framework and the implications of explanatory system
dissonance
ABSTRACT: Explanatory systems make machine learning models more transparent. However,
they are often inconsistent. In order to quantify and isolate possible
scenarios leading to this discrepancy, this paper compares two explanatory
systems, SHAP and LIME, based on the correlation of their respective importance
scores using 14 machine learning models (7 regression and 7 classification) and
4 tabular datasets (2 regression and 2 classification). We make two novel
findings. Firstly, the magnitude of importance is not significant in
explanation consistency. The correlations between SHAP and LIME importance
scores for the most important features may or may not be more variable than the
correlation between SHAP and LIME importance scores averaged across all
features. Secondly, the similarity between SHAP and LIME importance scores
cannot predict model accuracy. In the process of our research, we construct an
open-source library, XAISuite, that unifies the process of training and
explaining models. Finally, this paper contributes a generalized framework to
better explain machine learning models and optimize their performance.
|
2304.08577 | Artsiom Sanakoyeu | Yuming Du, Robin Kips, Albert Pumarola, Sebastian Starke, Ali Thabet,
Artsiom Sanakoyeu | Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking
Inputs with Diffusion Model | CVPR 2023, project page: https://dulucas.github.io/agrol/ | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the recent surge in popularity of AR/VR applications, realistic and
accurate control of 3D full-body avatars has become a highly demanded feature.
A particular challenge is that only a sparse tracking signal is available from
standalone HMDs (Head Mounted Devices), often limited to tracking the user's
head and wrists. While this signal is resourceful for reconstructing the upper
body motion, the lower body is not tracked and must be synthesized from the
limited information provided by the upper body joints. In this paper, we
present AGRoL, a novel conditional diffusion model specifically designed to
track full bodies given sparse upper-body tracking signals. Our model is based
on a simple multi-layer perceptron (MLP) architecture and a novel conditioning
scheme for motion data. It can predict accurate and smooth full-body motion,
particularly the challenging lower body movement. Unlike common diffusion
architectures, our compact architecture can run in real-time, making it
suitable for online body-tracking applications. We train and evaluate our model
on AMASS motion capture dataset, and demonstrate that our approach outperforms
state-of-the-art methods in generated motion accuracy and smoothness. We
further justify our design choices through extensive experiments and ablation
studies.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 19:35:13 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Du",
"Yuming",
""
],
[
"Kips",
"Robin",
""
],
[
"Pumarola",
"Albert",
""
],
[
"Starke",
"Sebastian",
""
],
[
"Thabet",
"Ali",
""
],
[
"Sanakoyeu",
"Artsiom",
""
]
] | TITLE: Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking
Inputs with Diffusion Model
ABSTRACT: With the recent surge in popularity of AR/VR applications, realistic and
accurate control of 3D full-body avatars has become a highly demanded feature.
A particular challenge is that only a sparse tracking signal is available from
standalone HMDs (Head Mounted Devices), often limited to tracking the user's
head and wrists. While this signal is resourceful for reconstructing the upper
body motion, the lower body is not tracked and must be synthesized from the
limited information provided by the upper body joints. In this paper, we
present AGRoL, a novel conditional diffusion model specifically designed to
track full bodies given sparse upper-body tracking signals. Our model is based
on a simple multi-layer perceptron (MLP) architecture and a novel conditioning
scheme for motion data. It can predict accurate and smooth full-body motion,
particularly the challenging lower body movement. Unlike common diffusion
architectures, our compact architecture can run in real-time, making it
suitable for online body-tracking applications. We train and evaluate our model
on AMASS motion capture dataset, and demonstrate that our approach outperforms
state-of-the-art methods in generated motion accuracy and smoothness. We
further justify our design choices through extensive experiments and ablation
studies.
|
2304.08580 | Pooya Fayyazsanavi | Pooya Fayyazsanavi, Zhiqiang Wan, Will Hutchcroft, Ivaylo Boyadzhiev,
Yuguang Li, Jana Kosecka, Sing Bing Kang | U2RLE: Uncertainty-Guided 2-Stage Room Layout Estimation | To be Appear on CVPR 2023 | null | null | null | cs.CV cs.RO | http://creativecommons.org/licenses/by/4.0/ | While the existing deep learning-based room layout estimation techniques
demonstrate good overall accuracy, they are less effective for distant
floor-wall boundary. To tackle this problem, we propose a novel
uncertainty-guided approach for layout boundary estimation introducing new
two-stage CNN architecture termed U2RLE. The initial stage predicts both
floor-wall boundary and its uncertainty and is followed by the refinement of
boundaries with high positional uncertainty using a different, distance-aware
loss. Finally, outputs from the two stages are merged to produce the room
layout. Experiments using ZInD and Structure3D datasets show that U2RLE
improves over current state-of-the-art, being able to handle both near and far
walls better. In particular, U2RLE outperforms current state-of-the-art
techniques for the most distant walls.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 19:43:08 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Fayyazsanavi",
"Pooya",
""
],
[
"Wan",
"Zhiqiang",
""
],
[
"Hutchcroft",
"Will",
""
],
[
"Boyadzhiev",
"Ivaylo",
""
],
[
"Li",
"Yuguang",
""
],
[
"Kosecka",
"Jana",
""
],
[
"Kang",
"Sing Bing",
""
]
] | TITLE: U2RLE: Uncertainty-Guided 2-Stage Room Layout Estimation
ABSTRACT: While the existing deep learning-based room layout estimation techniques
demonstrate good overall accuracy, they are less effective for distant
floor-wall boundary. To tackle this problem, we propose a novel
uncertainty-guided approach for layout boundary estimation introducing new
two-stage CNN architecture termed U2RLE. The initial stage predicts both
floor-wall boundary and its uncertainty and is followed by the refinement of
boundaries with high positional uncertainty using a different, distance-aware
loss. Finally, outputs from the two stages are merged to produce the room
layout. Experiments using ZInD and Structure3D datasets show that U2RLE
improves over current state-of-the-art, being able to handle both near and far
walls better. In particular, U2RLE outperforms current state-of-the-art
techniques for the most distant walls.
|
2304.08582 | Swapnil Mane | Swapnil Mane, Vaibhav Khatavkar | Researchers eye-view of sarcasm detection in social media textual
content | 8 pages | International Journal of Creative Research Thoughts (IJCRT),
ISSN:2320-2882, Volume.9, Issue 8, pp.150-157, 2021 | 10.1729/Journal.31277 | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The enormous use of sarcastic text in all forms of communication in social
media will have a physiological effect on target users. Each user has a
different approach to misusing and recognising sarcasm. Sarcasm detection is
difficult even for users, and this will depend on many things such as
perspective, context, special symbols. So, that will be a challenging task for
machines to differentiate sarcastic sentences from non-sarcastic sentences.
There are no exact rules based on which model will accurately detect sarcasm
from many text corpus in the current situation. So, one needs to focus on
optimistic and forthcoming approaches in the sarcasm detection domain. This
paper discusses various sarcasm detection techniques and concludes with some
approaches, related datasets with optimal features, and the researcher's
challenges.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 19:45:10 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Mane",
"Swapnil",
""
],
[
"Khatavkar",
"Vaibhav",
""
]
] | TITLE: Researchers eye-view of sarcasm detection in social media textual
content
ABSTRACT: The enormous use of sarcastic text in all forms of communication in social
media will have a physiological effect on target users. Each user has a
different approach to misusing and recognising sarcasm. Sarcasm detection is
difficult even for users, and this will depend on many things such as
perspective, context, special symbols. So, that will be a challenging task for
machines to differentiate sarcastic sentences from non-sarcastic sentences.
There are no exact rules based on which model will accurately detect sarcasm
from many text corpus in the current situation. So, one needs to focus on
optimistic and forthcoming approaches in the sarcasm detection domain. This
paper discusses various sarcasm detection techniques and concludes with some
approaches, related datasets with optimal features, and the researcher's
challenges.
|
2304.08592 | Sunghyun Park | Sunghyun Park, Sunghyo Chung, Jungsoo Lee, Jaegul Choo | Improving Scene Text Recognition for Character-Level Long-Tailed
Distribution | 17 pages | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Despite the recent remarkable improvements in scene text recognition (STR),
the majority of the studies focused mainly on the English language, which only
includes few number of characters. However, STR models show a large performance
degradation on languages with a numerous number of characters (e.g., Chinese
and Korean), especially on characters that rarely appear due to the long-tailed
distribution of characters in such languages. To address such an issue, we
conducted an empirical analysis using synthetic datasets with different
character-level distributions (e.g., balanced and long-tailed distributions).
While increasing a substantial number of tail classes without considering the
context helps the model to correctly recognize characters individually,
training with such a synthetic dataset interferes the model with learning the
contextual information (i.e., relation among characters), which is also
important for predicting the whole word. Based on this motivation, we propose a
novel Context-Aware and Free Experts Network (CAFE-Net) using two experts: 1)
context-aware expert learns the contextual representation trained with a
long-tailed dataset composed of common words used in everyday life and 2)
context-free expert focuses on correctly predicting individual characters by
utilizing a dataset with a balanced number of characters. By training two
experts to focus on learning contextual and visual representations,
respectively, we propose a novel confidence ensemble method to compensate the
limitation of each expert. Through the experiments, we demonstrate that
CAFE-Net improves the STR performance on languages containing numerous number
of characters. Moreover, we show that CAFE-Net is easily applicable to various
STR models.
| [
{
"version": "v1",
"created": "Fri, 31 Mar 2023 06:11:33 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Park",
"Sunghyun",
""
],
[
"Chung",
"Sunghyo",
""
],
[
"Lee",
"Jungsoo",
""
],
[
"Choo",
"Jaegul",
""
]
] | TITLE: Improving Scene Text Recognition for Character-Level Long-Tailed
Distribution
ABSTRACT: Despite the recent remarkable improvements in scene text recognition (STR),
the majority of the studies focused mainly on the English language, which only
includes few number of characters. However, STR models show a large performance
degradation on languages with a numerous number of characters (e.g., Chinese
and Korean), especially on characters that rarely appear due to the long-tailed
distribution of characters in such languages. To address such an issue, we
conducted an empirical analysis using synthetic datasets with different
character-level distributions (e.g., balanced and long-tailed distributions).
While increasing a substantial number of tail classes without considering the
context helps the model to correctly recognize characters individually,
training with such a synthetic dataset interferes the model with learning the
contextual information (i.e., relation among characters), which is also
important for predicting the whole word. Based on this motivation, we propose a
novel Context-Aware and Free Experts Network (CAFE-Net) using two experts: 1)
context-aware expert learns the contextual representation trained with a
long-tailed dataset composed of common words used in everyday life and 2)
context-free expert focuses on correctly predicting individual characters by
utilizing a dataset with a balanced number of characters. By training two
experts to focus on learning contextual and visual representations,
respectively, we propose a novel confidence ensemble method to compensate the
limitation of each expert. Through the experiments, we demonstrate that
CAFE-Net improves the STR performance on languages containing numerous number
of characters. Moreover, we show that CAFE-Net is easily applicable to various
STR models.
|
2304.08593 | Harry Rubin-Falcone | Harry Rubin-Falcone and Joyce Lee and Jenna Wiens | Forecasting with Sparse but Informative Variables: A Case Study in
Predicting Blood Glucose | 10 pages, 9 figures, 5 tables, accepted to AAAI23 | AAAI 2023 | null | null | cs.LG | http://creativecommons.org/licenses/by/4.0/ | In time-series forecasting, future target values may be affected by both
intrinsic and extrinsic effects. When forecasting blood glucose, for example,
intrinsic effects can be inferred from the history of the target signal alone
(\textit{i.e.} blood glucose), but accurately modeling the impact of extrinsic
effects requires auxiliary signals, like the amount of carbohydrates ingested.
Standard forecasting techniques often assume that extrinsic and intrinsic
effects vary at similar rates. However, when auxiliary signals are generated at
a much lower frequency than the target variable (e.g., blood glucose
measurements are made every 5 minutes, while meals occur once every few hours),
even well-known extrinsic effects (e.g., carbohydrates increase blood glucose)
may prove difficult to learn. To better utilize these \textit{sparse but
informative variables} (SIVs), we introduce a novel encoder/decoder forecasting
approach that accurately learns the per-timepoint effect of the SIV, by (i)
isolating it from intrinsic effects and (ii) restricting its learned effect
based on domain knowledge. On a simulated dataset pertaining to the task of
blood glucose forecasting, when the SIV is accurately recorded our approach
outperforms baseline approaches in terms of rMSE (13.07 [95% CI: 11.77,14.16]
vs. 14.14 [12.69,15.27]). In the presence of a corrupted SIV, the proposed
approach can still result in lower error compared to the baseline but the
advantage is reduced as noise increases. By isolating their effects and
incorporating domain knowledge, our approach makes it possible to better
utilize SIVs in forecasting.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 20:16:09 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Rubin-Falcone",
"Harry",
""
],
[
"Lee",
"Joyce",
""
],
[
"Wiens",
"Jenna",
""
]
] | TITLE: Forecasting with Sparse but Informative Variables: A Case Study in
Predicting Blood Glucose
ABSTRACT: In time-series forecasting, future target values may be affected by both
intrinsic and extrinsic effects. When forecasting blood glucose, for example,
intrinsic effects can be inferred from the history of the target signal alone
(\textit{i.e.} blood glucose), but accurately modeling the impact of extrinsic
effects requires auxiliary signals, like the amount of carbohydrates ingested.
Standard forecasting techniques often assume that extrinsic and intrinsic
effects vary at similar rates. However, when auxiliary signals are generated at
a much lower frequency than the target variable (e.g., blood glucose
measurements are made every 5 minutes, while meals occur once every few hours),
even well-known extrinsic effects (e.g., carbohydrates increase blood glucose)
may prove difficult to learn. To better utilize these \textit{sparse but
informative variables} (SIVs), we introduce a novel encoder/decoder forecasting
approach that accurately learns the per-timepoint effect of the SIV, by (i)
isolating it from intrinsic effects and (ii) restricting its learned effect
based on domain knowledge. On a simulated dataset pertaining to the task of
blood glucose forecasting, when the SIV is accurately recorded our approach
outperforms baseline approaches in terms of rMSE (13.07 [95% CI: 11.77,14.16]
vs. 14.14 [12.69,15.27]). In the presence of a corrupted SIV, the proposed
approach can still result in lower error compared to the baseline but the
advantage is reduced as noise increases. By isolating their effects and
incorporating domain knowledge, our approach makes it possible to better
utilize SIVs in forecasting.
|
2304.08594 | Marina Neseem | Marina Neseem, Ahmed Agiza, Sherief Reda | AdaMTL: Adaptive Input-dependent Inference for Efficient Multi-Task
Learning | This paper will appear in the proceedings of CVPR 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern Augmented reality applications require performing multiple tasks on
each input frame simultaneously. Multi-task learning (MTL) represents an
effective approach where multiple tasks share an encoder to extract
representative features from the input frame, followed by task-specific
decoders to generate predictions for each task. Generally, the shared encoder
in MTL models needs to have a large representational capacity in order to
generalize well to various tasks and input data, which has a negative effect on
the inference latency. In this paper, we argue that due to the large variations
in the complexity of the input frames, some computations might be unnecessary
for the output. Therefore, we introduce AdaMTL, an adaptive framework that
learns task-aware inference policies for the MTL models in an input-dependent
manner. Specifically, we attach a task-aware lightweight policy network to the
shared encoder and co-train it alongside the MTL model to recognize unnecessary
computations. During runtime, our task-aware policy network decides which parts
of the model to activate depending on the input frame and the target
computational complexity. Extensive experiments on the PASCAL dataset
demonstrate that AdaMTL reduces the computational complexity by 43% while
improving the accuracy by 1.32% compared to single-task models. Combined with
SOTA MTL methodologies, AdaMTL boosts the accuracy by 7.8% while improving the
efficiency by 3.1X. When deployed on Vuzix M4000 smart glasses, AdaMTL reduces
the inference latency and the energy consumption by up to 21.8% and 37.5%,
respectively, compared to the static MTL model. Our code is publicly available
at https://github.com/scale-lab/AdaMTL.git.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 20:17:44 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Neseem",
"Marina",
""
],
[
"Agiza",
"Ahmed",
""
],
[
"Reda",
"Sherief",
""
]
] | TITLE: AdaMTL: Adaptive Input-dependent Inference for Efficient Multi-Task
Learning
ABSTRACT: Modern Augmented reality applications require performing multiple tasks on
each input frame simultaneously. Multi-task learning (MTL) represents an
effective approach where multiple tasks share an encoder to extract
representative features from the input frame, followed by task-specific
decoders to generate predictions for each task. Generally, the shared encoder
in MTL models needs to have a large representational capacity in order to
generalize well to various tasks and input data, which has a negative effect on
the inference latency. In this paper, we argue that due to the large variations
in the complexity of the input frames, some computations might be unnecessary
for the output. Therefore, we introduce AdaMTL, an adaptive framework that
learns task-aware inference policies for the MTL models in an input-dependent
manner. Specifically, we attach a task-aware lightweight policy network to the
shared encoder and co-train it alongside the MTL model to recognize unnecessary
computations. During runtime, our task-aware policy network decides which parts
of the model to activate depending on the input frame and the target
computational complexity. Extensive experiments on the PASCAL dataset
demonstrate that AdaMTL reduces the computational complexity by 43% while
improving the accuracy by 1.32% compared to single-task models. Combined with
SOTA MTL methodologies, AdaMTL boosts the accuracy by 7.8% while improving the
efficiency by 3.1X. When deployed on Vuzix M4000 smart glasses, AdaMTL reduces
the inference latency and the energy consumption by up to 21.8% and 37.5%,
respectively, compared to the static MTL model. Our code is publicly available
at https://github.com/scale-lab/AdaMTL.git.
|
2304.08597 | Yash Garg | Haoxiang Zhang, Juliana Freire, Yash Garg | eTOP: Early Termination of Pipelines for Faster Training of AutoML
Systems | NA | null | null | null | cs.LG cs.IR | http://creativecommons.org/licenses/by/4.0/ | Recent advancements in software and hardware technologies have enabled the
use of AI/ML models in everyday applications has significantly improved the
quality of service rendered. However, for a given application, finding the
right AI/ML model is a complex and costly process, that involves the
generation, training, and evaluation of multiple interlinked steps (called
pipelines), such as data pre-processing, feature engineering, selection, and
model tuning. These pipelines are complex (in structure) and costly (both in
compute resource and time) to execute end-to-end, with a hyper-parameter
associated with each step. AutoML systems automate the search of these
hyper-parameters but are slow, as they rely on optimizing the pipeline's end
output. We propose the eTOP Framework which works on top of any AutoML system
and decides whether or not to execute the pipeline to the end or terminate at
an intermediate step. Experimental evaluation on 26 benchmark datasets and
integration of eTOPwith MLBox4 reduces the training time of the AutoML system
upto 40x than baseline MLBox.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 20:22:30 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zhang",
"Haoxiang",
""
],
[
"Freire",
"Juliana",
""
],
[
"Garg",
"Yash",
""
]
] | TITLE: eTOP: Early Termination of Pipelines for Faster Training of AutoML
Systems
ABSTRACT: Recent advancements in software and hardware technologies have enabled the
use of AI/ML models in everyday applications has significantly improved the
quality of service rendered. However, for a given application, finding the
right AI/ML model is a complex and costly process, that involves the
generation, training, and evaluation of multiple interlinked steps (called
pipelines), such as data pre-processing, feature engineering, selection, and
model tuning. These pipelines are complex (in structure) and costly (both in
compute resource and time) to execute end-to-end, with a hyper-parameter
associated with each step. AutoML systems automate the search of these
hyper-parameters but are slow, as they rely on optimizing the pipeline's end
output. We propose the eTOP Framework which works on top of any AutoML system
and decides whether or not to execute the pipeline to the end or terminate at
an intermediate step. Experimental evaluation on 26 benchmark datasets and
integration of eTOPwith MLBox4 reduces the training time of the AutoML system
upto 40x than baseline MLBox.
|
2304.08640 | Baixiang Huang | Baixiang Huang, Bryan Hooi, Kai Shu | TAP: A Comprehensive Data Repository for Traffic Accident Prediction in
Road Networks | 10 pages, 5 figures | null | null | null | cs.LG cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Road safety is a major global public health concern. Effective traffic crash
prediction can play a critical role in reducing road traffic accidents.
However, Existing machine learning approaches tend to focus on predicting
traffic accidents in isolation, without considering the potential relationships
between different accident locations within road networks. To incorporate graph
structure information, graph-based approaches such as Graph Neural Networks
(GNNs) can be naturally applied. However, applying GNNs to the accident
prediction problem faces challenges due to the lack of suitable
graph-structured traffic accident datasets. To bridge this gap, we have
constructed a real-world graph-based Traffic Accident Prediction (TAP) data
repository, along with two representative tasks: accident occurrence prediction
and accident severity prediction. With nationwide coverage, real-world network
topology, and rich geospatial features, this data repository can be used for a
variety of traffic-related tasks. We further comprehensively evaluate eleven
state-of-the-art GNN variants and two non-graph-based machine learning methods
using the created datasets. Significantly facilitated by the proposed data, we
develop a novel Traffic Accident Vulnerability Estimation via Linkage (TRAVEL)
model, which is designed to capture angular and directional information from
road networks. We demonstrate that the proposed model consistently outperforms
the baselines. The data and code are available on GitHub
(https://github.com/baixianghuang/travel).
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 22:18:58 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Huang",
"Baixiang",
""
],
[
"Hooi",
"Bryan",
""
],
[
"Shu",
"Kai",
""
]
] | TITLE: TAP: A Comprehensive Data Repository for Traffic Accident Prediction in
Road Networks
ABSTRACT: Road safety is a major global public health concern. Effective traffic crash
prediction can play a critical role in reducing road traffic accidents.
However, Existing machine learning approaches tend to focus on predicting
traffic accidents in isolation, without considering the potential relationships
between different accident locations within road networks. To incorporate graph
structure information, graph-based approaches such as Graph Neural Networks
(GNNs) can be naturally applied. However, applying GNNs to the accident
prediction problem faces challenges due to the lack of suitable
graph-structured traffic accident datasets. To bridge this gap, we have
constructed a real-world graph-based Traffic Accident Prediction (TAP) data
repository, along with two representative tasks: accident occurrence prediction
and accident severity prediction. With nationwide coverage, real-world network
topology, and rich geospatial features, this data repository can be used for a
variety of traffic-related tasks. We further comprehensively evaluate eleven
state-of-the-art GNN variants and two non-graph-based machine learning methods
using the created datasets. Significantly facilitated by the proposed data, we
develop a novel Traffic Accident Vulnerability Estimation via Linkage (TRAVEL)
model, which is designed to capture angular and directional information from
road networks. We demonstrate that the proposed model consistently outperforms
the baselines. The data and code are available on GitHub
(https://github.com/baixianghuang/travel).
|
2304.08653 | Polina Zablotskaia | Polina Zablotskaia, Du Phan, Joshua Maynez, Shashi Narayan, Jie Ren,
Jeremiah Liu | On Uncertainty Calibration and Selective Generation in Probabilistic
Neural Summarization: A Benchmark Study | null | null | null | null | cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | Modern deep models for summarization attains impressive benchmark
performance, but they are prone to generating miscalibrated predictive
uncertainty. This means that they assign high confidence to low-quality
predictions, leading to compromised reliability and trustworthiness in
real-world applications. Probabilistic deep learning methods are common
solutions to the miscalibration problem. However, their relative effectiveness
in complex autoregressive summarization tasks are not well-understood. In this
work, we thoroughly investigate different state-of-the-art probabilistic
methods' effectiveness in improving the uncertainty quality of the neural
summarization models, across three large-scale benchmarks with varying
difficulty. We show that the probabilistic methods consistently improve the
model's generation and uncertainty quality, leading to improved selective
generation performance (i.e., abstaining from low-quality summaries) in
practice. We also reveal notable failure patterns of probabilistic methods
widely-adopted in NLP community (e.g., Deep Ensemble and Monte Carlo Dropout),
cautioning the importance of choosing appropriate method for the data setting.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 23:06:28 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zablotskaia",
"Polina",
""
],
[
"Phan",
"Du",
""
],
[
"Maynez",
"Joshua",
""
],
[
"Narayan",
"Shashi",
""
],
[
"Ren",
"Jie",
""
],
[
"Liu",
"Jeremiah",
""
]
] | TITLE: On Uncertainty Calibration and Selective Generation in Probabilistic
Neural Summarization: A Benchmark Study
ABSTRACT: Modern deep models for summarization attains impressive benchmark
performance, but they are prone to generating miscalibrated predictive
uncertainty. This means that they assign high confidence to low-quality
predictions, leading to compromised reliability and trustworthiness in
real-world applications. Probabilistic deep learning methods are common
solutions to the miscalibration problem. However, their relative effectiveness
in complex autoregressive summarization tasks are not well-understood. In this
work, we thoroughly investigate different state-of-the-art probabilistic
methods' effectiveness in improving the uncertainty quality of the neural
summarization models, across three large-scale benchmarks with varying
difficulty. We show that the probabilistic methods consistently improve the
model's generation and uncertainty quality, leading to improved selective
generation performance (i.e., abstaining from low-quality summaries) in
practice. We also reveal notable failure patterns of probabilistic methods
widely-adopted in NLP community (e.g., Deep Ensemble and Monte Carlo Dropout),
cautioning the importance of choosing appropriate method for the data setting.
|
2304.08660 | Alex Junho Lee | Alex Junho Lee, Seungwon Song, Hyungtae Lim, Woojoo Lee and Hyun Myung | (LC)$^2$: LiDAR-Camera Loop Constraints For Cross-Modal Place
Recognition | 8 pages, 11 figures, Accepted to IEEE Robotics and Automation Letters
(RA-L) | null | null | null | cs.RO cs.AI | http://creativecommons.org/licenses/by/4.0/ | Localization has been a challenging task for autonomous navigation. A loop
detection algorithm must overcome environmental changes for the place
recognition and re-localization of robots. Therefore, deep learning has been
extensively studied for the consistent transformation of measurements into
localization descriptors. Street view images are easily accessible; however,
images are vulnerable to appearance changes. LiDAR can robustly provide precise
structural information. However, constructing a point cloud database is
expensive, and point clouds exist only in limited places. Different from
previous works that train networks to produce shared embedding directly between
the 2D image and 3D point cloud, we transform both data into 2.5D depth images
for matching. In this work, we propose a novel cross-matching method, called
(LC)$^2$, for achieving LiDAR localization without a prior point cloud map. To
this end, LiDAR measurements are expressed in the form of range images before
matching them to reduce the modality discrepancy. Subsequently, the network is
trained to extract localization descriptors from disparity and range images.
Next, the best matches are employed as a loop factor in a pose graph. Using
public datasets that include multiple sessions in significantly different
lighting conditions, we demonstrated that LiDAR-based navigation systems could
be optimized from image databases and vice versa.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 23:20:16 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Lee",
"Alex Junho",
""
],
[
"Song",
"Seungwon",
""
],
[
"Lim",
"Hyungtae",
""
],
[
"Lee",
"Woojoo",
""
],
[
"Myung",
"Hyun",
""
]
] | TITLE: (LC)$^2$: LiDAR-Camera Loop Constraints For Cross-Modal Place
Recognition
ABSTRACT: Localization has been a challenging task for autonomous navigation. A loop
detection algorithm must overcome environmental changes for the place
recognition and re-localization of robots. Therefore, deep learning has been
extensively studied for the consistent transformation of measurements into
localization descriptors. Street view images are easily accessible; however,
images are vulnerable to appearance changes. LiDAR can robustly provide precise
structural information. However, constructing a point cloud database is
expensive, and point clouds exist only in limited places. Different from
previous works that train networks to produce shared embedding directly between
the 2D image and 3D point cloud, we transform both data into 2.5D depth images
for matching. In this work, we propose a novel cross-matching method, called
(LC)$^2$, for achieving LiDAR localization without a prior point cloud map. To
this end, LiDAR measurements are expressed in the form of range images before
matching them to reduce the modality discrepancy. Subsequently, the network is
trained to extract localization descriptors from disparity and range images.
Next, the best matches are employed as a loop factor in a pose graph. Using
public datasets that include multiple sessions in significantly different
lighting conditions, we demonstrated that LiDAR-based navigation systems could
be optimized from image databases and vice versa.
|
2304.08673 | Nishant Panda | Nishant Panda, Natalie Klein, Dominic Yang, Patrick Gasda and Diane
Oyen | Semi-supervised Learning of Pushforwards For Domain Translation &
Adaptation | 19 pages, 7 figures | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given two probability densities on related data spaces, we seek a map pushing
one density to the other while satisfying application-dependent constraints.
For maps to have utility in a broad application space (including domain
translation, domain adaptation, and generative modeling), the map must be
available to apply on out-of-sample data points and should correspond to a
probabilistic model over the two spaces. Unfortunately, existing approaches,
which are primarily based on optimal transport, do not address these needs. In
this paper, we introduce a novel pushforward map learning algorithm that
utilizes normalizing flows to parameterize the map. We first re-formulate the
classical optimal transport problem to be map-focused and propose a learning
algorithm to select from all possible maps under the constraint that the map
minimizes a probability distance and application-specific regularizers; thus,
our method can be seen as solving a modified optimal transport problem. Once
the map is learned, it can be used to map samples from a source domain to a
target domain. In addition, because the map is parameterized as a composition
of normalizing flows, it models the empirical distributions over the two data
spaces and allows both sampling and likelihood evaluation for both data sets.
We compare our method (parOT) to related optimal transport approaches in the
context of domain adaptation and domain translation on benchmark data sets.
Finally, to illustrate the impact of our work on applied problems, we apply
parOT to a real scientific application: spectral calibration for
high-dimensional measurements from two vastly different environments
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 00:35:32 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Panda",
"Nishant",
""
],
[
"Klein",
"Natalie",
""
],
[
"Yang",
"Dominic",
""
],
[
"Gasda",
"Patrick",
""
],
[
"Oyen",
"Diane",
""
]
] | TITLE: Semi-supervised Learning of Pushforwards For Domain Translation &
Adaptation
ABSTRACT: Given two probability densities on related data spaces, we seek a map pushing
one density to the other while satisfying application-dependent constraints.
For maps to have utility in a broad application space (including domain
translation, domain adaptation, and generative modeling), the map must be
available to apply on out-of-sample data points and should correspond to a
probabilistic model over the two spaces. Unfortunately, existing approaches,
which are primarily based on optimal transport, do not address these needs. In
this paper, we introduce a novel pushforward map learning algorithm that
utilizes normalizing flows to parameterize the map. We first re-formulate the
classical optimal transport problem to be map-focused and propose a learning
algorithm to select from all possible maps under the constraint that the map
minimizes a probability distance and application-specific regularizers; thus,
our method can be seen as solving a modified optimal transport problem. Once
the map is learned, it can be used to map samples from a source domain to a
target domain. In addition, because the map is parameterized as a composition
of normalizing flows, it models the empirical distributions over the two data
spaces and allows both sampling and likelihood evaluation for both data sets.
We compare our method (parOT) to related optimal transport approaches in the
context of domain adaptation and domain translation on benchmark data sets.
Finally, to illustrate the impact of our work on applied problems, we apply
parOT to a real scientific application: spectral calibration for
high-dimensional measurements from two vastly different environments
|
2304.08687 | Meiqi Hu | Meiqi Hu, Chen Wu, Liangpei Zhang | GlobalMind: Global Multi-head Interactive Self-attention Network for
Hyperspectral Change Detection | 14 page, 18 figures | null | null | null | cs.CV eess.IV | http://creativecommons.org/licenses/by-sa/4.0/ | High spectral resolution imagery of the Earth's surface enables users to
monitor changes over time in fine-grained scale, playing an increasingly
important role in agriculture, defense, and emergency response. However, most
current algorithms are still confined to describing local features and fail to
incorporate a global perspective, which limits their ability to capture
interactions between global features, thus usually resulting in incomplete
change regions. In this paper, we propose a Global Multi-head INteractive
self-attention change Detection network (GlobalMind) to explore the implicit
correlation between different surface objects and variant land cover
transformations, acquiring a comprehensive understanding of the data and
accurate change detection result. Firstly, a simple but effective Global Axial
Segmentation (GAS) strategy is designed to expand the self-attention
computation along the row space or column space of hyperspectral images,
allowing the global connection with high efficiency. Secondly, with GAS, the
global spatial multi-head interactive self-attention (Global-M) module is
crafted to mine the abundant spatial-spectral feature involving potential
correlations between the ground objects from the entire rich and complex
hyperspectral space. Moreover, to acquire the accurate and complete
cross-temporal changes, we devise a global temporal interactive multi-head
self-attention (GlobalD) module which incorporates the relevance and variation
of bi-temporal spatial-spectral features, deriving the integrate potential same
kind of changes in the local and global range with the combination of GAS. We
perform extensive experiments on five mostly used hyperspectral datasets, and
our method outperforms the state-of-the-art algorithms with high accuracy and
efficiency.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 01:43:17 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Hu",
"Meiqi",
""
],
[
"Wu",
"Chen",
""
],
[
"Zhang",
"Liangpei",
""
]
] | TITLE: GlobalMind: Global Multi-head Interactive Self-attention Network for
Hyperspectral Change Detection
ABSTRACT: High spectral resolution imagery of the Earth's surface enables users to
monitor changes over time in fine-grained scale, playing an increasingly
important role in agriculture, defense, and emergency response. However, most
current algorithms are still confined to describing local features and fail to
incorporate a global perspective, which limits their ability to capture
interactions between global features, thus usually resulting in incomplete
change regions. In this paper, we propose a Global Multi-head INteractive
self-attention change Detection network (GlobalMind) to explore the implicit
correlation between different surface objects and variant land cover
transformations, acquiring a comprehensive understanding of the data and
accurate change detection result. Firstly, a simple but effective Global Axial
Segmentation (GAS) strategy is designed to expand the self-attention
computation along the row space or column space of hyperspectral images,
allowing the global connection with high efficiency. Secondly, with GAS, the
global spatial multi-head interactive self-attention (Global-M) module is
crafted to mine the abundant spatial-spectral feature involving potential
correlations between the ground objects from the entire rich and complex
hyperspectral space. Moreover, to acquire the accurate and complete
cross-temporal changes, we devise a global temporal interactive multi-head
self-attention (GlobalD) module which incorporates the relevance and variation
of bi-temporal spatial-spectral features, deriving the integrate potential same
kind of changes in the local and global range with the combination of GAS. We
perform extensive experiments on five mostly used hyperspectral datasets, and
our method outperforms the state-of-the-art algorithms with high accuracy and
efficiency.
|
2304.08703 | Giang Cao | Hoang-Giang Cao, Weihao Zeng, I-Chen Wu | Learning Sim-to-Real Dense Object Descriptors for Robotic Manipulation | Accepted to International Conference on Robotics and Automation
(ICRA) 2023 | null | null | null | cs.RO cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | It is crucial to address the following issues for ubiquitous robotics
manipulation applications: (a) vision-based manipulation tasks require the
robot to visually learn and understand the object with rich information like
dense object descriptors; and (b) sim-to-real transfer in robotics aims to
close the gap between simulated and real data. In this paper, we present
Sim-to-Real Dense Object Nets (SRDONs), a dense object descriptor that not only
understands the object via appropriate representation but also maps simulated
and real data to a unified feature space with pixel consistency. We proposed an
object-to-object matching method for image pairs from different scenes and
different domains. This method helps reduce the effort of training data from
real-world by taking advantage of public datasets, such as GraspNet. With
sim-to-real object representation consistency, our SRDONs can serve as a
building block for a variety of sim-to-real manipulation tasks. We demonstrate
in experiments that pre-trained SRDONs significantly improve performances on
unseen objects and unseen visual environments for various robotic tasks with
zero real-world training.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 02:28:55 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Cao",
"Hoang-Giang",
""
],
[
"Zeng",
"Weihao",
""
],
[
"Wu",
"I-Chen",
""
]
] | TITLE: Learning Sim-to-Real Dense Object Descriptors for Robotic Manipulation
ABSTRACT: It is crucial to address the following issues for ubiquitous robotics
manipulation applications: (a) vision-based manipulation tasks require the
robot to visually learn and understand the object with rich information like
dense object descriptors; and (b) sim-to-real transfer in robotics aims to
close the gap between simulated and real data. In this paper, we present
Sim-to-Real Dense Object Nets (SRDONs), a dense object descriptor that not only
understands the object via appropriate representation but also maps simulated
and real data to a unified feature space with pixel consistency. We proposed an
object-to-object matching method for image pairs from different scenes and
different domains. This method helps reduce the effort of training data from
real-world by taking advantage of public datasets, such as GraspNet. With
sim-to-real object representation consistency, our SRDONs can serve as a
building block for a variety of sim-to-real manipulation tasks. We demonstrate
in experiments that pre-trained SRDONs significantly improve performances on
unseen objects and unseen visual environments for various robotic tasks with
zero real-world training.
|
2304.08706 | Peng-Tao Jiang | Jiaxiong Qiu, Peng-Tao Jiang, Yifan Zhu, Ze-Xin Yin, Ming-Ming Cheng,
Bo Ren | Looking Through the Glass: Neural Surface Reconstruction Against High
Specular Reflections | 17 pages, 20 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Neural implicit methods have achieved high-quality 3D object surfaces under
slight specular highlights. However, high specular reflections (HSR) often
appear in front of target objects when we capture them through glasses. The
complex ambiguity in these scenes violates the multi-view consistency, then
makes it challenging for recent methods to reconstruct target objects
correctly. To remedy this issue, we present a novel surface reconstruction
framework, NeuS-HSR, based on implicit neural rendering. In NeuS-HSR, the
object surface is parameterized as an implicit signed distance function (SDF).
To reduce the interference of HSR, we propose decomposing the rendered image
into two appearances: the target object and the auxiliary plane. We design a
novel auxiliary plane module by combining physical assumptions and neural
networks to generate the auxiliary plane appearance. Extensive experiments on
synthetic and real-world datasets demonstrate that NeuS-HSR outperforms
state-of-the-art approaches for accurate and robust target surface
reconstruction against HSR. Code is available at
https://github.com/JiaxiongQ/NeuS-HSR.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 02:34:58 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Qiu",
"Jiaxiong",
""
],
[
"Jiang",
"Peng-Tao",
""
],
[
"Zhu",
"Yifan",
""
],
[
"Yin",
"Ze-Xin",
""
],
[
"Cheng",
"Ming-Ming",
""
],
[
"Ren",
"Bo",
""
]
] | TITLE: Looking Through the Glass: Neural Surface Reconstruction Against High
Specular Reflections
ABSTRACT: Neural implicit methods have achieved high-quality 3D object surfaces under
slight specular highlights. However, high specular reflections (HSR) often
appear in front of target objects when we capture them through glasses. The
complex ambiguity in these scenes violates the multi-view consistency, then
makes it challenging for recent methods to reconstruct target objects
correctly. To remedy this issue, we present a novel surface reconstruction
framework, NeuS-HSR, based on implicit neural rendering. In NeuS-HSR, the
object surface is parameterized as an implicit signed distance function (SDF).
To reduce the interference of HSR, we propose decomposing the rendered image
into two appearances: the target object and the auxiliary plane. We design a
novel auxiliary plane module by combining physical assumptions and neural
networks to generate the auxiliary plane appearance. Extensive experiments on
synthetic and real-world datasets demonstrate that NeuS-HSR outperforms
state-of-the-art approaches for accurate and robust target surface
reconstruction against HSR. Code is available at
https://github.com/JiaxiongQ/NeuS-HSR.
|
2304.08708 | Hao Cai | Xiaoping Xie, Hao Cai, Can Li, Fei Ding | A Voice Disease Detection Method Based on MFCCs and Shallow CNN | null | null | null | null | cs.SD eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The incidence rate of voice diseases is increasing year by year. The use of
software for remote diagnosis is a technical development trend and has
important practical value. Among voice diseases, common diseases that cause
hoarseness include spasmodic dysphonia, vocal cord paralysis, vocal nodule, and
vocal cord polyp. This paper presents a voice disease detection method that can
be applied in a wide range of clinical. We cooperated with Xiangya Hospital of
Central South University to collect voice samples from sixty-one different
patients. The Mel Frequency Cepstrum Coefficient (MFCC) parameters are
extracted as input features to describe the voice in the form of data. An
innovative model combining MFCC parameters and single convolution layer CNN is
proposed for fast calculation and classification. The highest accuracy we
achieved was 92%, it is fully ahead of the original research results and
internationally advanced. And we use Advanced Voice Function Assessment
Databases (AVFAD) to evaluate the generalization ability of the method we
proposed, which achieved an accuracy rate of 98%. Experiments on clinical and
standard datasets show that for the pathological detection of voice diseases,
our method has greatly improved in accuracy and computational efficiency.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 02:41:59 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Xie",
"Xiaoping",
""
],
[
"Cai",
"Hao",
""
],
[
"Li",
"Can",
""
],
[
"Ding",
"Fei",
""
]
] | TITLE: A Voice Disease Detection Method Based on MFCCs and Shallow CNN
ABSTRACT: The incidence rate of voice diseases is increasing year by year. The use of
software for remote diagnosis is a technical development trend and has
important practical value. Among voice diseases, common diseases that cause
hoarseness include spasmodic dysphonia, vocal cord paralysis, vocal nodule, and
vocal cord polyp. This paper presents a voice disease detection method that can
be applied in a wide range of clinical. We cooperated with Xiangya Hospital of
Central South University to collect voice samples from sixty-one different
patients. The Mel Frequency Cepstrum Coefficient (MFCC) parameters are
extracted as input features to describe the voice in the form of data. An
innovative model combining MFCC parameters and single convolution layer CNN is
proposed for fast calculation and classification. The highest accuracy we
achieved was 92%, it is fully ahead of the original research results and
internationally advanced. And we use Advanced Voice Function Assessment
Databases (AVFAD) to evaluate the generalization ability of the method we
proposed, which achieved an accuracy rate of 98%. Experiments on clinical and
standard datasets show that for the pathological detection of voice diseases,
our method has greatly improved in accuracy and computational efficiency.
|
2304.08802 | Stein Stroobants | Stein Stroobants, Julien Dupeyroux, Guido C.H.E. de Croon | Neuromorphic computing for attitude estimation onboard quadrotors | null | Neuromorphic Computing and Engineering 2.3 (2022): 034005 | 10.1088/2634-4386/ac7ee0 | null | cs.RO | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Compelling evidence has been given for the high energy efficiency and update
rates of neuromorphic processors, with performance beyond what standard Von
Neumann architectures can achieve. Such promising features could be
advantageous in critical embedded systems, especially in robotics. To date, the
constraints inherent in robots (e.g., size and weight, battery autonomy,
available sensors, computing resources, processing time, etc.), and
particularly in aerial vehicles, severely hamper the performance of
fully-autonomous on-board control, including sensor processing and state
estimation. In this work, we propose a spiking neural network (SNN) capable of
estimating the pitch and roll angles of a quadrotor in highly dynamic movements
from 6-degree of freedom Inertial Measurement Unit (IMU) data. With only 150
neurons and a limited training dataset obtained using a quadrotor in a real
world setup, the network shows competitive results as compared to
state-of-the-art, non-neuromorphic attitude estimators. The proposed
architecture was successfully tested on the Loihi neuromorphic processor
on-board a quadrotor to estimate the attitude when flying. Our results show the
robustness of neuromorphic attitude estimation and pave the way towards
energy-efficient, fully autonomous control of quadrotors with dedicated
neuromorphic computing systems.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:07:22 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Stroobants",
"Stein",
""
],
[
"Dupeyroux",
"Julien",
""
],
[
"de Croon",
"Guido C. H. E.",
""
]
] | TITLE: Neuromorphic computing for attitude estimation onboard quadrotors
ABSTRACT: Compelling evidence has been given for the high energy efficiency and update
rates of neuromorphic processors, with performance beyond what standard Von
Neumann architectures can achieve. Such promising features could be
advantageous in critical embedded systems, especially in robotics. To date, the
constraints inherent in robots (e.g., size and weight, battery autonomy,
available sensors, computing resources, processing time, etc.), and
particularly in aerial vehicles, severely hamper the performance of
fully-autonomous on-board control, including sensor processing and state
estimation. In this work, we propose a spiking neural network (SNN) capable of
estimating the pitch and roll angles of a quadrotor in highly dynamic movements
from 6-degree of freedom Inertial Measurement Unit (IMU) data. With only 150
neurons and a limited training dataset obtained using a quadrotor in a real
world setup, the network shows competitive results as compared to
state-of-the-art, non-neuromorphic attitude estimators. The proposed
architecture was successfully tested on the Loihi neuromorphic processor
on-board a quadrotor to estimate the attitude when flying. Our results show the
robustness of neuromorphic attitude estimation and pave the way towards
energy-efficient, fully autonomous control of quadrotors with dedicated
neuromorphic computing systems.
|
2304.08803 | Guoliang Xu | Guoliang Xu, Jianqin Yin | MLP-AIR: An Efficient MLP-Based Method for Actor Interaction Relation
Learning in Group Activity Recognition | Submit to Neurocomputing | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The task of Group Activity Recognition (GAR) aims to predict the activity
category of the group by learning the actor spatial-temporal interaction
relation in the group. Therefore, an effective actor relation learning method
is crucial for the GAR task. The previous works mainly learn the interaction
relation by the well-designed GCNs or Transformers. For example, to infer the
actor interaction relation, GCNs need a learnable adjacency, and Transformers
need to calculate the self-attention. Although the above methods can model the
interaction relation effectively, they also increase the complexity of the
model (the number of parameters and computations). In this paper, we design a
novel MLP-based method for Actor Interaction Relation learning (MLP-AIR) in
GAR. Compared with GCNs and Transformers, our method has a competitive but
conceptually and technically simple alternative, significantly reducing the
complexity. Specifically, MLP-AIR includes three sub-modules: MLP-based Spatial
relation modeling module (MLP-S), MLP-based Temporal relation modeling module
(MLP-T), and MLP-based Relation refining module (MLP-R). MLP-S is used to model
the spatial relation between different actors in each frame. MLP-T is used to
model the temporal relation between different frames for each actor. MLP-R is
used further to refine the relation between different dimensions of relation
features to improve the feature's expression ability. To evaluate the MLP-AIR,
we conduct extensive experiments on two widely used benchmarks, including the
Volleyball and Collective Activity datasets. Experimental results demonstrate
that MLP-AIR can get competitive results but with low complexity.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:07:23 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Xu",
"Guoliang",
""
],
[
"Yin",
"Jianqin",
""
]
] | TITLE: MLP-AIR: An Efficient MLP-Based Method for Actor Interaction Relation
Learning in Group Activity Recognition
ABSTRACT: The task of Group Activity Recognition (GAR) aims to predict the activity
category of the group by learning the actor spatial-temporal interaction
relation in the group. Therefore, an effective actor relation learning method
is crucial for the GAR task. The previous works mainly learn the interaction
relation by the well-designed GCNs or Transformers. For example, to infer the
actor interaction relation, GCNs need a learnable adjacency, and Transformers
need to calculate the self-attention. Although the above methods can model the
interaction relation effectively, they also increase the complexity of the
model (the number of parameters and computations). In this paper, we design a
novel MLP-based method for Actor Interaction Relation learning (MLP-AIR) in
GAR. Compared with GCNs and Transformers, our method has a competitive but
conceptually and technically simple alternative, significantly reducing the
complexity. Specifically, MLP-AIR includes three sub-modules: MLP-based Spatial
relation modeling module (MLP-S), MLP-based Temporal relation modeling module
(MLP-T), and MLP-based Relation refining module (MLP-R). MLP-S is used to model
the spatial relation between different actors in each frame. MLP-T is used to
model the temporal relation between different frames for each actor. MLP-R is
used further to refine the relation between different dimensions of relation
features to improve the feature's expression ability. To evaluate the MLP-AIR,
we conduct extensive experiments on two widely used benchmarks, including the
Volleyball and Collective Activity datasets. Experimental results demonstrate
that MLP-AIR can get competitive results but with low complexity.
|
2304.08809 | Yi Li | Yi Li, Kyle Min, Subarna Tripathi, Nuno Vasconcelos | SViTT: Temporal Learning of Sparse Video-Text Transformers | CVPR 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Do video-text transformers learn to model temporal relationships across
frames? Despite their immense capacity and the abundance of multimodal training
data, recent work has revealed the strong tendency of video-text models towards
frame-based spatial representations, while temporal reasoning remains largely
unsolved. In this work, we identify several key challenges in temporal learning
of video-text transformers: the spatiotemporal trade-off from limited network
size; the curse of dimensionality for multi-frame modeling; and the diminishing
returns of semantic information by extending clip length. Guided by these
findings, we propose SViTT, a sparse video-text architecture that performs
multi-frame reasoning with significantly lower cost than naive transformers
with dense attention. Analogous to graph-based networks, SViTT employs two
forms of sparsity: edge sparsity that limits the query-key communications
between tokens in self-attention, and node sparsity that discards uninformative
visual tokens. Trained with a curriculum which increases model sparsity with
the clip length, SViTT outperforms dense transformer baselines on multiple
video-text retrieval and question answering benchmarks, with a fraction of
computational cost. Project page: http://svcl.ucsd.edu/projects/svitt.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:17:58 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Li",
"Yi",
""
],
[
"Min",
"Kyle",
""
],
[
"Tripathi",
"Subarna",
""
],
[
"Vasconcelos",
"Nuno",
""
]
] | TITLE: SViTT: Temporal Learning of Sparse Video-Text Transformers
ABSTRACT: Do video-text transformers learn to model temporal relationships across
frames? Despite their immense capacity and the abundance of multimodal training
data, recent work has revealed the strong tendency of video-text models towards
frame-based spatial representations, while temporal reasoning remains largely
unsolved. In this work, we identify several key challenges in temporal learning
of video-text transformers: the spatiotemporal trade-off from limited network
size; the curse of dimensionality for multi-frame modeling; and the diminishing
returns of semantic information by extending clip length. Guided by these
findings, we propose SViTT, a sparse video-text architecture that performs
multi-frame reasoning with significantly lower cost than naive transformers
with dense attention. Analogous to graph-based networks, SViTT employs two
forms of sparsity: edge sparsity that limits the query-key communications
between tokens in self-attention, and node sparsity that discards uninformative
visual tokens. Trained with a curriculum which increases model sparsity with
the clip length, SViTT outperforms dense transformer baselines on multiple
video-text retrieval and question answering benchmarks, with a fraction of
computational cost. Project page: http://svcl.ucsd.edu/projects/svitt.
|
2304.08820 | Jinming Su | Jinming Su, Ruihong Yin, Shuaibin Zhang and Junfeng Luo | Motion-state Alignment for Video Semantic Segmentation | Accepted by CVPR Workshops 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In recent years, video semantic segmentation has made great progress with
advanced deep neural networks. However, there still exist two main challenges
\ie, information inconsistency and computation cost. To deal with the two
difficulties, we propose a novel motion-state alignment framework for video
semantic segmentation to keep both motion and state consistency. In the
framework, we first construct a motion alignment branch armed with an efficient
decoupled transformer to capture dynamic semantics, guaranteeing region-level
temporal consistency. Then, a state alignment branch composed of a stage
transformer is designed to enrich feature spaces for the current frame to
extract static semantics and achieve pixel-level state consistency. Next, by a
semantic assignment mechanism, the region descriptor of each semantic category
is gained from dynamic semantics and linked with pixel descriptors from static
semantics. Benefiting from the alignment of these two kinds of effective
information, the proposed method picks up dynamic and static semantics in a
targeted way, so that video semantic regions are consistently segmented to
obtain precise locations with low computational complexity. Extensive
experiments on Cityscapes and CamVid datasets show that the proposed approach
outperforms state-of-the-art methods and validates the effectiveness of the
motion-state alignment framework.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:34:46 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Su",
"Jinming",
""
],
[
"Yin",
"Ruihong",
""
],
[
"Zhang",
"Shuaibin",
""
],
[
"Luo",
"Junfeng",
""
]
] | TITLE: Motion-state Alignment for Video Semantic Segmentation
ABSTRACT: In recent years, video semantic segmentation has made great progress with
advanced deep neural networks. However, there still exist two main challenges
\ie, information inconsistency and computation cost. To deal with the two
difficulties, we propose a novel motion-state alignment framework for video
semantic segmentation to keep both motion and state consistency. In the
framework, we first construct a motion alignment branch armed with an efficient
decoupled transformer to capture dynamic semantics, guaranteeing region-level
temporal consistency. Then, a state alignment branch composed of a stage
transformer is designed to enrich feature spaces for the current frame to
extract static semantics and achieve pixel-level state consistency. Next, by a
semantic assignment mechanism, the region descriptor of each semantic category
is gained from dynamic semantics and linked with pixel descriptors from static
semantics. Benefiting from the alignment of these two kinds of effective
information, the proposed method picks up dynamic and static semantics in a
targeted way, so that video semantic regions are consistently segmented to
obtain precise locations with low computational complexity. Extensive
experiments on Cityscapes and CamVid datasets show that the proposed approach
outperforms state-of-the-art methods and validates the effectiveness of the
motion-state alignment framework.
|
2304.08821 | Yuwei Yin | Yuwei Yin, Jean Kaddour, Xiang Zhang, Yixin Nie, Zhenguang Liu,
Lingpeng Kong, Qi Liu | TTIDA: Controllable Generative Data Augmentation via Text-to-Text and
Text-to-Image Models | null | null | null | null | cs.CV cs.CL cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data augmentation has been established as an efficacious approach to
supplement useful information for low-resource datasets. Traditional
augmentation techniques such as noise injection and image transformations have
been widely used. In addition, generative data augmentation (GDA) has been
shown to produce more diverse and flexible data. While generative adversarial
networks (GANs) have been frequently used for GDA, they lack diversity and
controllability compared to text-to-image diffusion models. In this paper, we
propose TTIDA (Text-to-Text-to-Image Data Augmentation) to leverage the
capabilities of large-scale pre-trained Text-to-Text (T2T) and Text-to-Image
(T2I) generative models for data augmentation. By conditioning the T2I model on
detailed descriptions produced by T2T models, we are able to generate
photo-realistic labeled images in a flexible and controllable manner.
Experiments on in-domain classification, cross-domain classification, and image
captioning tasks show consistent improvements over other data augmentation
baselines. Analytical studies in varied settings, including few-shot,
long-tail, and adversarial, further reinforce the effectiveness of TTIDA in
enhancing performance and increasing robustness.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:40:30 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Yin",
"Yuwei",
""
],
[
"Kaddour",
"Jean",
""
],
[
"Zhang",
"Xiang",
""
],
[
"Nie",
"Yixin",
""
],
[
"Liu",
"Zhenguang",
""
],
[
"Kong",
"Lingpeng",
""
],
[
"Liu",
"Qi",
""
]
] | TITLE: TTIDA: Controllable Generative Data Augmentation via Text-to-Text and
Text-to-Image Models
ABSTRACT: Data augmentation has been established as an efficacious approach to
supplement useful information for low-resource datasets. Traditional
augmentation techniques such as noise injection and image transformations have
been widely used. In addition, generative data augmentation (GDA) has been
shown to produce more diverse and flexible data. While generative adversarial
networks (GANs) have been frequently used for GDA, they lack diversity and
controllability compared to text-to-image diffusion models. In this paper, we
propose TTIDA (Text-to-Text-to-Image Data Augmentation) to leverage the
capabilities of large-scale pre-trained Text-to-Text (T2T) and Text-to-Image
(T2I) generative models for data augmentation. By conditioning the T2I model on
detailed descriptions produced by T2T models, we are able to generate
photo-realistic labeled images in a flexible and controllable manner.
Experiments on in-domain classification, cross-domain classification, and image
captioning tasks show consistent improvements over other data augmentation
baselines. Analytical studies in varied settings, including few-shot,
long-tail, and adversarial, further reinforce the effectiveness of TTIDA in
enhancing performance and increasing robustness.
|
2304.08823 | V\'esteinn Sn{\ae}bjarnarson | V\'esteinn Sn{\ae}bjarnarson, Annika Simonsen, Goran Glava\v{s} and
Ivan Vuli\'c | Transfer to a Low-Resource Language via Close Relatives: The Case Study
on Faroese | null | null | null | null | cs.CL | http://creativecommons.org/licenses/by/4.0/ | Multilingual language models have pushed state-of-the-art in cross-lingual
NLP transfer. The majority of zero-shot cross-lingual transfer, however, use
one and the same massively multilingual transformer (e.g., mBERT or XLM-R) to
transfer to all target languages, irrespective of their typological,
etymological, and phylogenetic relations to other languages. In particular,
readily available data and models of resource-rich sibling languages are often
ignored. In this work, we empirically show, in a case study for Faroese -- a
low-resource language from a high-resource language family -- that by
leveraging the phylogenetic information and departing from the
'one-size-fits-all' paradigm, one can improve cross-lingual transfer to
low-resource languages. In particular, we leverage abundant resources of other
Scandinavian languages (i.e., Danish, Norwegian, Swedish, and Icelandic) for
the benefit of Faroese. Our evaluation results show that we can substantially
improve the transfer performance to Faroese by exploiting data and models of
closely-related high-resource languages. Further, we release a new web corpus
of Faroese and Faroese datasets for named entity recognition (NER), semantic
text similarity (STS), and new language models trained on all Scandinavian
languages.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:42:38 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Snæbjarnarson",
"Vésteinn",
""
],
[
"Simonsen",
"Annika",
""
],
[
"Glavaš",
"Goran",
""
],
[
"Vulić",
"Ivan",
""
]
] | TITLE: Transfer to a Low-Resource Language via Close Relatives: The Case Study
on Faroese
ABSTRACT: Multilingual language models have pushed state-of-the-art in cross-lingual
NLP transfer. The majority of zero-shot cross-lingual transfer, however, use
one and the same massively multilingual transformer (e.g., mBERT or XLM-R) to
transfer to all target languages, irrespective of their typological,
etymological, and phylogenetic relations to other languages. In particular,
readily available data and models of resource-rich sibling languages are often
ignored. In this work, we empirically show, in a case study for Faroese -- a
low-resource language from a high-resource language family -- that by
leveraging the phylogenetic information and departing from the
'one-size-fits-all' paradigm, one can improve cross-lingual transfer to
low-resource languages. In particular, we leverage abundant resources of other
Scandinavian languages (i.e., Danish, Norwegian, Swedish, and Icelandic) for
the benefit of Faroese. Our evaluation results show that we can substantially
improve the transfer performance to Faroese by exploiting data and models of
closely-related high-resource languages. Further, we release a new web corpus
of Faroese and Faroese datasets for named entity recognition (NER), semantic
text similarity (STS), and new language models trained on all Scandinavian
languages.
|
2304.08826 | Jinming Su | Jinming Su, Ruihong Yin, Xingyue Chen and Junfeng Luo | Perceive, Excavate and Purify: A Novel Object Mining Framework for
Instance Segmentation | Accepted by CVPR Workshops 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Recently, instance segmentation has made great progress with the rapid
development of deep neural networks. However, there still exist two main
challenges including discovering indistinguishable objects and modeling the
relationship between instances. To deal with these difficulties, we propose a
novel object mining framework for instance segmentation. In this framework, we
first introduce the semantics perceiving subnetwork to capture pixels that may
belong to an obvious instance from the bottom up. Then, we propose an object
excavating mechanism to discover indistinguishable objects. In the mechanism,
preliminary perceived semantics are regarded as original instances with
classifications and locations, and then indistinguishable objects around these
original instances are mined, which ensures that hard objects are fully
excavated. Next, an instance purifying strategy is put forward to model the
relationship between instances, which pulls the similar instances close and
pushes away different instances to keep intra-instance similarity and
inter-instance discrimination. In this manner, the same objects are combined as
the one instance and different objects are distinguished as independent
instances. Extensive experiments on the COCO dataset show that the proposed
approach outperforms state-of-the-art methods, which validates the
effectiveness of the proposed object mining framework.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 08:47:03 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Su",
"Jinming",
""
],
[
"Yin",
"Ruihong",
""
],
[
"Chen",
"Xingyue",
""
],
[
"Luo",
"Junfeng",
""
]
] | TITLE: Perceive, Excavate and Purify: A Novel Object Mining Framework for
Instance Segmentation
ABSTRACT: Recently, instance segmentation has made great progress with the rapid
development of deep neural networks. However, there still exist two main
challenges including discovering indistinguishable objects and modeling the
relationship between instances. To deal with these difficulties, we propose a
novel object mining framework for instance segmentation. In this framework, we
first introduce the semantics perceiving subnetwork to capture pixels that may
belong to an obvious instance from the bottom up. Then, we propose an object
excavating mechanism to discover indistinguishable objects. In the mechanism,
preliminary perceived semantics are regarded as original instances with
classifications and locations, and then indistinguishable objects around these
original instances are mined, which ensures that hard objects are fully
excavated. Next, an instance purifying strategy is put forward to model the
relationship between instances, which pulls the similar instances close and
pushes away different instances to keep intra-instance similarity and
inter-instance discrimination. In this manner, the same objects are combined as
the one instance and different objects are distinguished as independent
instances. Extensive experiments on the COCO dataset show that the proposed
approach outperforms state-of-the-art methods, which validates the
effectiveness of the proposed object mining framework.
|
2304.08852 | Hassan Imani | Hassan Imani, Md Baharul Islam, Lai-Kuan Wong | Saliency-aware Stereoscopic Video Retargeting | 8 pages excluding references. CVPRW conference | null | null | null | cs.CV | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Stereo video retargeting aims to resize an image to a desired aspect ratio.
The quality of retargeted videos can be significantly impacted by the stereo
videos spatial, temporal, and disparity coherence, all of which can be impacted
by the retargeting process. Due to the lack of a publicly accessible annotated
dataset, there is little research on deep learning-based methods for stereo
video retargeting. This paper proposes an unsupervised deep learning-based
stereo video retargeting network. Our model first detects the salient objects
and shifts and warps all objects such that it minimizes the distortion of the
salient parts of the stereo frames. We use 1D convolution for shifting the
salient objects and design a stereo video Transformer to assist the retargeting
process. To train the network, we use the parallax attention mechanism to fuse
the left and right views and feed the retargeted frames to a reconstruction
module that reverses the retargeted frames to the input frames. Therefore, the
network is trained in an unsupervised manner. Extensive qualitative and
quantitative experiments and ablation studies on KITTI stereo 2012 and 2015
datasets demonstrate the efficiency of the proposed method over the existing
state-of-the-art methods. The code is available at
https://github.com/z65451/SVR/.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 09:38:33 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Imani",
"Hassan",
""
],
[
"Islam",
"Md Baharul",
""
],
[
"Wong",
"Lai-Kuan",
""
]
] | TITLE: Saliency-aware Stereoscopic Video Retargeting
ABSTRACT: Stereo video retargeting aims to resize an image to a desired aspect ratio.
The quality of retargeted videos can be significantly impacted by the stereo
videos spatial, temporal, and disparity coherence, all of which can be impacted
by the retargeting process. Due to the lack of a publicly accessible annotated
dataset, there is little research on deep learning-based methods for stereo
video retargeting. This paper proposes an unsupervised deep learning-based
stereo video retargeting network. Our model first detects the salient objects
and shifts and warps all objects such that it minimizes the distortion of the
salient parts of the stereo frames. We use 1D convolution for shifting the
salient objects and design a stereo video Transformer to assist the retargeting
process. To train the network, we use the parallax attention mechanism to fuse
the left and right views and feed the retargeted frames to a reconstruction
module that reverses the retargeted frames to the input frames. Therefore, the
network is trained in an unsupervised manner. Extensive qualitative and
quantitative experiments and ablation studies on KITTI stereo 2012 and 2015
datasets demonstrate the efficiency of the proposed method over the existing
state-of-the-art methods. The code is available at
https://github.com/z65451/SVR/.
|
2304.08876 | Chang Xu | Chang Xu, Jian Ding, Jinwang Wang, Wen Yang, Huai Yu, Lei Yu, Gui-Song
Xia | Dynamic Coarse-to-Fine Learning for Oriented Tiny Object Detection | Accepted by CVPR2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Detecting arbitrarily oriented tiny objects poses intense challenges to
existing detectors, especially for label assignment. Despite the exploration of
adaptive label assignment in recent oriented object detectors, the extreme
geometry shape and limited feature of oriented tiny objects still induce severe
mismatch and imbalance issues. Specifically, the position prior, positive
sample feature, and instance are mismatched, and the learning of extreme-shaped
objects is biased and unbalanced due to little proper feature supervision. To
tackle these issues, we propose a dynamic prior along with the coarse-to-fine
assigner, dubbed DCFL. For one thing, we model the prior, label assignment, and
object representation all in a dynamic manner to alleviate the mismatch issue.
For another, we leverage the coarse prior matching and finer posterior
constraint to dynamically assign labels, providing appropriate and relatively
balanced supervision for diverse instances. Extensive experiments on six
datasets show substantial improvements to the baseline. Notably, we obtain the
state-of-the-art performance for one-stage detectors on the DOTA-v1.5,
DOTA-v2.0, and DIOR-R datasets under single-scale training and testing. Codes
are available at https://github.com/Chasel-Tsui/mmrotate-dcfl.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 10:09:22 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Xu",
"Chang",
""
],
[
"Ding",
"Jian",
""
],
[
"Wang",
"Jinwang",
""
],
[
"Yang",
"Wen",
""
],
[
"Yu",
"Huai",
""
],
[
"Yu",
"Lei",
""
],
[
"Xia",
"Gui-Song",
""
]
] | TITLE: Dynamic Coarse-to-Fine Learning for Oriented Tiny Object Detection
ABSTRACT: Detecting arbitrarily oriented tiny objects poses intense challenges to
existing detectors, especially for label assignment. Despite the exploration of
adaptive label assignment in recent oriented object detectors, the extreme
geometry shape and limited feature of oriented tiny objects still induce severe
mismatch and imbalance issues. Specifically, the position prior, positive
sample feature, and instance are mismatched, and the learning of extreme-shaped
objects is biased and unbalanced due to little proper feature supervision. To
tackle these issues, we propose a dynamic prior along with the coarse-to-fine
assigner, dubbed DCFL. For one thing, we model the prior, label assignment, and
object representation all in a dynamic manner to alleviate the mismatch issue.
For another, we leverage the coarse prior matching and finer posterior
constraint to dynamically assign labels, providing appropriate and relatively
balanced supervision for diverse instances. Extensive experiments on six
datasets show substantial improvements to the baseline. Notably, we obtain the
state-of-the-art performance for one-stage detectors on the DOTA-v1.5,
DOTA-v2.0, and DIOR-R datasets under single-scale training and testing. Codes
are available at https://github.com/Chasel-Tsui/mmrotate-dcfl.
|
2304.08881 | Ragnhild Holden Helland | Ragnhild Holden Helland, Alexandros Ferles, Andr\'e Pedersen, Ivar
Kommers, Hilko Ardon, Frederik Barkhof, Lorenzo Bello, Mitchel S. Berger,
Tora Dun{\aa}s, Marco Conti Nibali, Julia Furtner, Shawn Hervey-Jumper,
Albert J. S. Idema, Barbara Kiesel, Rishi Nandoe Tewari, Emmanuel Mandonnet,
Domenique M.J. M\"uller, Pierre A. Robe, Marco Rossi, Lisa M. Sagberg,
Tommaso Sciortino, Tom Aalders, Michiel Wagemakers, Georg Widhalm, Marnix G.
Witte, Aeilko H. Zwinderman, Paulina L. Majewska, Asgeir S. Jakola, Ole
Solheim, Philip C. De Witt Hamer, Ingerid Reinertsen, Roelant S. Eijgelaar,
and David Bouget | Segmentation of glioblastomas in early post-operative multi-modal MRI
with deep neural networks | 13 pages, 4 figures, 4 tables | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by/4.0/ | Extent of resection after surgery is one of the main prognostic factors for
patients diagnosed with glioblastoma. To achieve this, accurate segmentation
and classification of residual tumor from post-operative MR images is
essential. The current standard method for estimating it is subject to high
inter- and intra-rater variability, and an automated method for segmentation of
residual tumor in early post-operative MRI could lead to a more accurate
estimation of extent of resection. In this study, two state-of-the-art neural
network architectures for pre-operative segmentation were trained for the task.
The models were extensively validated on a multicenter dataset with nearly 1000
patients, from 12 hospitals in Europe and the United States. The best
performance achieved was a 61\% Dice score, and the best classification
performance was about 80\% balanced accuracy, with a demonstrated ability to
generalize across hospitals. In addition, the segmentation performance of the
best models was on par with human expert raters. The predicted segmentations
can be used to accurately classify the patients into those with residual tumor,
and those with gross total resection.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 10:14:45 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Helland",
"Ragnhild Holden",
""
],
[
"Ferles",
"Alexandros",
""
],
[
"Pedersen",
"André",
""
],
[
"Kommers",
"Ivar",
""
],
[
"Ardon",
"Hilko",
""
],
[
"Barkhof",
"Frederik",
""
],
[
"Bello",
"Lorenzo",
""
],
[
"Berger",
"Mitchel S.",
""
],
[
"Dunås",
"Tora",
""
],
[
"Nibali",
"Marco Conti",
""
],
[
"Furtner",
"Julia",
""
],
[
"Hervey-Jumper",
"Shawn",
""
],
[
"Idema",
"Albert J. S.",
""
],
[
"Kiesel",
"Barbara",
""
],
[
"Tewari",
"Rishi Nandoe",
""
],
[
"Mandonnet",
"Emmanuel",
""
],
[
"Müller",
"Domenique M. J.",
""
],
[
"Robe",
"Pierre A.",
""
],
[
"Rossi",
"Marco",
""
],
[
"Sagberg",
"Lisa M.",
""
],
[
"Sciortino",
"Tommaso",
""
],
[
"Aalders",
"Tom",
""
],
[
"Wagemakers",
"Michiel",
""
],
[
"Widhalm",
"Georg",
""
],
[
"Witte",
"Marnix G.",
""
],
[
"Zwinderman",
"Aeilko H.",
""
],
[
"Majewska",
"Paulina L.",
""
],
[
"Jakola",
"Asgeir S.",
""
],
[
"Solheim",
"Ole",
""
],
[
"Hamer",
"Philip C. De Witt",
""
],
[
"Reinertsen",
"Ingerid",
""
],
[
"Eijgelaar",
"Roelant S.",
""
],
[
"Bouget",
"David",
""
]
] | TITLE: Segmentation of glioblastomas in early post-operative multi-modal MRI
with deep neural networks
ABSTRACT: Extent of resection after surgery is one of the main prognostic factors for
patients diagnosed with glioblastoma. To achieve this, accurate segmentation
and classification of residual tumor from post-operative MR images is
essential. The current standard method for estimating it is subject to high
inter- and intra-rater variability, and an automated method for segmentation of
residual tumor in early post-operative MRI could lead to a more accurate
estimation of extent of resection. In this study, two state-of-the-art neural
network architectures for pre-operative segmentation were trained for the task.
The models were extensively validated on a multicenter dataset with nearly 1000
patients, from 12 hospitals in Europe and the United States. The best
performance achieved was a 61\% Dice score, and the best classification
performance was about 80\% balanced accuracy, with a demonstrated ability to
generalize across hospitals. In addition, the segmentation performance of the
best models was on par with human expert raters. The predicted segmentations
can be used to accurately classify the patients into those with residual tumor,
and those with gross total resection.
|
2304.08894 | Zihan Wang | Zihan Wang and Gang wu and Haotong Wang | Discreetly Exploiting Inter-session Information for Session-based
Recommendation | null | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Limited intra-session information is the performance bottleneck of the early
GNN based SBR models. Therefore, some GNN based SBR models have evolved to
introduce additional inter-session information to facilitate the next-item
prediction. However, we found that the introduction of inter-session
information may bring interference to these models. The possible reasons are
twofold. First, inter-session dependencies are not differentiated at the
factor-level. Second, measuring inter-session weight by similarity is not
enough. In this paper, we propose DEISI to solve the problems. For the first
problem, DEISI differentiates the types of inter-session dependencies at the
factor-level with the help of DRL technology. For the second problem, DEISI
introduces stability as a new metric for weighting inter-session dependencies
together with the similarity. Moreover, CL is used to improve the robustness of
the model. Extensive experiments on three datasets show the superior
performance of the DEISI model compared with the state-of-the-art models.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 10:43:36 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Wang",
"Zihan",
""
],
[
"wu",
"Gang",
""
],
[
"Wang",
"Haotong",
""
]
] | TITLE: Discreetly Exploiting Inter-session Information for Session-based
Recommendation
ABSTRACT: Limited intra-session information is the performance bottleneck of the early
GNN based SBR models. Therefore, some GNN based SBR models have evolved to
introduce additional inter-session information to facilitate the next-item
prediction. However, we found that the introduction of inter-session
information may bring interference to these models. The possible reasons are
twofold. First, inter-session dependencies are not differentiated at the
factor-level. Second, measuring inter-session weight by similarity is not
enough. In this paper, we propose DEISI to solve the problems. For the first
problem, DEISI differentiates the types of inter-session dependencies at the
factor-level with the help of DRL technology. For the second problem, DEISI
introduces stability as a new metric for weighting inter-session dependencies
together with the similarity. Moreover, CL is used to improve the robustness of
the model. Extensive experiments on three datasets show the superior
performance of the DEISI model compared with the state-of-the-art models.
|
2304.08901 | Alpay Sabuncuo\u{g}lu | Alpay Sabuncuoglu, T. Metin Sezgin | Multimodal Group Activity Dataset for Classroom Engagement Level
Prediction | null | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | We collected a new dataset that includes approximately eight hours of
audiovisual recordings of a group of students and their self-evaluation scores
for classroom engagement. The dataset and data analysis scripts are available
on our open-source repository. We developed baseline face-based and
group-activity-based image and video recognition models. Our image models yield
45-85% test accuracy with face-area inputs on person-based classification task.
Our video models achieved up to 71% test accuracy on group-level prediction
using group activity video inputs. In this technical report, we shared the
details of our end-to-end human-centered engagement analysis pipeline from data
collection to model development.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 11:09:02 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Sabuncuoglu",
"Alpay",
""
],
[
"Sezgin",
"T. Metin",
""
]
] | TITLE: Multimodal Group Activity Dataset for Classroom Engagement Level
Prediction
ABSTRACT: We collected a new dataset that includes approximately eight hours of
audiovisual recordings of a group of students and their self-evaluation scores
for classroom engagement. The dataset and data analysis scripts are available
on our open-source repository. We developed baseline face-based and
group-activity-based image and video recognition models. Our image models yield
45-85% test accuracy with face-area inputs on person-based classification task.
Our video models achieved up to 71% test accuracy on group-level prediction
using group activity video inputs. In this technical report, we shared the
details of our end-to-end human-centered engagement analysis pipeline from data
collection to model development.
|
2304.08911 | Haolan Zhan | Haolan Zhan and Xuming Lin and Shaobo Cui and Zhongzhou Zhao and Wei
Zhou and Haiqing Chen | Towards Zero-Shot Personalized Table-to-Text Generation with Contrastive
Persona Distillation | Accepted by ICASSP 2023 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing neural methods have shown great potentials towards generating
informative text from structured tabular data as well as maintaining high
content fidelity. However, few of them shed light on generating personalized
expressions, which often requires well-aligned persona-table-text datasets that
are difficult to obtain. To overcome these obstacles, we explore personalized
table-to-text generation under a zero-shot setting, by assuming no well-aligned
persona-table-text triples are required during training. To this end, we
firstly collect a set of unpaired persona information and then propose a
semi-supervised approach with contrastive persona distillation (S2P-CPD) to
generate personalized context. Specifically, tabular data and persona
information are firstly represented as latent variables separately. Then, we
devise a latent space fusion technique to distill persona information into the
table representation. Besides, a contrastive-based discriminator is employed to
guarantee the style consistency between the generated context and its
corresponding persona. Experimental results on two benchmarks demonstrate
S2P-CPD's ability on keeping both content fidelity and personalized
expressions.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 11:32:33 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zhan",
"Haolan",
""
],
[
"Lin",
"Xuming",
""
],
[
"Cui",
"Shaobo",
""
],
[
"Zhao",
"Zhongzhou",
""
],
[
"Zhou",
"Wei",
""
],
[
"Chen",
"Haiqing",
""
]
] | TITLE: Towards Zero-Shot Personalized Table-to-Text Generation with Contrastive
Persona Distillation
ABSTRACT: Existing neural methods have shown great potentials towards generating
informative text from structured tabular data as well as maintaining high
content fidelity. However, few of them shed light on generating personalized
expressions, which often requires well-aligned persona-table-text datasets that
are difficult to obtain. To overcome these obstacles, we explore personalized
table-to-text generation under a zero-shot setting, by assuming no well-aligned
persona-table-text triples are required during training. To this end, we
firstly collect a set of unpaired persona information and then propose a
semi-supervised approach with contrastive persona distillation (S2P-CPD) to
generate personalized context. Specifically, tabular data and persona
information are firstly represented as latent variables separately. Then, we
devise a latent space fusion technique to distill persona information into the
table representation. Besides, a contrastive-based discriminator is employed to
guarantee the style consistency between the generated context and its
corresponding persona. Experimental results on two benchmarks demonstrate
S2P-CPD's ability on keeping both content fidelity and personalized
expressions.
|
2304.08912 | Yen-Chieh Lien | Yen-Chieh Lien, Hamed Zamani, W. Bruce Croft | Generalized Weak Supervision for Neural Information Retrieval | null | null | null | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Neural ranking models (NRMs) have demonstrated effective performance in
several information retrieval (IR) tasks. However, training NRMs often requires
large-scale training data, which is difficult and expensive to obtain. To
address this issue, one can train NRMs via weak supervision, where a large
dataset is automatically generated using an existing ranking model (called the
weak labeler) for training NRMs. Weakly supervised NRMs can generalize from the
observed data and significantly outperform the weak labeler. This paper
generalizes this idea through an iterative re-labeling process, demonstrating
that weakly supervised models can iteratively play the role of weak labeler and
significantly improve ranking performance without using manually labeled data.
The proposed Generalized Weak Supervision (GWS) solution is generic and
orthogonal to the ranking model architecture. This paper offers four
implementations of GWS: self-labeling, cross-labeling, joint cross- and
self-labeling, and greedy multi-labeling. GWS also benefits from a query
importance weighting mechanism based on query performance prediction methods to
reduce noise in the generated training data. We further draw a theoretical
connection between self-labeling and Expectation-Maximization. Our experiments
on two passage retrieval benchmarks suggest that all implementations of GWS
lead to substantial improvements compared to weak supervision in all cases.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 11:34:05 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Lien",
"Yen-Chieh",
""
],
[
"Zamani",
"Hamed",
""
],
[
"Croft",
"W. Bruce",
""
]
] | TITLE: Generalized Weak Supervision for Neural Information Retrieval
ABSTRACT: Neural ranking models (NRMs) have demonstrated effective performance in
several information retrieval (IR) tasks. However, training NRMs often requires
large-scale training data, which is difficult and expensive to obtain. To
address this issue, one can train NRMs via weak supervision, where a large
dataset is automatically generated using an existing ranking model (called the
weak labeler) for training NRMs. Weakly supervised NRMs can generalize from the
observed data and significantly outperform the weak labeler. This paper
generalizes this idea through an iterative re-labeling process, demonstrating
that weakly supervised models can iteratively play the role of weak labeler and
significantly improve ranking performance without using manually labeled data.
The proposed Generalized Weak Supervision (GWS) solution is generic and
orthogonal to the ranking model architecture. This paper offers four
implementations of GWS: self-labeling, cross-labeling, joint cross- and
self-labeling, and greedy multi-labeling. GWS also benefits from a query
importance weighting mechanism based on query performance prediction methods to
reduce noise in the generated training data. We further draw a theoretical
connection between self-labeling and Expectation-Maximization. Our experiments
on two passage retrieval benchmarks suggest that all implementations of GWS
lead to substantial improvements compared to weak supervision in all cases.
|
2304.08913 | Jakub Kudela | Jakub Kudela, Martin Juricek | Computational and Exploratory Landscape Analysis of the GKLS Generator | null | null | null | null | cs.NE cs.AI math.OC | http://creativecommons.org/licenses/by/4.0/ | The GKLS generator is one of the most used testbeds for benchmarking global
optimization algorithms. In this paper, we conduct both a computational
analysis and the Exploratory Landscape Analysis (ELA) of the GKLS generator. We
utilize both canonically used and newly generated classes of GKLS-generated
problems and show their use in benchmarking three state-of-the-art methods
(from evolutionary and deterministic communities) in dimensions 5 and 10. We
show that the GKLS generator produces ``needle in a haystack'' type problems
that become extremely difficult to optimize in higher dimensions. Furthermore,
we conduct the ELA on the GKLS generator and then compare it to the ELA of two
other widely used benchmark sets (BBOB and CEC 2014), and discuss the
meaningfulness of the results.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 11:34:56 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Kudela",
"Jakub",
""
],
[
"Juricek",
"Martin",
""
]
] | TITLE: Computational and Exploratory Landscape Analysis of the GKLS Generator
ABSTRACT: The GKLS generator is one of the most used testbeds for benchmarking global
optimization algorithms. In this paper, we conduct both a computational
analysis and the Exploratory Landscape Analysis (ELA) of the GKLS generator. We
utilize both canonically used and newly generated classes of GKLS-generated
problems and show their use in benchmarking three state-of-the-art methods
(from evolutionary and deterministic communities) in dimensions 5 and 10. We
show that the GKLS generator produces ``needle in a haystack'' type problems
that become extremely difficult to optimize in higher dimensions. Furthermore,
we conduct the ELA on the GKLS generator and then compare it to the ELA of two
other widely used benchmark sets (BBOB and CEC 2014), and discuss the
meaningfulness of the results.
|
2304.08915 | Peng Zeng | Peng Zeng, Xiaotian Song, Andrew Lensen, Yuwei Ou, Yanan Sun, Mengjie
Zhang, Jiancheng Lv | Differentiable Genetic Programming for High-dimensional Symbolic
Regression | null | null | null | null | cs.NE cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Symbolic regression (SR) is the process of discovering hidden relationships
from data with mathematical expressions, which is considered an effective way
to reach interpretable machine learning (ML). Genetic programming (GP) has been
the dominator in solving SR problems. However, as the scale of SR problems
increases, GP often poorly demonstrates and cannot effectively address the
real-world high-dimensional problems. This limitation is mainly caused by the
stochastic evolutionary nature of traditional GP in constructing the trees. In
this paper, we propose a differentiable approach named DGP to construct GP
trees towards high-dimensional SR for the first time. Specifically, a new data
structure called differentiable symbolic tree is proposed to relax the discrete
structure to be continuous, thus a gradient-based optimizer can be presented
for the efficient optimization. In addition, a sampling method is proposed to
eliminate the discrepancy caused by the above relaxation for valid symbolic
expressions. Furthermore, a diversification mechanism is introduced to promote
the optimizer escaping from local optima for globally better solutions. With
these designs, the proposed DGP method can efficiently search for the GP trees
with higher performance, thus being capable of dealing with high-dimensional
SR. To demonstrate the effectiveness of DGP, we conducted various experiments
against the state of the arts based on both GP and deep neural networks. The
experiment results reveal that DGP can outperform these chosen peer competitors
on high-dimensional regression benchmarks with dimensions varying from tens to
thousands. In addition, on the synthetic SR problems, the proposed DGP method
can also achieve the best recovery rate even with different noisy levels. It is
believed this work can facilitate SR being a powerful alternative to
interpretable ML for a broader range of real-world problems.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 11:39:45 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Zeng",
"Peng",
""
],
[
"Song",
"Xiaotian",
""
],
[
"Lensen",
"Andrew",
""
],
[
"Ou",
"Yuwei",
""
],
[
"Sun",
"Yanan",
""
],
[
"Zhang",
"Mengjie",
""
],
[
"Lv",
"Jiancheng",
""
]
] | TITLE: Differentiable Genetic Programming for High-dimensional Symbolic
Regression
ABSTRACT: Symbolic regression (SR) is the process of discovering hidden relationships
from data with mathematical expressions, which is considered an effective way
to reach interpretable machine learning (ML). Genetic programming (GP) has been
the dominator in solving SR problems. However, as the scale of SR problems
increases, GP often poorly demonstrates and cannot effectively address the
real-world high-dimensional problems. This limitation is mainly caused by the
stochastic evolutionary nature of traditional GP in constructing the trees. In
this paper, we propose a differentiable approach named DGP to construct GP
trees towards high-dimensional SR for the first time. Specifically, a new data
structure called differentiable symbolic tree is proposed to relax the discrete
structure to be continuous, thus a gradient-based optimizer can be presented
for the efficient optimization. In addition, a sampling method is proposed to
eliminate the discrepancy caused by the above relaxation for valid symbolic
expressions. Furthermore, a diversification mechanism is introduced to promote
the optimizer escaping from local optima for globally better solutions. With
these designs, the proposed DGP method can efficiently search for the GP trees
with higher performance, thus being capable of dealing with high-dimensional
SR. To demonstrate the effectiveness of DGP, we conducted various experiments
against the state of the arts based on both GP and deep neural networks. The
experiment results reveal that DGP can outperform these chosen peer competitors
on high-dimensional regression benchmarks with dimensions varying from tens to
thousands. In addition, on the synthetic SR problems, the proposed DGP method
can also achieve the best recovery rate even with different noisy levels. It is
believed this work can facilitate SR being a powerful alternative to
interpretable ML for a broader range of real-world problems.
|
2304.08989 | Xiaolei Diao | Fausto Giunchiglia, Xiaolei Diao, Mayukh Bagchi | Incremental Image Labeling via Iterative Refinement | null | IWCIM@ICASSP 2023 | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data quality is critical for multimedia tasks, while various types of
systematic flaws are found in image benchmark datasets, as discussed in recent
work. In particular, the existence of the semantic gap problem leads to a
many-to-many mapping between the information extracted from an image and its
linguistic description. This unavoidable bias further leads to poor performance
on current computer vision tasks. To address this issue, we introduce a
Knowledge Representation (KR)-based methodology to provide guidelines driving
the labeling process, thereby indirectly introducing intended semantics in ML
models. Specifically, an iterative refinement-based annotation method is
proposed to optimize data labeling by organizing objects in a classification
hierarchy according to their visual properties, ensuring that they are aligned
with their linguistic descriptions. Preliminary results verify the
effectiveness of the proposed method.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 13:37:22 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Giunchiglia",
"Fausto",
""
],
[
"Diao",
"Xiaolei",
""
],
[
"Bagchi",
"Mayukh",
""
]
] | TITLE: Incremental Image Labeling via Iterative Refinement
ABSTRACT: Data quality is critical for multimedia tasks, while various types of
systematic flaws are found in image benchmark datasets, as discussed in recent
work. In particular, the existence of the semantic gap problem leads to a
many-to-many mapping between the information extracted from an image and its
linguistic description. This unavoidable bias further leads to poor performance
on current computer vision tasks. To address this issue, we introduce a
Knowledge Representation (KR)-based methodology to provide guidelines driving
the labeling process, thereby indirectly introducing intended semantics in ML
models. Specifically, an iterative refinement-based annotation method is
proposed to optimize data labeling by organizing objects in a classification
hierarchy according to their visual properties, ensuring that they are aligned
with their linguistic descriptions. Preliminary results verify the
effectiveness of the proposed method.
|
2304.08991 | Hyunjae Lee | Hyunjae Lee | D2CSE: Difference-aware Deep continuous prompts for Contrastive Sentence
Embeddings | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper describes Difference-aware Deep continuous prompt for Contrastive
Sentence Embeddings (D2CSE) that learns sentence embeddings. Compared to
state-of-the-art approaches, D2CSE computes sentence vectors that are
exceptional to distinguish a subtle difference in similar sentences by
employing a simple neural architecture for continuous prompts. Unlike existing
architectures that require multiple pretrained language models (PLMs) to
process a pair of the original and corrupted (subtly modified) sentences, D2CSE
avoids cumbersome fine-tuning of multiple PLMs by only optimizing continuous
prompts by performing multiple tasks -- i.e., contrastive learning and
conditional replaced token detection all done in a self-guided manner. D2CSE
overloads a single PLM on continuous prompts and greatly saves memory
consumption as a result. The number of training parameters in D2CSE is reduced
to about 1\% of existing approaches while substantially improving the quality
of sentence embeddings. We evaluate D2CSE on seven Semantic Textual Similarity
(STS) benchmarks, using three different metrics, namely, Spearman's rank
correlation, recall@K for a retrieval task, and the anisotropy of an embedding
space measured in alignment and uniformity. Our empirical results suggest that
shallow (not too meticulously devised) continuous prompts can be honed
effectively for multiple NLP tasks and lead to improvements upon existing
state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 13:45:07 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Lee",
"Hyunjae",
""
]
] | TITLE: D2CSE: Difference-aware Deep continuous prompts for Contrastive Sentence
Embeddings
ABSTRACT: This paper describes Difference-aware Deep continuous prompt for Contrastive
Sentence Embeddings (D2CSE) that learns sentence embeddings. Compared to
state-of-the-art approaches, D2CSE computes sentence vectors that are
exceptional to distinguish a subtle difference in similar sentences by
employing a simple neural architecture for continuous prompts. Unlike existing
architectures that require multiple pretrained language models (PLMs) to
process a pair of the original and corrupted (subtly modified) sentences, D2CSE
avoids cumbersome fine-tuning of multiple PLMs by only optimizing continuous
prompts by performing multiple tasks -- i.e., contrastive learning and
conditional replaced token detection all done in a self-guided manner. D2CSE
overloads a single PLM on continuous prompts and greatly saves memory
consumption as a result. The number of training parameters in D2CSE is reduced
to about 1\% of existing approaches while substantially improving the quality
of sentence embeddings. We evaluate D2CSE on seven Semantic Textual Similarity
(STS) benchmarks, using three different metrics, namely, Spearman's rank
correlation, recall@K for a retrieval task, and the anisotropy of an embedding
space measured in alignment and uniformity. Our empirical results suggest that
shallow (not too meticulously devised) continuous prompts can be honed
effectively for multiple NLP tasks and lead to improvements upon existing
state-of-the-art approaches.
|
2304.08993 | Rui Li | Rui Li, Dong Gong, Wei Yin, Hao Chen, Yu Zhu, Kaixuan Wang, Xiaozhi
Chen, Jinqiu Sun, Yanning Zhang | Learning to Fuse Monocular and Multi-view Cues for Multi-frame Depth
Estimation in Dynamic Scenes | Accepted by CVPR 2023. Code and models are available at:
https://github.com/ruili3/dynamic-multiframe-depth | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-frame depth estimation generally achieves high accuracy relying on the
multi-view geometric consistency. When applied in dynamic scenes, e.g.,
autonomous driving, this consistency is usually violated in the dynamic areas,
leading to corrupted estimations. Many multi-frame methods handle dynamic areas
by identifying them with explicit masks and compensating the multi-view cues
with monocular cues represented as local monocular depth or features. The
improvements are limited due to the uncontrolled quality of the masks and the
underutilized benefits of the fusion of the two types of cues. In this paper,
we propose a novel method to learn to fuse the multi-view and monocular cues
encoded as volumes without needing the heuristically crafted masks. As unveiled
in our analyses, the multi-view cues capture more accurate geometric
information in static areas, and the monocular cues capture more useful
contexts in dynamic areas. To let the geometric perception learned from
multi-view cues in static areas propagate to the monocular representation in
dynamic areas and let monocular cues enhance the representation of multi-view
cost volume, we propose a cross-cue fusion (CCF) module, which includes the
cross-cue attention (CCA) to encode the spatially non-local relative
intra-relations from each source to enhance the representation of the other.
Experiments on real-world datasets prove the significant effectiveness and
generalization ability of the proposed method.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 13:55:24 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Li",
"Rui",
""
],
[
"Gong",
"Dong",
""
],
[
"Yin",
"Wei",
""
],
[
"Chen",
"Hao",
""
],
[
"Zhu",
"Yu",
""
],
[
"Wang",
"Kaixuan",
""
],
[
"Chen",
"Xiaozhi",
""
],
[
"Sun",
"Jinqiu",
""
],
[
"Zhang",
"Yanning",
""
]
] | TITLE: Learning to Fuse Monocular and Multi-view Cues for Multi-frame Depth
Estimation in Dynamic Scenes
ABSTRACT: Multi-frame depth estimation generally achieves high accuracy relying on the
multi-view geometric consistency. When applied in dynamic scenes, e.g.,
autonomous driving, this consistency is usually violated in the dynamic areas,
leading to corrupted estimations. Many multi-frame methods handle dynamic areas
by identifying them with explicit masks and compensating the multi-view cues
with monocular cues represented as local monocular depth or features. The
improvements are limited due to the uncontrolled quality of the masks and the
underutilized benefits of the fusion of the two types of cues. In this paper,
we propose a novel method to learn to fuse the multi-view and monocular cues
encoded as volumes without needing the heuristically crafted masks. As unveiled
in our analyses, the multi-view cues capture more accurate geometric
information in static areas, and the monocular cues capture more useful
contexts in dynamic areas. To let the geometric perception learned from
multi-view cues in static areas propagate to the monocular representation in
dynamic areas and let monocular cues enhance the representation of multi-view
cost volume, we propose a cross-cue fusion (CCF) module, which includes the
cross-cue attention (CCA) to encode the spatially non-local relative
intra-relations from each source to enhance the representation of the other.
Experiments on real-world datasets prove the significant effectiveness and
generalization ability of the proposed method.
|
2304.08994 | Alexander Naumann | Alexander Naumann, Felix Hertlein, Laura D\"orr, Kai Furmans | Parcel3D: Shape Reconstruction from Single RGB Images for Applications
in Transportation Logistics | Accepted at CVPR workshop on Vision-based InduStrial InspectiON
(VISION) 2023, see
https://vision-based-industrial-inspection.github.io/cvpr-2023/ | null | null | null | cs.CV cs.LG cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We focus on enabling damage and tampering detection in logistics and tackle
the problem of 3D shape reconstruction of potentially damaged parcels. As input
we utilize single RGB images, which corresponds to use-cases where only simple
handheld devices are available, e.g. for postmen during delivery or clients on
delivery. We present a novel synthetic dataset, named Parcel3D, that is based
on the Google Scanned Objects (GSO) dataset and consists of more than 13,000
images of parcels with full 3D annotations. The dataset contains intact, i.e.
cuboid-shaped, parcels and damaged parcels, which were generated in
simulations. We work towards detecting mishandling of parcels by presenting a
novel architecture called CubeRefine R-CNN, which combines estimating a 3D
bounding box with an iterative mesh refinement. We benchmark our approach on
Parcel3D and an existing dataset of cuboid-shaped parcels in real-world
scenarios. Our results show, that while training on Parcel3D enables transfer
to the real world, enabling reliable deployment in real-world scenarios is
still challenging. CubeRefine R-CNN yields competitive performance in terms of
Mesh AP and is the only model that directly enables deformation assessment by
3D mesh comparison and tampering detection by comparing viewpoint invariant
parcel side surface representations. Dataset and code are available at
https://a-nau.github.io/parcel3d.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 13:55:51 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Naumann",
"Alexander",
""
],
[
"Hertlein",
"Felix",
""
],
[
"Dörr",
"Laura",
""
],
[
"Furmans",
"Kai",
""
]
] | TITLE: Parcel3D: Shape Reconstruction from Single RGB Images for Applications
in Transportation Logistics
ABSTRACT: We focus on enabling damage and tampering detection in logistics and tackle
the problem of 3D shape reconstruction of potentially damaged parcels. As input
we utilize single RGB images, which corresponds to use-cases where only simple
handheld devices are available, e.g. for postmen during delivery or clients on
delivery. We present a novel synthetic dataset, named Parcel3D, that is based
on the Google Scanned Objects (GSO) dataset and consists of more than 13,000
images of parcels with full 3D annotations. The dataset contains intact, i.e.
cuboid-shaped, parcels and damaged parcels, which were generated in
simulations. We work towards detecting mishandling of parcels by presenting a
novel architecture called CubeRefine R-CNN, which combines estimating a 3D
bounding box with an iterative mesh refinement. We benchmark our approach on
Parcel3D and an existing dataset of cuboid-shaped parcels in real-world
scenarios. Our results show, that while training on Parcel3D enables transfer
to the real world, enabling reliable deployment in real-world scenarios is
still challenging. CubeRefine R-CNN yields competitive performance in terms of
Mesh AP and is the only model that directly enables deformation assessment by
3D mesh comparison and tampering detection by comparing viewpoint invariant
parcel side surface representations. Dataset and code are available at
https://a-nau.github.io/parcel3d.
|
2304.09012 | Andrey Sobolevsky | Andrey Sobolevsky, Guillaume-Alexandre Bilodeau, Jinghui Cheng, Jin
L.C. Guo | GUILGET: GUI Layout GEneration with Transformer | 12 pages, 5 figures, Canadian AI Conference 2023 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Sketching out Graphical User Interface (GUI) layout is part of the pipeline
of designing a GUI and a crucial task for the success of a software
application. Arranging all components inside a GUI layout manually is a
time-consuming task. In order to assist designers, we developed a method named
GUILGET to automatically generate GUI layouts from positional constraints
represented as GUI arrangement graphs (GUI-AGs). The goal is to support the
initial step of GUI design by producing realistic and diverse GUI layouts. The
existing image layout generation techniques often cannot incorporate GUI design
constraints. Thus, GUILGET needs to adapt existing techniques to generate GUI
layouts that obey to constraints specific to GUI designs. GUILGET is based on
transformers in order to capture the semantic in relationships between elements
from GUI-AG. Moreover, the model learns constraints through the minimization of
losses responsible for placing each component inside its parent layout, for not
letting components overlap if they are inside the same parent, and for
component alignment. Our experiments, which are conducted on the CLAY dataset,
reveal that our model has the best understanding of relationships from GUI-AG
and has the best performances in most of evaluation metrics. Therefore, our
work contributes to improved GUI layout generation by proposing a novel method
that effectively accounts for the constraints on GUI elements and paves the
road for a more efficient GUI design pipeline.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 14:27:34 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Sobolevsky",
"Andrey",
""
],
[
"Bilodeau",
"Guillaume-Alexandre",
""
],
[
"Cheng",
"Jinghui",
""
],
[
"Guo",
"Jin L. C.",
""
]
] | TITLE: GUILGET: GUI Layout GEneration with Transformer
ABSTRACT: Sketching out Graphical User Interface (GUI) layout is part of the pipeline
of designing a GUI and a crucial task for the success of a software
application. Arranging all components inside a GUI layout manually is a
time-consuming task. In order to assist designers, we developed a method named
GUILGET to automatically generate GUI layouts from positional constraints
represented as GUI arrangement graphs (GUI-AGs). The goal is to support the
initial step of GUI design by producing realistic and diverse GUI layouts. The
existing image layout generation techniques often cannot incorporate GUI design
constraints. Thus, GUILGET needs to adapt existing techniques to generate GUI
layouts that obey to constraints specific to GUI designs. GUILGET is based on
transformers in order to capture the semantic in relationships between elements
from GUI-AG. Moreover, the model learns constraints through the minimization of
losses responsible for placing each component inside its parent layout, for not
letting components overlap if they are inside the same parent, and for
component alignment. Our experiments, which are conducted on the CLAY dataset,
reveal that our model has the best understanding of relationships from GUI-AG
and has the best performances in most of evaluation metrics. Therefore, our
work contributes to improved GUI layout generation by proposing a novel method
that effectively accounts for the constraints on GUI elements and paves the
road for a more efficient GUI design pipeline.
|
2304.09061 | Guillaume Salha-Galvan | Walid Bendada and Guillaume Salha-Galvan and Thomas Bouab\c{c}a and
Tristan Cazenave | A Scalable Framework for Automatic Playlist Continuation on Music
Streaming Services | Accepted as a Full Paper at the SIGIR 2023 conference | null | 10.1145/3539618.3591628 | null | cs.IR cs.LG cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Music streaming services often aim to recommend songs for users to extend the
playlists they have created on these services. However, extending playlists
while preserving their musical characteristics and matching user preferences
remains a challenging task, commonly referred to as Automatic Playlist
Continuation (APC). Besides, while these services often need to select the best
songs to recommend in real-time and among large catalogs with millions of
candidates, recent research on APC mainly focused on models with few
scalability guarantees and evaluated on relatively small datasets. In this
paper, we introduce a general framework to build scalable yet effective APC
models for large-scale applications. Based on a represent-then-aggregate
strategy, it ensures scalability by design while remaining flexible enough to
incorporate a wide range of representation learning and sequence modeling
techniques, e.g., based on Transformers. We demonstrate the relevance of this
framework through in-depth experimental validation on Spotify's Million
Playlist Dataset (MPD), the largest public dataset for APC. We also describe
how, in 2022, we successfully leveraged this framework to improve APC in
production on Deezer. We report results from a large-scale online A/B test on
this service, emphasizing the practical impact of our approach in such a
real-world application.
| [
{
"version": "v1",
"created": "Wed, 12 Apr 2023 08:46:04 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Bendada",
"Walid",
""
],
[
"Salha-Galvan",
"Guillaume",
""
],
[
"Bouabça",
"Thomas",
""
],
[
"Cazenave",
"Tristan",
""
]
] | TITLE: A Scalable Framework for Automatic Playlist Continuation on Music
Streaming Services
ABSTRACT: Music streaming services often aim to recommend songs for users to extend the
playlists they have created on these services. However, extending playlists
while preserving their musical characteristics and matching user preferences
remains a challenging task, commonly referred to as Automatic Playlist
Continuation (APC). Besides, while these services often need to select the best
songs to recommend in real-time and among large catalogs with millions of
candidates, recent research on APC mainly focused on models with few
scalability guarantees and evaluated on relatively small datasets. In this
paper, we introduce a general framework to build scalable yet effective APC
models for large-scale applications. Based on a represent-then-aggregate
strategy, it ensures scalability by design while remaining flexible enough to
incorporate a wide range of representation learning and sequence modeling
techniques, e.g., based on Transformers. We demonstrate the relevance of this
framework through in-depth experimental validation on Spotify's Million
Playlist Dataset (MPD), the largest public dataset for APC. We also describe
how, in 2022, we successfully leveraged this framework to improve APC in
production on Deezer. We report results from a large-scale online A/B test on
this service, emphasizing the practical impact of our approach in such a
real-world application.
|
2304.09068 | Sainyam Galhotra | Sainyam Galhotra and Yue Gong and Raul Castro Fernandez | METAM: Goal-Oriented Data Discovery | ICDE 2023 paper | null | null | null | cs.DB cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Data is a central component of machine learning and causal inference tasks.
The availability of large amounts of data from sources such as open data
repositories, data lakes and data marketplaces creates an opportunity to
augment data and boost those tasks' performance. However, augmentation
techniques rely on a user manually discovering and shortlisting useful
candidate augmentations. Existing solutions do not leverage the synergy between
discovery and augmentation, thus under exploiting data.
In this paper, we introduce METAM, a novel goal-oriented framework that
queries the downstream task with a candidate dataset, forming a feedback loop
that automatically steers the discovery and augmentation process. To select
candidates efficiently, METAM leverages properties of the: i) data, ii) utility
function, and iii) solution set size. We show METAM's theoretical guarantees
and demonstrate those empirically on a broad set of tasks. All in all, we
demonstrate the promise of goal-oriented data discovery to modern data science
applications.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 15:42:25 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Galhotra",
"Sainyam",
""
],
[
"Gong",
"Yue",
""
],
[
"Fernandez",
"Raul Castro",
""
]
] | TITLE: METAM: Goal-Oriented Data Discovery
ABSTRACT: Data is a central component of machine learning and causal inference tasks.
The availability of large amounts of data from sources such as open data
repositories, data lakes and data marketplaces creates an opportunity to
augment data and boost those tasks' performance. However, augmentation
techniques rely on a user manually discovering and shortlisting useful
candidate augmentations. Existing solutions do not leverage the synergy between
discovery and augmentation, thus under exploiting data.
In this paper, we introduce METAM, a novel goal-oriented framework that
queries the downstream task with a candidate dataset, forming a feedback loop
that automatically steers the discovery and augmentation process. To select
candidates efficiently, METAM leverages properties of the: i) data, ii) utility
function, and iii) solution set size. We show METAM's theoretical guarantees
and demonstrate those empirically on a broad set of tasks. All in all, we
demonstrate the promise of goal-oriented data discovery to modern data science
applications.
|
2304.09082 | Michael Robinson | Michael Robinson, Tate Altman, Denley Lam, Letitia W. Li | Unsupervised clustering of file dialects according to monotonic
decompositions of mixtures | null | null | null | null | cs.PL cs.CL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes an unsupervised classification method that partitions a
set of files into non-overlapping dialects based upon their behaviors,
determined by messages produced by a collection of programs that consume them.
The pattern of messages can be used as the signature of a particular kind of
behavior, with the understanding that some messages are likely to co-occur,
while others are not. Patterns of messages can be used to classify files into
dialects. A dialect is defined by a subset of messages, called the required
messages. Once files are conditioned upon dialect and its required messages,
the remaining messages are statistically independent.
With this definition of dialect in hand, we present a greedy algorithm that
deduces candidate dialects from a dataset consisting of a matrix of
file-message data, demonstrate its performance on several file formats, and
prove conditions under which it is optimal. We show that an analyst needs to
consider fewer dialects than distinct message patterns, which reduces their
cognitive load when studying a complex format.
| [
{
"version": "v1",
"created": "Thu, 9 Feb 2023 21:15:36 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Robinson",
"Michael",
""
],
[
"Altman",
"Tate",
""
],
[
"Lam",
"Denley",
""
],
[
"Li",
"Letitia W.",
""
]
] | TITLE: Unsupervised clustering of file dialects according to monotonic
decompositions of mixtures
ABSTRACT: This paper proposes an unsupervised classification method that partitions a
set of files into non-overlapping dialects based upon their behaviors,
determined by messages produced by a collection of programs that consume them.
The pattern of messages can be used as the signature of a particular kind of
behavior, with the understanding that some messages are likely to co-occur,
while others are not. Patterns of messages can be used to classify files into
dialects. A dialect is defined by a subset of messages, called the required
messages. Once files are conditioned upon dialect and its required messages,
the remaining messages are statistically independent.
With this definition of dialect in hand, we present a greedy algorithm that
deduces candidate dialects from a dataset consisting of a matrix of
file-message data, demonstrate its performance on several file formats, and
prove conditions under which it is optimal. We show that an analyst needs to
consider fewer dialects than distinct message patterns, which reduces their
cognitive load when studying a complex format.
|
2304.09085 | Haoxuan Li | Haoxuan Li, Yanghao Xiao, Chunyuan Zheng, Peng Wu | Balancing Unobserved Confounding with a Few Unbiased Ratings in Debiased
Recommendations | Accepted Paper in WWW'23 | Proceedings of the ACM Web Conference 2023 (WWW '23), April 30-May
4, 2023, Austin, TX, USA | 10.1145/3543507.3583495 | null | cs.IR cs.LG | http://creativecommons.org/publicdomain/zero/1.0/ | Recommender systems are seen as an effective tool to address information
overload, but it is widely known that the presence of various biases makes
direct training on large-scale observational data result in sub-optimal
prediction performance. In contrast, unbiased ratings obtained from randomized
controlled trials or A/B tests are considered to be the golden standard, but
are costly and small in scale in reality. To exploit both types of data, recent
works proposed to use unbiased ratings to correct the parameters of the
propensity or imputation models trained on the biased dataset. However, the
existing methods fail to obtain accurate predictions in the presence of
unobserved confounding or model misspecification. In this paper, we propose a
theoretically guaranteed model-agnostic balancing approach that can be applied
to any existing debiasing method with the aim of combating unobserved
confounding and model misspecification. The proposed approach makes full use of
unbiased data by alternatively correcting model parameters learned with biased
data, and adaptively learning balance coefficients of biased samples for
further debiasing. Extensive real-world experiments are conducted along with
the deployment of our proposal on four representative debiasing methods to
demonstrate the effectiveness.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 08:56:55 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Li",
"Haoxuan",
""
],
[
"Xiao",
"Yanghao",
""
],
[
"Zheng",
"Chunyuan",
""
],
[
"Wu",
"Peng",
""
]
] | TITLE: Balancing Unobserved Confounding with a Few Unbiased Ratings in Debiased
Recommendations
ABSTRACT: Recommender systems are seen as an effective tool to address information
overload, but it is widely known that the presence of various biases makes
direct training on large-scale observational data result in sub-optimal
prediction performance. In contrast, unbiased ratings obtained from randomized
controlled trials or A/B tests are considered to be the golden standard, but
are costly and small in scale in reality. To exploit both types of data, recent
works proposed to use unbiased ratings to correct the parameters of the
propensity or imputation models trained on the biased dataset. However, the
existing methods fail to obtain accurate predictions in the presence of
unobserved confounding or model misspecification. In this paper, we propose a
theoretically guaranteed model-agnostic balancing approach that can be applied
to any existing debiasing method with the aim of combating unobserved
confounding and model misspecification. The proposed approach makes full use of
unbiased data by alternatively correcting model parameters learned with biased
data, and adaptively learning balance coefficients of biased samples for
further debiasing. Extensive real-world experiments are conducted along with
the deployment of our proposal on four representative debiasing methods to
demonstrate the effectiveness.
|
2304.09093 | Huy Dao | Huy Dao, Dung D. Le, Cuong Chu | Improving Items and Contexts Understanding with Descriptive Graph for
Conversational Recommendation | 14 pages, 3 figures, 9 tables | null | null | null | cs.IR cs.CL cs.LG | http://creativecommons.org/licenses/by/4.0/ | State-of-the-art methods on conversational recommender systems (CRS) leverage
external knowledge to enhance both items' and contextual words' representations
to achieve high quality recommendations and responses generation. However, the
representations of the items and words are usually modeled in two separated
semantic spaces, which leads to misalignment issue between them. Consequently,
this will cause the CRS to only achieve a sub-optimal ranking performance,
especially when there is a lack of sufficient information from the user's
input. To address limitations of previous works, we propose a new CRS framework
KLEVER, which jointly models items and their associated contextual words in the
same semantic space. Particularly, we construct an item descriptive graph from
the rich items' textual features, such as item description and categories.
Based on the constructed descriptive graph, KLEVER jointly learns the
embeddings of the words and items, towards enhancing both recommender and
dialog generation modules. Extensive experiments on benchmarking CRS dataset
demonstrate that KLEVER achieves superior performance, especially when the
information from the users' responses is lacking.
| [
{
"version": "v1",
"created": "Tue, 11 Apr 2023 21:21:46 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Dao",
"Huy",
""
],
[
"Le",
"Dung D.",
""
],
[
"Chu",
"Cuong",
""
]
] | TITLE: Improving Items and Contexts Understanding with Descriptive Graph for
Conversational Recommendation
ABSTRACT: State-of-the-art methods on conversational recommender systems (CRS) leverage
external knowledge to enhance both items' and contextual words' representations
to achieve high quality recommendations and responses generation. However, the
representations of the items and words are usually modeled in two separated
semantic spaces, which leads to misalignment issue between them. Consequently,
this will cause the CRS to only achieve a sub-optimal ranking performance,
especially when there is a lack of sufficient information from the user's
input. To address limitations of previous works, we propose a new CRS framework
KLEVER, which jointly models items and their associated contextual words in the
same semantic space. Particularly, we construct an item descriptive graph from
the rich items' textual features, such as item description and categories.
Based on the constructed descriptive graph, KLEVER jointly learns the
embeddings of the words and items, towards enhancing both recommender and
dialog generation modules. Extensive experiments on benchmarking CRS dataset
demonstrate that KLEVER achieves superior performance, especially when the
information from the users' responses is lacking.
|
2304.09101 | Di Hong | Di Hong, Jiangrong Shen, Yu Qi, Yueming Wang | LaSNN: Layer-wise ANN-to-SNN Distillation for Effective and Efficient
Training in Deep Spiking Neural Networks | null | null | null | null | cs.NE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Spiking Neural Networks (SNNs) are biologically realistic and practically
promising in low-power computation because of their event-driven mechanism.
Usually, the training of SNNs suffers accuracy loss on various tasks, yielding
an inferior performance compared with ANNs. A conversion scheme is proposed to
obtain competitive accuracy by mapping trained ANNs' parameters to SNNs with
the same structures. However, an enormous number of time steps are required for
these converted SNNs, thus losing the energy-efficient benefit. Utilizing both
the accuracy advantages of ANNs and the computing efficiency of SNNs, a novel
SNN training framework is proposed, namely layer-wise ANN-to-SNN knowledge
distillation (LaSNN). In order to achieve competitive accuracy and reduced
inference latency, LaSNN transfers the learning from a well-trained ANN to a
small SNN by distilling the knowledge other than converting the parameters of
ANN. The information gap between heterogeneous ANN and SNN is bridged by
introducing the attention scheme, the knowledge in an ANN is effectively
compressed and then efficiently transferred by utilizing our layer-wise
distillation paradigm. We conduct detailed experiments to demonstrate the
effectiveness, efficacy, and scalability of LaSNN on three benchmark data sets
(CIFAR-10, CIFAR-100, and Tiny ImageNet). We achieve competitive top-1 accuracy
compared to ANNs and 20x faster inference than converted SNNs with similar
performance. More importantly, LaSNN is dexterous and extensible that can be
effortlessly developed for SNNs with different architectures/depths and input
encoding methods, contributing to their potential development.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 03:49:35 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Hong",
"Di",
""
],
[
"Shen",
"Jiangrong",
""
],
[
"Qi",
"Yu",
""
],
[
"Wang",
"Yueming",
""
]
] | TITLE: LaSNN: Layer-wise ANN-to-SNN Distillation for Effective and Efficient
Training in Deep Spiking Neural Networks
ABSTRACT: Spiking Neural Networks (SNNs) are biologically realistic and practically
promising in low-power computation because of their event-driven mechanism.
Usually, the training of SNNs suffers accuracy loss on various tasks, yielding
an inferior performance compared with ANNs. A conversion scheme is proposed to
obtain competitive accuracy by mapping trained ANNs' parameters to SNNs with
the same structures. However, an enormous number of time steps are required for
these converted SNNs, thus losing the energy-efficient benefit. Utilizing both
the accuracy advantages of ANNs and the computing efficiency of SNNs, a novel
SNN training framework is proposed, namely layer-wise ANN-to-SNN knowledge
distillation (LaSNN). In order to achieve competitive accuracy and reduced
inference latency, LaSNN transfers the learning from a well-trained ANN to a
small SNN by distilling the knowledge other than converting the parameters of
ANN. The information gap between heterogeneous ANN and SNN is bridged by
introducing the attention scheme, the knowledge in an ANN is effectively
compressed and then efficiently transferred by utilizing our layer-wise
distillation paradigm. We conduct detailed experiments to demonstrate the
effectiveness, efficacy, and scalability of LaSNN on three benchmark data sets
(CIFAR-10, CIFAR-100, and Tiny ImageNet). We achieve competitive top-1 accuracy
compared to ANNs and 20x faster inference than converted SNNs with similar
performance. More importantly, LaSNN is dexterous and extensible that can be
effortlessly developed for SNNs with different architectures/depths and input
encoding methods, contributing to their potential development.
|
2304.09102 | Joy He-Yueya | Joy He-Yueya, Gabriel Poesia, Rose E. Wang, Noah D. Goodman | Solving Math Word Problems by Combining Language Models With Symbolic
Solvers | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Automatically generating high-quality step-by-step solutions to math word
problems has many applications in education. Recently, combining large language
models (LLMs) with external tools to perform complex reasoning and calculation
has emerged as a promising direction for solving math word problems, but prior
approaches such as Program-Aided Language model (PAL) are biased towards simple
procedural problems and less effective for problems that require declarative
reasoning. We propose an approach that combines an LLM that can incrementally
formalize word problems as a set of variables and equations with an external
symbolic solver that can solve the equations. Our approach achieves comparable
accuracy to the original PAL on the GSM8K benchmark of math word problems and
outperforms PAL by an absolute 20% on ALGEBRA, a new dataset of more
challenging word problems extracted from Algebra textbooks. Our work highlights
the benefits of using declarative and incremental representations when
interfacing with an external tool for solving complex math word problems. Our
data and prompts are publicly available at
https://github.com/joyheyueya/declarative-math-word-problem.
| [
{
"version": "v1",
"created": "Sun, 16 Apr 2023 04:16:06 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"He-Yueya",
"Joy",
""
],
[
"Poesia",
"Gabriel",
""
],
[
"Wang",
"Rose E.",
""
],
[
"Goodman",
"Noah D.",
""
]
] | TITLE: Solving Math Word Problems by Combining Language Models With Symbolic
Solvers
ABSTRACT: Automatically generating high-quality step-by-step solutions to math word
problems has many applications in education. Recently, combining large language
models (LLMs) with external tools to perform complex reasoning and calculation
has emerged as a promising direction for solving math word problems, but prior
approaches such as Program-Aided Language model (PAL) are biased towards simple
procedural problems and less effective for problems that require declarative
reasoning. We propose an approach that combines an LLM that can incrementally
formalize word problems as a set of variables and equations with an external
symbolic solver that can solve the equations. Our approach achieves comparable
accuracy to the original PAL on the GSM8K benchmark of math word problems and
outperforms PAL by an absolute 20% on ALGEBRA, a new dataset of more
challenging word problems extracted from Algebra textbooks. Our work highlights
the benefits of using declarative and incremental representations when
interfacing with an external tool for solving complex math word problems. Our
data and prompts are publicly available at
https://github.com/joyheyueya/declarative-math-word-problem.
|
2304.09105 | Md Mostafizur Rahman Ph.D. | Md Mostafizur Rahman, Daisuke Kikuta, Satyen Abrol, Yu Hirate,
Toyotaro Suzumura, Pablo Loyola, Takuma Ebisu, Manoj Kondapaka | Exploring 360-Degree View of Customers for Lookalike Modeling | null | SIGIR 2023 | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Lookalike models are based on the assumption that user similarity plays an
important role towards product selling and enhancing the existing advertising
campaigns from a very large user base. Challenges associated to these models
reside on the heterogeneity of the user base and its sparsity. In this work, we
propose a novel framework that unifies the customers different behaviors or
features such as demographics, buying behaviors on different platforms,
customer loyalty behaviors and build a lookalike model to improve customer
targeting for Rakuten Group, Inc. Extensive experiments on real e-commerce and
travel datasets demonstrate the effectiveness of our proposed lookalike model
for user targeting task.
| [
{
"version": "v1",
"created": "Mon, 17 Apr 2023 14:01:12 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Rahman",
"Md Mostafizur",
""
],
[
"Kikuta",
"Daisuke",
""
],
[
"Abrol",
"Satyen",
""
],
[
"Hirate",
"Yu",
""
],
[
"Suzumura",
"Toyotaro",
""
],
[
"Loyola",
"Pablo",
""
],
[
"Ebisu",
"Takuma",
""
],
[
"Kondapaka",
"Manoj",
""
]
] | TITLE: Exploring 360-Degree View of Customers for Lookalike Modeling
ABSTRACT: Lookalike models are based on the assumption that user similarity plays an
important role towards product selling and enhancing the existing advertising
campaigns from a very large user base. Challenges associated to these models
reside on the heterogeneity of the user base and its sparsity. In this work, we
propose a novel framework that unifies the customers different behaviors or
features such as demographics, buying behaviors on different platforms,
customer loyalty behaviors and build a lookalike model to improve customer
targeting for Rakuten Group, Inc. Extensive experiments on real e-commerce and
travel datasets demonstrate the effectiveness of our proposed lookalike model
for user targeting task.
|
2304.09131 | Liang Pan | Liang Pan, Xinyi Chen, Zhongang Cai, Junzhe Zhang, Haiyu Zhao, Shuai
Yi, Ziwei Liu | Variational Relational Point Completion Network for Robust 3D
Classification | 12 pages, 10 figures, accepted by PAMI. project webpage:
https://mvp-dataset.github.io/. arXiv admin note: substantial text overlap
with arXiv:2104.10154 | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Real-scanned point clouds are often incomplete due to viewpoint, occlusion,
and noise, which hampers 3D geometric modeling and perception. Existing point
cloud completion methods tend to generate global shape skeletons and hence lack
fine local details. Furthermore, they mostly learn a deterministic
partial-to-complete mapping, but overlook structural relations in man-made
objects. To tackle these challenges, this paper proposes a variational
framework, Variational Relational point Completion Network (VRCNet) with two
appealing properties: 1) Probabilistic Modeling. In particular, we propose a
dual-path architecture to enable principled probabilistic modeling across
partial and complete clouds. One path consumes complete point clouds for
reconstruction by learning a point VAE. The other path generates complete
shapes for partial point clouds, whose embedded distribution is guided by
distribution obtained from the reconstruction path during training. 2)
Relational Enhancement. Specifically, we carefully design point self-attention
kernel and point selective kernel module to exploit relational point features,
which refines local shape details conditioned on the coarse completion. In
addition, we contribute multi-view partial point cloud datasets (MVP and MVP-40
dataset) containing over 200,000 high-quality scans, which render partial 3D
shapes from 26 uniformly distributed camera poses for each 3D CAD model.
Extensive experiments demonstrate that VRCNet outperforms state-of-the-art
methods on all standard point cloud completion benchmarks. Notably, VRCNet
shows great generalizability and robustness on real-world point cloud scans.
Moreover, we can achieve robust 3D classification for partial point clouds with
the help of VRCNet, which can highly increase classification accuracy.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 17:03:20 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Pan",
"Liang",
""
],
[
"Chen",
"Xinyi",
""
],
[
"Cai",
"Zhongang",
""
],
[
"Zhang",
"Junzhe",
""
],
[
"Zhao",
"Haiyu",
""
],
[
"Yi",
"Shuai",
""
],
[
"Liu",
"Ziwei",
""
]
] | TITLE: Variational Relational Point Completion Network for Robust 3D
Classification
ABSTRACT: Real-scanned point clouds are often incomplete due to viewpoint, occlusion,
and noise, which hampers 3D geometric modeling and perception. Existing point
cloud completion methods tend to generate global shape skeletons and hence lack
fine local details. Furthermore, they mostly learn a deterministic
partial-to-complete mapping, but overlook structural relations in man-made
objects. To tackle these challenges, this paper proposes a variational
framework, Variational Relational point Completion Network (VRCNet) with two
appealing properties: 1) Probabilistic Modeling. In particular, we propose a
dual-path architecture to enable principled probabilistic modeling across
partial and complete clouds. One path consumes complete point clouds for
reconstruction by learning a point VAE. The other path generates complete
shapes for partial point clouds, whose embedded distribution is guided by
distribution obtained from the reconstruction path during training. 2)
Relational Enhancement. Specifically, we carefully design point self-attention
kernel and point selective kernel module to exploit relational point features,
which refines local shape details conditioned on the coarse completion. In
addition, we contribute multi-view partial point cloud datasets (MVP and MVP-40
dataset) containing over 200,000 high-quality scans, which render partial 3D
shapes from 26 uniformly distributed camera poses for each 3D CAD model.
Extensive experiments demonstrate that VRCNet outperforms state-of-the-art
methods on all standard point cloud completion benchmarks. Notably, VRCNet
shows great generalizability and robustness on real-world point cloud scans.
Moreover, we can achieve robust 3D classification for partial point clouds with
the help of VRCNet, which can highly increase classification accuracy.
|
2304.09151 | Noah Constant | Hyung Won Chung, Noah Constant, Xavier Garcia, Adam Roberts, Yi Tay,
Sharan Narang, Orhan Firat | UniMax: Fairer and more Effective Language Sampling for Large-Scale
Multilingual Pretraining | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pretrained multilingual large language models have typically used heuristic
temperature-based sampling to balance between different languages. However
previous work has not systematically evaluated the efficacy of different
pretraining language distributions across model scales. In this paper, we
propose a new sampling method, UniMax, that delivers more uniform coverage of
head languages while mitigating overfitting on tail languages by explicitly
capping the number of repeats over each language's corpus. We perform an
extensive series of ablations testing a range of sampling strategies on a suite
of multilingual benchmarks, while varying model scale. We find that UniMax
outperforms standard temperature-based sampling, and the benefits persist as
scale increases. As part of our contribution, we release: (i) an improved and
refreshed mC4 multilingual corpus consisting of 29 trillion characters across
107 languages, and (ii) a suite of pretrained umT5 model checkpoints trained
with UniMax sampling.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 17:45:50 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Chung",
"Hyung Won",
""
],
[
"Constant",
"Noah",
""
],
[
"Garcia",
"Xavier",
""
],
[
"Roberts",
"Adam",
""
],
[
"Tay",
"Yi",
""
],
[
"Narang",
"Sharan",
""
],
[
"Firat",
"Orhan",
""
]
] | TITLE: UniMax: Fairer and more Effective Language Sampling for Large-Scale
Multilingual Pretraining
ABSTRACT: Pretrained multilingual large language models have typically used heuristic
temperature-based sampling to balance between different languages. However
previous work has not systematically evaluated the efficacy of different
pretraining language distributions across model scales. In this paper, we
propose a new sampling method, UniMax, that delivers more uniform coverage of
head languages while mitigating overfitting on tail languages by explicitly
capping the number of repeats over each language's corpus. We perform an
extensive series of ablations testing a range of sampling strategies on a suite
of multilingual benchmarks, while varying model scale. We find that UniMax
outperforms standard temperature-based sampling, and the benefits persist as
scale increases. As part of our contribution, we release: (i) an improved and
refreshed mC4 multilingual corpus consisting of 29 trillion characters across
107 languages, and (ii) a suite of pretrained umT5 model checkpoints trained
with UniMax sampling.
|
2304.09162 | Xinyue Li | Xinyue Li, Jian Wang, Wei Song, Yanling Du, Zhixiang Liu | Robust Calibrate Proxy Loss for Deep Metric Learning | null | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The mainstream researche in deep metric learning can be divided into two
genres: proxy-based and pair-based methods. Proxy-based methods have attracted
extensive attention due to the lower training complexity and fast network
convergence. However, these methods have limitations as the poxy optimization
is done by network, which makes it challenging for the proxy to accurately
represent the feature distrubtion of the real class of data. In this paper, we
propose a Calibrate Proxy (CP) structure, which uses the real sample
information to improve the similarity calculation in proxy-based loss and
introduces a calibration loss to constraint the proxy optimization towards the
center of the class features. At the same time, we set a small number of
proxies for each class to alleviate the impact of intra-class differences on
retrieval performance. The effectiveness of our method is evaluated by
extensive experiments on three public datasets and multiple synthetic
label-noise datasets. The results show that our approach can effectively
improve the performance of commonly used proxy-based losses on both regular and
noisy datasets.
| [
{
"version": "v1",
"created": "Thu, 6 Apr 2023 02:43:10 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Li",
"Xinyue",
""
],
[
"Wang",
"Jian",
""
],
[
"Song",
"Wei",
""
],
[
"Du",
"Yanling",
""
],
[
"Liu",
"Zhixiang",
""
]
] | TITLE: Robust Calibrate Proxy Loss for Deep Metric Learning
ABSTRACT: The mainstream researche in deep metric learning can be divided into two
genres: proxy-based and pair-based methods. Proxy-based methods have attracted
extensive attention due to the lower training complexity and fast network
convergence. However, these methods have limitations as the poxy optimization
is done by network, which makes it challenging for the proxy to accurately
represent the feature distrubtion of the real class of data. In this paper, we
propose a Calibrate Proxy (CP) structure, which uses the real sample
information to improve the similarity calculation in proxy-based loss and
introduces a calibration loss to constraint the proxy optimization towards the
center of the class features. At the same time, we set a small number of
proxies for each class to alleviate the impact of intra-class differences on
retrieval performance. The effectiveness of our method is evaluated by
extensive experiments on three public datasets and multiple synthetic
label-noise datasets. The results show that our approach can effectively
improve the performance of commonly used proxy-based losses on both regular and
noisy datasets.
|
2304.09164 | Paolo Iacono | Paolo Iacono and Naimul Khan | Structure Preserving Cycle-GAN for Unsupervised Medical Image Domain
Adaptation | 11 pages, 4 figures, submitted to Machine Learning for Healthcare
2023 | null | null | null | eess.IV cs.CV cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The presence of domain shift in medical imaging is a common issue, which can
greatly impact the performance of segmentation models when dealing with unseen
image domains. Adversarial-based deep learning models, such as Cycle-GAN, have
become a common model for approaching unsupervised domain adaptation of medical
images. These models however, have no ability to enforce the preservation of
structures of interest when translating medical scans, which can lead to
potentially poor results for unsupervised domain adaptation within the context
of segmentation. This work introduces the Structure Preserving Cycle-GAN (SP
Cycle-GAN), which promotes medical structure preservation during image
translation through the enforcement of a segmentation loss term in the overall
Cycle-GAN training process. We demonstrate the structure preserving capability
of the SP Cycle-GAN both visually and through comparison of Dice score
segmentation performance for the unsupervised domain adaptation models. The SP
Cycle-GAN is able to outperform baseline approaches and standard Cycle-GAN
domain adaptation for binary blood vessel segmentation in the STARE and DRIVE
datasets, and multi-class Left Ventricle and Myocardium segmentation in the
multi-modal MM-WHS dataset. SP Cycle-GAN achieved a state of the art Myocardium
segmentation Dice score (DSC) of 0.7435 for the MR to CT MM-WHS domain
adaptation problem, and excelled in nearly all categories for the MM-WHS
dataset. SP Cycle-GAN also demonstrated a strong ability to preserve blood
vessel structure in the DRIVE to STARE domain adaptation problem, achieving a
4% DSC increase over a default Cycle-GAN implementation.
| [
{
"version": "v1",
"created": "Tue, 18 Apr 2023 17:53:21 GMT"
}
] | 2023-04-19T00:00:00 | [
[
"Iacono",
"Paolo",
""
],
[
"Khan",
"Naimul",
""
]
] | TITLE: Structure Preserving Cycle-GAN for Unsupervised Medical Image Domain
Adaptation
ABSTRACT: The presence of domain shift in medical imaging is a common issue, which can
greatly impact the performance of segmentation models when dealing with unseen
image domains. Adversarial-based deep learning models, such as Cycle-GAN, have
become a common model for approaching unsupervised domain adaptation of medical
images. These models however, have no ability to enforce the preservation of
structures of interest when translating medical scans, which can lead to
potentially poor results for unsupervised domain adaptation within the context
of segmentation. This work introduces the Structure Preserving Cycle-GAN (SP
Cycle-GAN), which promotes medical structure preservation during image
translation through the enforcement of a segmentation loss term in the overall
Cycle-GAN training process. We demonstrate the structure preserving capability
of the SP Cycle-GAN both visually and through comparison of Dice score
segmentation performance for the unsupervised domain adaptation models. The SP
Cycle-GAN is able to outperform baseline approaches and standard Cycle-GAN
domain adaptation for binary blood vessel segmentation in the STARE and DRIVE
datasets, and multi-class Left Ventricle and Myocardium segmentation in the
multi-modal MM-WHS dataset. SP Cycle-GAN achieved a state of the art Myocardium
segmentation Dice score (DSC) of 0.7435 for the MR to CT MM-WHS domain
adaptation problem, and excelled in nearly all categories for the MM-WHS
dataset. SP Cycle-GAN also demonstrated a strong ability to preserve blood
vessel structure in the DRIVE to STARE domain adaptation problem, achieving a
4% DSC increase over a default Cycle-GAN implementation.
|
1811.12624 | Elham J. Barezi Ms | Elham J. Barezi, Peyman Momeni, Pascale Fung | Modality-based Factorization for Multimodal Fusion | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a novel method, Modality-based Redundancy Reduction Fusion (MRRF),
for understanding and modulating the relative contribution of each modality in
multimodal inference tasks. This is achieved by obtaining an $(M+1)$-way tensor
to consider the high-order relationships between $M$ modalities and the output
layer of a neural network model. Applying a modality-based tensor factorization
method, which adopts different factors for different modalities, results in
removing information present in a modality that can be compensated by other
modalities, with respect to model outputs. This helps to understand the
relative utility of information in each modality. In addition it leads to a
less complicated model with less parameters and therefore could be applied as a
regularizer avoiding overfitting. We have applied this method to three
different multimodal datasets in sentiment analysis, personality trait
recognition, and emotion recognition. We are able to recognize relationships
and relative importance of different modalities in these tasks and achieves a
1\% to 4\% improvement on several evaluation measures compared to the
state-of-the-art for all three tasks.
| [
{
"version": "v1",
"created": "Fri, 30 Nov 2018 05:43:10 GMT"
},
{
"version": "v2",
"created": "Fri, 19 Jul 2019 04:15:48 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Apr 2023 02:48:28 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Barezi",
"Elham J.",
""
],
[
"Momeni",
"Peyman",
""
],
[
"Fung",
"Pascale",
""
]
] | TITLE: Modality-based Factorization for Multimodal Fusion
ABSTRACT: We propose a novel method, Modality-based Redundancy Reduction Fusion (MRRF),
for understanding and modulating the relative contribution of each modality in
multimodal inference tasks. This is achieved by obtaining an $(M+1)$-way tensor
to consider the high-order relationships between $M$ modalities and the output
layer of a neural network model. Applying a modality-based tensor factorization
method, which adopts different factors for different modalities, results in
removing information present in a modality that can be compensated by other
modalities, with respect to model outputs. This helps to understand the
relative utility of information in each modality. In addition it leads to a
less complicated model with less parameters and therefore could be applied as a
regularizer avoiding overfitting. We have applied this method to three
different multimodal datasets in sentiment analysis, personality trait
recognition, and emotion recognition. We are able to recognize relationships
and relative importance of different modalities in these tasks and achieves a
1\% to 4\% improvement on several evaluation measures compared to the
state-of-the-art for all three tasks.
|
2007.07627 | Bailin Deng | Juyong Zhang and Yuxin Yao and Bailin Deng | Fast and Robust Iterative Closest Point | null | IEEE Transactions on Pattern Analysis and Machine Intelligence,
2021 | 10.1109/TPAMI.2021.3054619 | null | cs.CV cs.GR cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The Iterative Closest Point (ICP) algorithm and its variants are a
fundamental technique for rigid registration between two point sets, with wide
applications in different areas from robotics to 3D reconstruction. The main
drawbacks for ICP are its slow convergence as well as its sensitivity to
outliers, missing data, and partial overlaps. Recent work such as Sparse ICP
achieves robustness via sparsity optimization at the cost of computational
speed. In this paper, we propose a new method for robust registration with fast
convergence. First, we show that the classical point-to-point ICP can be
treated as a majorization-minimization (MM) algorithm, and propose an Anderson
acceleration approach to speed up its convergence. In addition, we introduce a
robust error metric based on the Welsch's function, which is minimized
efficiently using the MM algorithm with Anderson acceleration. On challenging
datasets with noises and partial overlaps, we achieve similar or better
accuracy than Sparse ICP while being at least an order of magnitude faster.
Finally, we extend the robust formulation to point-to-plane ICP, and solve the
resulting problem using a similar Anderson-accelerated MM strategy. Our robust
ICP methods improve the registration accuracy on benchmark datasets while being
competitive in computational time.
| [
{
"version": "v1",
"created": "Wed, 15 Jul 2020 11:32:53 GMT"
},
{
"version": "v2",
"created": "Tue, 29 Dec 2020 14:24:22 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Apr 2023 19:47:28 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Zhang",
"Juyong",
""
],
[
"Yao",
"Yuxin",
""
],
[
"Deng",
"Bailin",
""
]
] | TITLE: Fast and Robust Iterative Closest Point
ABSTRACT: The Iterative Closest Point (ICP) algorithm and its variants are a
fundamental technique for rigid registration between two point sets, with wide
applications in different areas from robotics to 3D reconstruction. The main
drawbacks for ICP are its slow convergence as well as its sensitivity to
outliers, missing data, and partial overlaps. Recent work such as Sparse ICP
achieves robustness via sparsity optimization at the cost of computational
speed. In this paper, we propose a new method for robust registration with fast
convergence. First, we show that the classical point-to-point ICP can be
treated as a majorization-minimization (MM) algorithm, and propose an Anderson
acceleration approach to speed up its convergence. In addition, we introduce a
robust error metric based on the Welsch's function, which is minimized
efficiently using the MM algorithm with Anderson acceleration. On challenging
datasets with noises and partial overlaps, we achieve similar or better
accuracy than Sparse ICP while being at least an order of magnitude faster.
Finally, we extend the robust formulation to point-to-plane ICP, and solve the
resulting problem using a similar Anderson-accelerated MM strategy. Our robust
ICP methods improve the registration accuracy on benchmark datasets while being
competitive in computational time.
|
2103.00083 | Rasool Fakoor | Rasool Fakoor, Taesup Kim, Jonas Mueller, Alexander J. Smola, Ryan J.
Tibshirani | Flexible Model Aggregation for Quantile Regression | Accepted at JMLR 2023 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Quantile regression is a fundamental problem in statistical learning
motivated by a need to quantify uncertainty in predictions, or to model a
diverse population without being overly reductive. For instance,
epidemiological forecasts, cost estimates, and revenue predictions all benefit
from being able to quantify the range of possible values accurately. As such,
many models have been developed for this problem over many years of research in
statistics, machine learning, and related fields. Rather than proposing yet
another (new) algorithm for quantile regression we adopt a meta viewpoint: we
investigate methods for aggregating any number of conditional quantile models,
in order to improve accuracy and robustness. We consider weighted ensembles
where weights may vary over not only individual models, but also over quantile
levels, and feature values. All of the models we consider in this paper can be
fit using modern deep learning toolkits, and hence are widely accessible (from
an implementation point of view) and scalable. To improve the accuracy of the
predicted quantiles (or equivalently, prediction intervals), we develop tools
for ensuring that quantiles remain monotonically ordered, and apply conformal
calibration methods. These can be used without any modification of the original
library of base models. We also review some basic theory surrounding quantile
aggregation and related scoring rules, and contribute a few new results to this
literature (for example, the fact that post sorting or post isotonic regression
can only improve the weighted interval score). Finally, we provide an extensive
suite of empirical comparisons across 34 data sets from two different benchmark
repositories.
| [
{
"version": "v1",
"created": "Fri, 26 Feb 2021 23:21:16 GMT"
},
{
"version": "v2",
"created": "Tue, 16 Mar 2021 01:23:59 GMT"
},
{
"version": "v3",
"created": "Fri, 23 Jul 2021 17:11:29 GMT"
},
{
"version": "v4",
"created": "Thu, 14 Jul 2022 19:03:55 GMT"
},
{
"version": "v5",
"created": "Sat, 15 Apr 2023 08:40:57 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Fakoor",
"Rasool",
""
],
[
"Kim",
"Taesup",
""
],
[
"Mueller",
"Jonas",
""
],
[
"Smola",
"Alexander J.",
""
],
[
"Tibshirani",
"Ryan J.",
""
]
] | TITLE: Flexible Model Aggregation for Quantile Regression
ABSTRACT: Quantile regression is a fundamental problem in statistical learning
motivated by a need to quantify uncertainty in predictions, or to model a
diverse population without being overly reductive. For instance,
epidemiological forecasts, cost estimates, and revenue predictions all benefit
from being able to quantify the range of possible values accurately. As such,
many models have been developed for this problem over many years of research in
statistics, machine learning, and related fields. Rather than proposing yet
another (new) algorithm for quantile regression we adopt a meta viewpoint: we
investigate methods for aggregating any number of conditional quantile models,
in order to improve accuracy and robustness. We consider weighted ensembles
where weights may vary over not only individual models, but also over quantile
levels, and feature values. All of the models we consider in this paper can be
fit using modern deep learning toolkits, and hence are widely accessible (from
an implementation point of view) and scalable. To improve the accuracy of the
predicted quantiles (or equivalently, prediction intervals), we develop tools
for ensuring that quantiles remain monotonically ordered, and apply conformal
calibration methods. These can be used without any modification of the original
library of base models. We also review some basic theory surrounding quantile
aggregation and related scoring rules, and contribute a few new results to this
literature (for example, the fact that post sorting or post isotonic regression
can only improve the weighted interval score). Finally, we provide an extensive
suite of empirical comparisons across 34 data sets from two different benchmark
repositories.
|
2103.00959 | Yukuo Cen | Yukuo Cen, Zhenyu Hou, Yan Wang, Qibin Chen, Yizhen Luo, Zhongming Yu,
Hengrui Zhang, Xingcheng Yao, Aohan Zeng, Shiguang Guo, Yuxiao Dong, Yang
Yang, Peng Zhang, Guohao Dai, Yu Wang, Chang Zhou, Hongxia Yang, Jie Tang | CogDL: A Comprehensive Library for Graph Deep Learning | Accepted to WWW 2023. Website: https://github.com/THUDM/cogdl | null | null | null | cs.SI cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Graph neural networks (GNNs) have attracted tremendous attention from the
graph learning community in recent years. It has been widely adopted in various
real-world applications from diverse domains, such as social networks and
biological graphs. The research and applications of graph deep learning present
new challenges, including the sparse nature of graph data, complicated training
of GNNs, and non-standard evaluation of graph tasks. To tackle the issues, we
present CogDL, a comprehensive library for graph deep learning that allows
researchers and practitioners to conduct experiments, compare methods, and
build applications with ease and efficiency. In CogDL, we propose a unified
design for the training and evaluation of GNN models for various graph tasks,
making it unique among existing graph learning libraries. By utilizing this
unified trainer, CogDL can optimize the GNN training loop with several training
techniques, such as mixed precision training. Moreover, we develop efficient
sparse operators for CogDL, enabling it to become the most competitive graph
library for efficiency. Another important CogDL feature is its focus on ease of
use with the aim of facilitating open and reproducible research of graph
learning. We leverage CogDL to report and maintain benchmark results on
fundamental graph tasks, which can be reproduced and directly used by the
community.
| [
{
"version": "v1",
"created": "Mon, 1 Mar 2021 12:35:16 GMT"
},
{
"version": "v2",
"created": "Fri, 23 Jul 2021 03:29:00 GMT"
},
{
"version": "v3",
"created": "Thu, 3 Mar 2022 13:23:37 GMT"
},
{
"version": "v4",
"created": "Mon, 17 Apr 2023 10:44:27 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Cen",
"Yukuo",
""
],
[
"Hou",
"Zhenyu",
""
],
[
"Wang",
"Yan",
""
],
[
"Chen",
"Qibin",
""
],
[
"Luo",
"Yizhen",
""
],
[
"Yu",
"Zhongming",
""
],
[
"Zhang",
"Hengrui",
""
],
[
"Yao",
"Xingcheng",
""
],
[
"Zeng",
"Aohan",
""
],
[
"Guo",
"Shiguang",
""
],
[
"Dong",
"Yuxiao",
""
],
[
"Yang",
"Yang",
""
],
[
"Zhang",
"Peng",
""
],
[
"Dai",
"Guohao",
""
],
[
"Wang",
"Yu",
""
],
[
"Zhou",
"Chang",
""
],
[
"Yang",
"Hongxia",
""
],
[
"Tang",
"Jie",
""
]
] | TITLE: CogDL: A Comprehensive Library for Graph Deep Learning
ABSTRACT: Graph neural networks (GNNs) have attracted tremendous attention from the
graph learning community in recent years. It has been widely adopted in various
real-world applications from diverse domains, such as social networks and
biological graphs. The research and applications of graph deep learning present
new challenges, including the sparse nature of graph data, complicated training
of GNNs, and non-standard evaluation of graph tasks. To tackle the issues, we
present CogDL, a comprehensive library for graph deep learning that allows
researchers and practitioners to conduct experiments, compare methods, and
build applications with ease and efficiency. In CogDL, we propose a unified
design for the training and evaluation of GNN models for various graph tasks,
making it unique among existing graph learning libraries. By utilizing this
unified trainer, CogDL can optimize the GNN training loop with several training
techniques, such as mixed precision training. Moreover, we develop efficient
sparse operators for CogDL, enabling it to become the most competitive graph
library for efficiency. Another important CogDL feature is its focus on ease of
use with the aim of facilitating open and reproducible research of graph
learning. We leverage CogDL to report and maintain benchmark results on
fundamental graph tasks, which can be reproduced and directly used by the
community.
|
2104.02025 | Herschel Pangborn | Austin L. Nash, Herschel C. Pangborn, and Neera Jain | Robust Control Co-Design with Receding-Horizon MPC | null | American Control Conference (2021) | 10.23919/ACC50511.2021.9483216 | null | eess.SY cs.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Control co-design (CCD) is a technique for improving the closed-loop
performance of systems through the coordinated design of both plant parameters
and an optimal control policy. While model predictive control (MPC) is an
attractive control strategy for many systems, embedding it within a CCD
algorithm presents challenges because obtaining a closed-form solution for this
receding-horizon optimization strategy is often not feasible. This paper meets
that challenge by including a robust MPC formulation within the inner loop of a
CCD algorithm. As exemplified by application to an aircraft thermal management
system, the proposed algorithm closely matches the plant design of an open-loop
benchmark. However, unlike the open-loop approach, the proposed algorithm can
leverage MPC control variables designed a priori to achieve robust online
operation under disturbance profiles that differ from those used for design.
| [
{
"version": "v1",
"created": "Mon, 5 Apr 2021 17:24:30 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Apr 2023 23:34:11 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Nash",
"Austin L.",
""
],
[
"Pangborn",
"Herschel C.",
""
],
[
"Jain",
"Neera",
""
]
] | TITLE: Robust Control Co-Design with Receding-Horizon MPC
ABSTRACT: Control co-design (CCD) is a technique for improving the closed-loop
performance of systems through the coordinated design of both plant parameters
and an optimal control policy. While model predictive control (MPC) is an
attractive control strategy for many systems, embedding it within a CCD
algorithm presents challenges because obtaining a closed-form solution for this
receding-horizon optimization strategy is often not feasible. This paper meets
that challenge by including a robust MPC formulation within the inner loop of a
CCD algorithm. As exemplified by application to an aircraft thermal management
system, the proposed algorithm closely matches the plant design of an open-loop
benchmark. However, unlike the open-loop approach, the proposed algorithm can
leverage MPC control variables designed a priori to achieve robust online
operation under disturbance profiles that differ from those used for design.
|
2105.14559 | Jae Oh Woo | Jae Oh Woo | Active Learning in Bayesian Neural Networks with Balanced Entropy
Learning Principle | null | International Conference on Learning Representations 2023 | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Acquiring labeled data is challenging in many machine learning applications
with limited budgets. Active learning gives a procedure to select the most
informative data points and improve data efficiency by reducing the cost of
labeling. The info-max learning principle maximizing mutual information such as
BALD has been successful and widely adapted in various active learning
applications. However, this pool-based specific objective inherently introduces
a redundant selection and further requires a high computational cost for batch
selection. In this paper, we design and propose a new uncertainty measure,
Balanced Entropy Acquisition (BalEntAcq), which captures the information
balance between the uncertainty of underlying softmax probability and the label
variable. To do this, we approximate each marginal distribution by Beta
distribution. Beta approximation enables us to formulate BalEntAcq as a ratio
between an augmented entropy and the marginalized joint entropy. The
closed-form expression of BalEntAcq facilitates parallelization by estimating
two parameters in each marginal Beta distribution. BalEntAcq is a purely
standalone measure without requiring any relational computations with other
data points. Nevertheless, BalEntAcq captures a well-diversified selection near
the decision boundary with a margin, unlike other existing uncertainty measures
such as BALD, Entropy, or Mean Standard Deviation (MeanSD). Finally, we
demonstrate that our balanced entropy learning principle with BalEntAcq
consistently outperforms well-known linearly scalable active learning methods,
including a recently proposed PowerBALD, a simple but diversified version of
BALD, by showing experimental results obtained from MNIST, CIFAR-100, SVHN, and
TinyImageNet datasets.
| [
{
"version": "v1",
"created": "Sun, 30 May 2021 14:49:10 GMT"
},
{
"version": "v2",
"created": "Thu, 29 Sep 2022 07:40:12 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Apr 2023 05:45:25 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Woo",
"Jae Oh",
""
]
] | TITLE: Active Learning in Bayesian Neural Networks with Balanced Entropy
Learning Principle
ABSTRACT: Acquiring labeled data is challenging in many machine learning applications
with limited budgets. Active learning gives a procedure to select the most
informative data points and improve data efficiency by reducing the cost of
labeling. The info-max learning principle maximizing mutual information such as
BALD has been successful and widely adapted in various active learning
applications. However, this pool-based specific objective inherently introduces
a redundant selection and further requires a high computational cost for batch
selection. In this paper, we design and propose a new uncertainty measure,
Balanced Entropy Acquisition (BalEntAcq), which captures the information
balance between the uncertainty of underlying softmax probability and the label
variable. To do this, we approximate each marginal distribution by Beta
distribution. Beta approximation enables us to formulate BalEntAcq as a ratio
between an augmented entropy and the marginalized joint entropy. The
closed-form expression of BalEntAcq facilitates parallelization by estimating
two parameters in each marginal Beta distribution. BalEntAcq is a purely
standalone measure without requiring any relational computations with other
data points. Nevertheless, BalEntAcq captures a well-diversified selection near
the decision boundary with a margin, unlike other existing uncertainty measures
such as BALD, Entropy, or Mean Standard Deviation (MeanSD). Finally, we
demonstrate that our balanced entropy learning principle with BalEntAcq
consistently outperforms well-known linearly scalable active learning methods,
including a recently proposed PowerBALD, a simple but diversified version of
BALD, by showing experimental results obtained from MNIST, CIFAR-100, SVHN, and
TinyImageNet datasets.
|
2106.01132 | Benoit Dufumier | Benoit Dufumier, Pietro Gori, Ilaria Battaglia, Julie Victor, Antoine
Grigis, Edouard Duchesnay | Benchmarking CNN on 3D Anatomical Brain MRI: Architectures, Data
Augmentation and Deep Ensemble Learning | Technical report | null | null | null | cs.CV cs.LG eess.IV | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Deep Learning (DL) and specifically CNN models have become a de facto method
for a wide range of vision tasks, outperforming traditional machine learning
(ML) methods. Consequently, they drew a lot of attention in the neuroimaging
field in particular for phenotype prediction or computer-aided diagnosis.
However, most of the current studies often deal with small single-site cohorts,
along with a specific pre-processing pipeline and custom CNN architectures,
which make them difficult to compare to. We propose an extensive benchmark of
recent state-of-the-art (SOTA) 3D CNN, evaluating also the benefits of data
augmentation and deep ensemble learning, on both Voxel-Based Morphometry (VBM)
pre-processing and quasi-raw images. Experiments were conducted on a large
multi-site 3D brain anatomical MRI data-set comprising N=10k scans on 3
challenging tasks: age prediction, sex classification, and schizophrenia
diagnosis. We found that all models provide significantly better predictions
with VBM images than quasi-raw data. This finding evolved as the training set
approaches 10k samples where quasi-raw data almost reach the performance of
VBM. Moreover, we showed that linear models perform comparably with SOTA CNN on
VBM data. We also demonstrated that DenseNet and tiny-DenseNet, a lighter
version that we proposed, provide a good compromise in terms of performance in
all data regime. Therefore, we suggest to employ them as the architectures by
default. Critically, we also showed that current CNN are still very biased
towards the acquisition site, even when trained with N=10k multi-site images.
In this context, VBM pre-processing provides an efficient way to limit this
site effect. Surprisingly, we did not find any clear benefit from data
augmentation techniques. Finally, we proved that deep ensemble learning is well
suited to re-calibrate big CNN models without sacrificing performance.
| [
{
"version": "v1",
"created": "Wed, 2 Jun 2021 13:00:35 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Apr 2023 12:48:33 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Dufumier",
"Benoit",
""
],
[
"Gori",
"Pietro",
""
],
[
"Battaglia",
"Ilaria",
""
],
[
"Victor",
"Julie",
""
],
[
"Grigis",
"Antoine",
""
],
[
"Duchesnay",
"Edouard",
""
]
] | TITLE: Benchmarking CNN on 3D Anatomical Brain MRI: Architectures, Data
Augmentation and Deep Ensemble Learning
ABSTRACT: Deep Learning (DL) and specifically CNN models have become a de facto method
for a wide range of vision tasks, outperforming traditional machine learning
(ML) methods. Consequently, they drew a lot of attention in the neuroimaging
field in particular for phenotype prediction or computer-aided diagnosis.
However, most of the current studies often deal with small single-site cohorts,
along with a specific pre-processing pipeline and custom CNN architectures,
which make them difficult to compare to. We propose an extensive benchmark of
recent state-of-the-art (SOTA) 3D CNN, evaluating also the benefits of data
augmentation and deep ensemble learning, on both Voxel-Based Morphometry (VBM)
pre-processing and quasi-raw images. Experiments were conducted on a large
multi-site 3D brain anatomical MRI data-set comprising N=10k scans on 3
challenging tasks: age prediction, sex classification, and schizophrenia
diagnosis. We found that all models provide significantly better predictions
with VBM images than quasi-raw data. This finding evolved as the training set
approaches 10k samples where quasi-raw data almost reach the performance of
VBM. Moreover, we showed that linear models perform comparably with SOTA CNN on
VBM data. We also demonstrated that DenseNet and tiny-DenseNet, a lighter
version that we proposed, provide a good compromise in terms of performance in
all data regime. Therefore, we suggest to employ them as the architectures by
default. Critically, we also showed that current CNN are still very biased
towards the acquisition site, even when trained with N=10k multi-site images.
In this context, VBM pre-processing provides an efficient way to limit this
site effect. Surprisingly, we did not find any clear benefit from data
augmentation techniques. Finally, we proved that deep ensemble learning is well
suited to re-calibrate big CNN models without sacrificing performance.
|
2106.15214 | Arthur Marmin | Arthur Marmin and Jos\'e Henrique de Morais Goulart and C\'edric
F\'evotte | Joint Majorization-Minimization for Nonnegative Matrix Factorization
with the $\beta$-divergence | null | Signal Processing, Volume 209, 2023,109048,ISSN 0165-1684 | 10.1016/j.sigpro.2023.109048 | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This article proposes new multiplicative updates for nonnegative matrix
factorization (NMF) with the $\beta$-divergence objective function. Our new
updates are derived from a joint majorization-minimization (MM) scheme, in
which an auxiliary function (a tight upper bound of the objective function) is
built for the two factors jointly and minimized at each iteration. This is in
contrast with the classic approach in which a majorizer is derived for each
factor separately. Like that classic approach, our joint MM algorithm also
results in multiplicative updates that are simple to implement. They however
yield a significant drop of computation time (for equally good solutions), in
particular for some $\beta$-divergences of important applicative interest, such
as the squared Euclidean distance and the Kullback-Leibler or Itakura-Saito
divergences. We report experimental results using diverse datasets: face
images, an audio spectrogram, hyperspectral data and song play counts.
Depending on the value of $\beta$ and on the dataset, our joint MM approach can
yield CPU time reductions from about $13\%$ to $78\%$ in comparison to the
classic alternating scheme.
| [
{
"version": "v1",
"created": "Tue, 29 Jun 2021 09:58:21 GMT"
},
{
"version": "v2",
"created": "Mon, 20 Dec 2021 15:55:12 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Oct 2022 13:08:29 GMT"
},
{
"version": "v4",
"created": "Mon, 17 Apr 2023 08:30:42 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Marmin",
"Arthur",
""
],
[
"Goulart",
"José Henrique de Morais",
""
],
[
"Févotte",
"Cédric",
""
]
] | TITLE: Joint Majorization-Minimization for Nonnegative Matrix Factorization
with the $\beta$-divergence
ABSTRACT: This article proposes new multiplicative updates for nonnegative matrix
factorization (NMF) with the $\beta$-divergence objective function. Our new
updates are derived from a joint majorization-minimization (MM) scheme, in
which an auxiliary function (a tight upper bound of the objective function) is
built for the two factors jointly and minimized at each iteration. This is in
contrast with the classic approach in which a majorizer is derived for each
factor separately. Like that classic approach, our joint MM algorithm also
results in multiplicative updates that are simple to implement. They however
yield a significant drop of computation time (for equally good solutions), in
particular for some $\beta$-divergences of important applicative interest, such
as the squared Euclidean distance and the Kullback-Leibler or Itakura-Saito
divergences. We report experimental results using diverse datasets: face
images, an audio spectrogram, hyperspectral data and song play counts.
Depending on the value of $\beta$ and on the dataset, our joint MM approach can
yield CPU time reductions from about $13\%$ to $78\%$ in comparison to the
classic alternating scheme.
|
2107.06700 | Yinghua Yao | Yinghua Yao, Yuangang Pan, Ivor W.Tsang, Xin Yao | Differential-Critic GAN: Generating What You Want by a Cue of
Preferences | null | null | 10.1109/TNNLS.2022.3197313 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes Differential-Critic Generative Adversarial Network
(DiCGAN) to learn the distribution of user-desired data when only partial
instead of the entire dataset possesses the desired property. DiCGAN generates
desired data that meets the user's expectations and can assist in designing
biological products with desired properties. Existing approaches select the
desired samples first and train regular GANs on the selected samples to derive
the user-desired data distribution. However, the selection of the desired data
relies on global knowledge and supervision over the entire dataset. DiCGAN
introduces a differential critic that learns from pairwise preferences, which
are local knowledge and can be defined on a part of training data. The critic
is built by defining an additional ranking loss over the Wasserstein GAN's
critic. It endows the difference of critic values between each pair of samples
with the user preference and guides the generation of the desired data instead
of the whole data. For a more efficient solution to ensure data quality, we
further reformulate DiCGAN as a constrained optimization problem, based on
which we theoretically prove the convergence of our DiCGAN. Extensive
experiments on a diverse set of datasets with various applications demonstrate
that our DiCGAN achieves state-of-the-art performance in learning the
user-desired data distributions, especially in the cases of insufficient
desired data and limited supervision.
| [
{
"version": "v1",
"created": "Wed, 14 Jul 2021 13:44:07 GMT"
},
{
"version": "v2",
"created": "Sat, 6 Aug 2022 04:00:58 GMT"
},
{
"version": "v3",
"created": "Mon, 17 Apr 2023 02:23:11 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Yao",
"Yinghua",
""
],
[
"Pan",
"Yuangang",
""
],
[
"Tsang",
"Ivor W.",
""
],
[
"Yao",
"Xin",
""
]
] | TITLE: Differential-Critic GAN: Generating What You Want by a Cue of
Preferences
ABSTRACT: This paper proposes Differential-Critic Generative Adversarial Network
(DiCGAN) to learn the distribution of user-desired data when only partial
instead of the entire dataset possesses the desired property. DiCGAN generates
desired data that meets the user's expectations and can assist in designing
biological products with desired properties. Existing approaches select the
desired samples first and train regular GANs on the selected samples to derive
the user-desired data distribution. However, the selection of the desired data
relies on global knowledge and supervision over the entire dataset. DiCGAN
introduces a differential critic that learns from pairwise preferences, which
are local knowledge and can be defined on a part of training data. The critic
is built by defining an additional ranking loss over the Wasserstein GAN's
critic. It endows the difference of critic values between each pair of samples
with the user preference and guides the generation of the desired data instead
of the whole data. For a more efficient solution to ensure data quality, we
further reformulate DiCGAN as a constrained optimization problem, based on
which we theoretically prove the convergence of our DiCGAN. Extensive
experiments on a diverse set of datasets with various applications demonstrate
that our DiCGAN achieves state-of-the-art performance in learning the
user-desired data distributions, especially in the cases of insufficient
desired data and limited supervision.
|
2108.03499 | Luca Surace | Luca Surace (Universit\`a della Svizzera italiana), Marek Wernikowski
(West Pomeranian University of Technology), Cara Tursun (Universit\`a della
Svizzera italiana and University of Groningen), Karol Myszkowski (Max Planck
Institute for Informatics), Rados{\l}aw Mantiuk (West Pomeranian University
of Technology), Piotr Didyk (Universit\`a della Svizzera italiana) | Learning GAN-based Foveated Reconstruction to Recover Perceptually
Important Image Features | null | null | 10.1145/3583072 | null | cs.GR cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A foveated image can be entirely reconstructed from a sparse set of samples
distributed according to the retinal sensitivity of the human visual system,
which rapidly decreases with increasing eccentricity. The use of Generative
Adversarial Networks has recently been shown to be a promising solution for
such a task, as they can successfully hallucinate missing image information. As
in the case of other supervised learning approaches, the definition of the loss
function and the training strategy heavily influence the quality of the output.
In this work,we consider the problem of efficiently guiding the training of
foveated reconstruction techniques such that they are more aware of the
capabilities and limitations of the human visual system, and thus can
reconstruct visually important image features. Our primary goal is to make the
training procedure less sensitive to distortions that humans cannot detect and
focus on penalizing perceptually important artifacts. Given the nature of
GAN-based solutions, we focus on the sensitivity of human vision to
hallucination in case of input samples with different densities. We propose
psychophysical experiments, a dataset, and a procedure for training foveated
image reconstruction. The proposed strategy renders the generator network
flexible by penalizing only perceptually important deviations in the output. As
a result, the method emphasized the recovery of perceptually important image
features. We evaluated our strategy and compared it with alternative solutions
by using a newly trained objective metric, a recent foveated video quality
metric, and user experiments. Our evaluations revealed significant improvements
in the perceived image reconstruction quality compared with the standard
GAN-based training approach.
| [
{
"version": "v1",
"created": "Sat, 7 Aug 2021 18:39:49 GMT"
},
{
"version": "v2",
"created": "Wed, 25 May 2022 13:52:54 GMT"
},
{
"version": "v3",
"created": "Mon, 17 Apr 2023 16:42:28 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Surace",
"Luca",
"",
"Università della Svizzera italiana"
],
[
"Wernikowski",
"Marek",
"",
"West Pomeranian University of Technology"
],
[
"Tursun",
"Cara",
"",
"Università della\n Svizzera italiana and University of Groningen"
],
[
"Myszkowski",
"Karol",
"",
"Max Planck\n Institute for Informatics"
],
[
"Mantiuk",
"Radosław",
"",
"West Pomeranian University\n of Technology"
],
[
"Didyk",
"Piotr",
"",
"Università della Svizzera italiana"
]
] | TITLE: Learning GAN-based Foveated Reconstruction to Recover Perceptually
Important Image Features
ABSTRACT: A foveated image can be entirely reconstructed from a sparse set of samples
distributed according to the retinal sensitivity of the human visual system,
which rapidly decreases with increasing eccentricity. The use of Generative
Adversarial Networks has recently been shown to be a promising solution for
such a task, as they can successfully hallucinate missing image information. As
in the case of other supervised learning approaches, the definition of the loss
function and the training strategy heavily influence the quality of the output.
In this work,we consider the problem of efficiently guiding the training of
foveated reconstruction techniques such that they are more aware of the
capabilities and limitations of the human visual system, and thus can
reconstruct visually important image features. Our primary goal is to make the
training procedure less sensitive to distortions that humans cannot detect and
focus on penalizing perceptually important artifacts. Given the nature of
GAN-based solutions, we focus on the sensitivity of human vision to
hallucination in case of input samples with different densities. We propose
psychophysical experiments, a dataset, and a procedure for training foveated
image reconstruction. The proposed strategy renders the generator network
flexible by penalizing only perceptually important deviations in the output. As
a result, the method emphasized the recovery of perceptually important image
features. We evaluated our strategy and compared it with alternative solutions
by using a newly trained objective metric, a recent foveated video quality
metric, and user experiments. Our evaluations revealed significant improvements
in the perceived image reconstruction quality compared with the standard
GAN-based training approach.
|
2108.05271 | Georgi Karadzhov | Georgi Karadzhov, Tom Stafford, Andreas Vlachos | DeliData: A dataset for deliberation in multi-party problem solving | null | null | null | null | cs.CL cs.AI cs.CY | http://creativecommons.org/licenses/by/4.0/ | Group deliberation enables people to collaborate and solve problems, however,
it is understudied due to a lack of resources. To this end, we introduce the
first publicly available dataset containing collaborative conversations on
solving a well-established cognitive task, consisting of 500 group dialogues
and 14k utterances. In 64% of these conversations, the group members are able
to find a better solution than they had identified individually, and in 43.8%
of the groups who had a correct answer as their final solution, none of the
participants had solved the task correctly by themselves. Furthermore, we
propose a novel annotation schema that captures deliberation cues and release
all 14k utterances annotated with it. Finally, we use the proposed dataset to
develop and evaluate two methods for generating deliberation utterances. The
data collection platform, dataset and annotated corpus are publicly available
at https://delibot.xyz.
| [
{
"version": "v1",
"created": "Wed, 11 Aug 2021 15:13:07 GMT"
},
{
"version": "v2",
"created": "Sat, 7 May 2022 18:18:00 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Apr 2023 13:11:25 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Karadzhov",
"Georgi",
""
],
[
"Stafford",
"Tom",
""
],
[
"Vlachos",
"Andreas",
""
]
] | TITLE: DeliData: A dataset for deliberation in multi-party problem solving
ABSTRACT: Group deliberation enables people to collaborate and solve problems, however,
it is understudied due to a lack of resources. To this end, we introduce the
first publicly available dataset containing collaborative conversations on
solving a well-established cognitive task, consisting of 500 group dialogues
and 14k utterances. In 64% of these conversations, the group members are able
to find a better solution than they had identified individually, and in 43.8%
of the groups who had a correct answer as their final solution, none of the
participants had solved the task correctly by themselves. Furthermore, we
propose a novel annotation schema that captures deliberation cues and release
all 14k utterances annotated with it. Finally, we use the proposed dataset to
develop and evaluate two methods for generating deliberation utterances. The
data collection platform, dataset and annotated corpus are publicly available
at https://delibot.xyz.
|
2110.00637 | Haoyue Dai | Haoyue Dai, Rui Ding, Yuanyuan Jiang, Shi Han, Dongmei Zhang | ML4C: Seeing Causality Through Latent Vicinity | causal discovery, supervised causal learning, vicinity,
identifiability, learnability | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Supervised Causal Learning (SCL) aims to learn causal relations from
observational data by accessing previously seen datasets associated with ground
truth causal relations. This paper presents a first attempt at addressing a
fundamental question: What are the benefits from supervision and how does it
benefit? Starting from seeing that SCL is not better than random guessing if
the learning target is non-identifiable a priori, we propose a two-phase
paradigm for SCL by explicitly considering structure identifiability. Following
this paradigm, we tackle the problem of SCL on discrete data and propose ML4C.
The core of ML4C is a binary classifier with a novel learning target: it
classifies whether an Unshielded Triple (UT) is a v-structure or not.
Specifically, starting from an input dataset with the corresponding skeleton
provided, ML4C orients each UT once it is classified as a v-structure. These
v-structures are together used to construct the final output. To address the
fundamental question of SCL, we propose a principled method for ML4C
featurization: we exploit the vicinity of a given UT (i.e., the neighbors of UT
in skeleton), and derive features by considering the conditional dependencies
and structural entanglement within the vicinity. We further prove that ML4C is
asymptotically correct. Last but foremost, thorough experiments conducted on
benchmark datasets demonstrate that ML4C remarkably outperforms other
state-of-the-art algorithms in terms of accuracy, reliability, robustness and
tolerance. In summary, ML4C shows promising results on validating the
effectiveness of supervision for causal learning. Our codes are publicly
available at https://github.com/microsoft/ML4C.
| [
{
"version": "v1",
"created": "Fri, 1 Oct 2021 20:13:07 GMT"
},
{
"version": "v2",
"created": "Thu, 20 Oct 2022 05:14:57 GMT"
},
{
"version": "v3",
"created": "Wed, 26 Oct 2022 00:05:12 GMT"
},
{
"version": "v4",
"created": "Sun, 16 Apr 2023 05:01:45 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Dai",
"Haoyue",
""
],
[
"Ding",
"Rui",
""
],
[
"Jiang",
"Yuanyuan",
""
],
[
"Han",
"Shi",
""
],
[
"Zhang",
"Dongmei",
""
]
] | TITLE: ML4C: Seeing Causality Through Latent Vicinity
ABSTRACT: Supervised Causal Learning (SCL) aims to learn causal relations from
observational data by accessing previously seen datasets associated with ground
truth causal relations. This paper presents a first attempt at addressing a
fundamental question: What are the benefits from supervision and how does it
benefit? Starting from seeing that SCL is not better than random guessing if
the learning target is non-identifiable a priori, we propose a two-phase
paradigm for SCL by explicitly considering structure identifiability. Following
this paradigm, we tackle the problem of SCL on discrete data and propose ML4C.
The core of ML4C is a binary classifier with a novel learning target: it
classifies whether an Unshielded Triple (UT) is a v-structure or not.
Specifically, starting from an input dataset with the corresponding skeleton
provided, ML4C orients each UT once it is classified as a v-structure. These
v-structures are together used to construct the final output. To address the
fundamental question of SCL, we propose a principled method for ML4C
featurization: we exploit the vicinity of a given UT (i.e., the neighbors of UT
in skeleton), and derive features by considering the conditional dependencies
and structural entanglement within the vicinity. We further prove that ML4C is
asymptotically correct. Last but foremost, thorough experiments conducted on
benchmark datasets demonstrate that ML4C remarkably outperforms other
state-of-the-art algorithms in terms of accuracy, reliability, robustness and
tolerance. In summary, ML4C shows promising results on validating the
effectiveness of supervision for causal learning. Our codes are publicly
available at https://github.com/microsoft/ML4C.
|
2110.11073 | Kai Wang | Kai Wang, Zhene Zou, Minghao Zhao, Qilin Deng, Yue Shang, Yile Liang,
Runze Wu, Xudong Shen, Tangjie Lyu, Changjie Fan | RL4RS: A Real-World Dataset for Reinforcement Learning based Recommender
System | 4-th version, SIGIR2023 | null | null | null | cs.IR cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Reinforcement learning based recommender systems (RL-based RS) aim at
learning a good policy from a batch of collected data, by casting
recommendations to multi-step decision-making tasks. However, current RL-based
RS research commonly has a large reality gap. In this paper, we introduce the
first open-source real-world dataset, RL4RS, hoping to replace the artificial
datasets and semi-simulated RS datasets previous studies used due to the
resource limitation of the RL-based RS domain. Unlike academic RL research,
RL-based RS suffers from the difficulties of being well-validated before
deployment. We attempt to propose a new systematic evaluation framework,
including evaluation of environment simulation, evaluation on environments,
counterfactual policy evaluation, and evaluation on environments built from
test set. In summary, the RL4RS (Reinforcement Learning for Recommender
Systems), a new resource with special concerns on the reality gaps, contains
two real-world datasets, data understanding tools, tuned simulation
environments, related advanced RL baselines, batch RL baselines, and
counterfactual policy evaluation algorithms. The RL4RS suite can be found at
https://github.com/fuxiAIlab/RL4RS. In addition to the RL-based recommender
systems, we expect the resource to contribute to research in applied
reinforcement learning.
| [
{
"version": "v1",
"created": "Mon, 18 Oct 2021 12:48:02 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Feb 2022 09:12:53 GMT"
},
{
"version": "v3",
"created": "Wed, 16 Feb 2022 03:31:15 GMT"
},
{
"version": "v4",
"created": "Sun, 20 Feb 2022 13:08:08 GMT"
},
{
"version": "v5",
"created": "Mon, 17 Apr 2023 10:37:38 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Wang",
"Kai",
""
],
[
"Zou",
"Zhene",
""
],
[
"Zhao",
"Minghao",
""
],
[
"Deng",
"Qilin",
""
],
[
"Shang",
"Yue",
""
],
[
"Liang",
"Yile",
""
],
[
"Wu",
"Runze",
""
],
[
"Shen",
"Xudong",
""
],
[
"Lyu",
"Tangjie",
""
],
[
"Fan",
"Changjie",
""
]
] | TITLE: RL4RS: A Real-World Dataset for Reinforcement Learning based Recommender
System
ABSTRACT: Reinforcement learning based recommender systems (RL-based RS) aim at
learning a good policy from a batch of collected data, by casting
recommendations to multi-step decision-making tasks. However, current RL-based
RS research commonly has a large reality gap. In this paper, we introduce the
first open-source real-world dataset, RL4RS, hoping to replace the artificial
datasets and semi-simulated RS datasets previous studies used due to the
resource limitation of the RL-based RS domain. Unlike academic RL research,
RL-based RS suffers from the difficulties of being well-validated before
deployment. We attempt to propose a new systematic evaluation framework,
including evaluation of environment simulation, evaluation on environments,
counterfactual policy evaluation, and evaluation on environments built from
test set. In summary, the RL4RS (Reinforcement Learning for Recommender
Systems), a new resource with special concerns on the reality gaps, contains
two real-world datasets, data understanding tools, tuned simulation
environments, related advanced RL baselines, batch RL baselines, and
counterfactual policy evaluation algorithms. The RL4RS suite can be found at
https://github.com/fuxiAIlab/RL4RS. In addition to the RL-based recommender
systems, we expect the resource to contribute to research in applied
reinforcement learning.
|
2111.04178 | Fivos Kalogiannis | Fivos Kalogiannis, Ioannis Panageas, Emmanouil-Vasileios
Vlatakis-Gkaragkounis | Towards convergence to Nash equilibria in two-team zero-sum games | Paper accepted in ICLR 2023 | null | null | null | cs.GT cs.LG math.OC | http://creativecommons.org/licenses/by/4.0/ | Contemporary applications of machine learning in two-team e-sports and the
superior expressivity of multi-agent generative adversarial networks raise
important and overlooked theoretical questions regarding optimization in
two-team games. Formally, two-team zero-sum games are defined as multi-player
games where players are split into two competing sets of agents, each
experiencing a utility identical to that of their teammates and opposite to
that of the opposing team. We focus on the solution concept of Nash equilibria
(NE). We first show that computing NE for this class of games is
$\textit{hard}$ for the complexity class ${\mathrm{CLS}}$. To further examine
the capabilities of online learning algorithms in games with full-information
feedback, we propose a benchmark of a simple -- yet nontrivial -- family of
such games. These games do not enjoy the properties used to prove convergence
for relevant algorithms. In particular, we use a dynamical systems perspective
to demonstrate that gradient descent-ascent, its optimistic variant, optimistic
multiplicative weights update, and extra gradient fail to converge (even
locally) to a Nash equilibrium. On a brighter note, we propose a first-order
method that leverages control theory techniques and under some conditions
enjoys last-iterate local convergence to a Nash equilibrium. We also believe
our proposed method is of independent interest for general min-max
optimization.
| [
{
"version": "v1",
"created": "Sun, 7 Nov 2021 21:15:35 GMT"
},
{
"version": "v2",
"created": "Tue, 9 Nov 2021 11:07:16 GMT"
},
{
"version": "v3",
"created": "Mon, 29 Nov 2021 21:46:30 GMT"
},
{
"version": "v4",
"created": "Mon, 17 Apr 2023 01:57:55 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Kalogiannis",
"Fivos",
""
],
[
"Panageas",
"Ioannis",
""
],
[
"Vlatakis-Gkaragkounis",
"Emmanouil-Vasileios",
""
]
] | TITLE: Towards convergence to Nash equilibria in two-team zero-sum games
ABSTRACT: Contemporary applications of machine learning in two-team e-sports and the
superior expressivity of multi-agent generative adversarial networks raise
important and overlooked theoretical questions regarding optimization in
two-team games. Formally, two-team zero-sum games are defined as multi-player
games where players are split into two competing sets of agents, each
experiencing a utility identical to that of their teammates and opposite to
that of the opposing team. We focus on the solution concept of Nash equilibria
(NE). We first show that computing NE for this class of games is
$\textit{hard}$ for the complexity class ${\mathrm{CLS}}$. To further examine
the capabilities of online learning algorithms in games with full-information
feedback, we propose a benchmark of a simple -- yet nontrivial -- family of
such games. These games do not enjoy the properties used to prove convergence
for relevant algorithms. In particular, we use a dynamical systems perspective
to demonstrate that gradient descent-ascent, its optimistic variant, optimistic
multiplicative weights update, and extra gradient fail to converge (even
locally) to a Nash equilibrium. On a brighter note, we propose a first-order
method that leverages control theory techniques and under some conditions
enjoys last-iterate local convergence to a Nash equilibrium. We also believe
our proposed method is of independent interest for general min-max
optimization.
|
2111.10756 | Keng Ji Chow | Keng Ji Chow, Samson Tan, Min-Yen Kan | TraVLR: Now You See It, Now You Don't! A Bimodal Dataset for Evaluating
Visio-Linguistic Reasoning | The first two authors contributed equally | null | null | null | cs.CL cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Numerous visio-linguistic (V+L) representation learning methods have been
developed, yet existing datasets do not adequately evaluate the extent to which
they represent visual and linguistic concepts in a unified space. We propose
several novel evaluation settings for V+L models, including cross-modal
transfer. Furthermore, existing V+L benchmarks often report global accuracy
scores on the entire dataset, making it difficult to pinpoint the specific
reasoning tasks that models fail and succeed at. We present TraVLR, a synthetic
dataset comprising four V+L reasoning tasks. TraVLR's synthetic nature allows
us to constrain its training and testing distributions along task-relevant
dimensions, enabling the evaluation of out-of-distribution generalisation. Each
example in TraVLR redundantly encodes the scene in two modalities, allowing
either to be dropped or added during training or testing without losing
relevant information. We compare the performance of four state-of-the-art V+L
models, finding that while they perform well on test examples from the same
modality, they all fail at cross-modal transfer and have limited success
accommodating the addition or deletion of one modality. We release TraVLR as an
open challenge for the research community.
| [
{
"version": "v1",
"created": "Sun, 21 Nov 2021 07:22:44 GMT"
},
{
"version": "v2",
"created": "Sat, 4 Mar 2023 12:57:41 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Apr 2023 09:48:44 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Chow",
"Keng Ji",
""
],
[
"Tan",
"Samson",
""
],
[
"Kan",
"Min-Yen",
""
]
] | TITLE: TraVLR: Now You See It, Now You Don't! A Bimodal Dataset for Evaluating
Visio-Linguistic Reasoning
ABSTRACT: Numerous visio-linguistic (V+L) representation learning methods have been
developed, yet existing datasets do not adequately evaluate the extent to which
they represent visual and linguistic concepts in a unified space. We propose
several novel evaluation settings for V+L models, including cross-modal
transfer. Furthermore, existing V+L benchmarks often report global accuracy
scores on the entire dataset, making it difficult to pinpoint the specific
reasoning tasks that models fail and succeed at. We present TraVLR, a synthetic
dataset comprising four V+L reasoning tasks. TraVLR's synthetic nature allows
us to constrain its training and testing distributions along task-relevant
dimensions, enabling the evaluation of out-of-distribution generalisation. Each
example in TraVLR redundantly encodes the scene in two modalities, allowing
either to be dropped or added during training or testing without losing
relevant information. We compare the performance of four state-of-the-art V+L
models, finding that while they perform well on test examples from the same
modality, they all fail at cross-modal transfer and have limited success
accommodating the addition or deletion of one modality. We release TraVLR as an
open challenge for the research community.
|
2111.11969 | Qiang Nie | Qiang Nie, Ziwei Liu, Yunhui Liu | Lifting 2D Human Pose to 3D with Domain Adapted 3D Body Concept | 15 pages, a paper submitted to IJCV | Int J Comput Vis 131 (2023) 1250 - 1268 | 10.1007/s11263-023-01749-2 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Lifting the 2D human pose to the 3D pose is an important yet challenging
task. Existing 3D pose estimation suffers from 1) the inherent ambiguity
between the 2D and 3D data, and 2) the lack of well labeled 2D-3D pose pairs in
the wild. Human beings are able to imagine the human 3D pose from a 2D image or
a set of 2D body key-points with the least ambiguity, which should be
attributed to the prior knowledge of the human body that we have acquired in
our mind. Inspired by this, we propose a new framework that leverages the
labeled 3D human poses to learn a 3D concept of the human body to reduce the
ambiguity. To have consensus on the body concept from 2D pose, our key insight
is to treat the 2D human pose and the 3D human pose as two different domains.
By adapting the two domains, the body knowledge learned from 3D poses is
applied to 2D poses and guides the 2D pose encoder to generate informative 3D
"imagination" as embedding in pose lifting. Benefiting from the domain
adaptation perspective, the proposed framework unifies the supervised and
semi-supervised 3D pose estimation in a principled framework. Extensive
experiments demonstrate that the proposed approach can achieve state-of-the-art
performance on standard benchmarks. More importantly, it is validated that the
explicitly learned 3D body concept effectively alleviates the 2D-3D ambiguity
in 2D pose lifting, improves the generalization, and enables the network to
exploit the abundant unlabeled 2D data.
| [
{
"version": "v1",
"created": "Tue, 23 Nov 2021 16:02:12 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Nie",
"Qiang",
""
],
[
"Liu",
"Ziwei",
""
],
[
"Liu",
"Yunhui",
""
]
] | TITLE: Lifting 2D Human Pose to 3D with Domain Adapted 3D Body Concept
ABSTRACT: Lifting the 2D human pose to the 3D pose is an important yet challenging
task. Existing 3D pose estimation suffers from 1) the inherent ambiguity
between the 2D and 3D data, and 2) the lack of well labeled 2D-3D pose pairs in
the wild. Human beings are able to imagine the human 3D pose from a 2D image or
a set of 2D body key-points with the least ambiguity, which should be
attributed to the prior knowledge of the human body that we have acquired in
our mind. Inspired by this, we propose a new framework that leverages the
labeled 3D human poses to learn a 3D concept of the human body to reduce the
ambiguity. To have consensus on the body concept from 2D pose, our key insight
is to treat the 2D human pose and the 3D human pose as two different domains.
By adapting the two domains, the body knowledge learned from 3D poses is
applied to 2D poses and guides the 2D pose encoder to generate informative 3D
"imagination" as embedding in pose lifting. Benefiting from the domain
adaptation perspective, the proposed framework unifies the supervised and
semi-supervised 3D pose estimation in a principled framework. Extensive
experiments demonstrate that the proposed approach can achieve state-of-the-art
performance on standard benchmarks. More importantly, it is validated that the
explicitly learned 3D body concept effectively alleviates the 2D-3D ambiguity
in 2D pose lifting, improves the generalization, and enables the network to
exploit the abundant unlabeled 2D data.
|
2112.04079 | Li-Hsiang Shen | Li-Hsiang Shen, Yung-Ting Huang, Kai-Ten Feng | CoMP-Enhanced Flexible Functional Split for Mixed Services in Beyond 5G
Wireless Networks | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With explosively escalating service demands, beyond fifth generation (B5G)
aims to realize various requirements for multi-service networks, i.e., higher
performance of mixed enhanced mobile broadband (eMBB) and ultra-reliable
low-latency communication (URLLC) services than 5G. To flexibly serve diverse
traffic, various functional split options (FSOs) are specified by 5G protocols
enabling different network functions. In order to improve signal qualities for
edge users, we consider FSO-based coordinated multi-point (CoMP) transmission
as a prominent technique capable of supporting high traffic demands. However,
due to conventional confined hardware processing capability, a processor
sharing (PS) model is introduced to deal with high latency for multi-service
FSO-based networks. Therefore, it becomes essential to assign CoMP-enhanced
functional split modes under PS model. A more tractable FSO-based network in
terms of ergodic rate and reliability is derived by stochastic geometry
approach. Moreover, we have proposed CoMP-enhanced functional split mode
allocation (CFSMA) scheme to adaptively assign FSOs to provide enhanced mixed
throughput and latency-aware services. The simulation results have validated
analytical derivation and demonstrated that the proposed CFSMA scheme optimizes
system spectrum efficiency while guaranteeing stringent latency requirement.
The proposed CFSMA scheme with the designed PS FFS-CoMP system outperforms the
benchmarks of conventional FCFS scheduling, non-FSO network, fixed FSOs, and
limited available FSO selections in open literature.
| [
{
"version": "v1",
"created": "Wed, 8 Dec 2021 02:02:30 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Apr 2023 21:25:21 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Shen",
"Li-Hsiang",
""
],
[
"Huang",
"Yung-Ting",
""
],
[
"Feng",
"Kai-Ten",
""
]
] | TITLE: CoMP-Enhanced Flexible Functional Split for Mixed Services in Beyond 5G
Wireless Networks
ABSTRACT: With explosively escalating service demands, beyond fifth generation (B5G)
aims to realize various requirements for multi-service networks, i.e., higher
performance of mixed enhanced mobile broadband (eMBB) and ultra-reliable
low-latency communication (URLLC) services than 5G. To flexibly serve diverse
traffic, various functional split options (FSOs) are specified by 5G protocols
enabling different network functions. In order to improve signal qualities for
edge users, we consider FSO-based coordinated multi-point (CoMP) transmission
as a prominent technique capable of supporting high traffic demands. However,
due to conventional confined hardware processing capability, a processor
sharing (PS) model is introduced to deal with high latency for multi-service
FSO-based networks. Therefore, it becomes essential to assign CoMP-enhanced
functional split modes under PS model. A more tractable FSO-based network in
terms of ergodic rate and reliability is derived by stochastic geometry
approach. Moreover, we have proposed CoMP-enhanced functional split mode
allocation (CFSMA) scheme to adaptively assign FSOs to provide enhanced mixed
throughput and latency-aware services. The simulation results have validated
analytical derivation and demonstrated that the proposed CFSMA scheme optimizes
system spectrum efficiency while guaranteeing stringent latency requirement.
The proposed CFSMA scheme with the designed PS FFS-CoMP system outperforms the
benchmarks of conventional FCFS scheduling, non-FSO network, fixed FSOs, and
limited available FSO selections in open literature.
|
2112.06437 | Jiachen Xu | Jiachen Xu, Junlin Guo, James Zimmer-Dauphinee, Quan Liu, Yuxuan Shi,
Zuhayr Asad, D. Mitchell Wilkes, Parker VanValkenburgh, Steven A. Wernke,
Yuankai Huo | Semi-Supervised Contrastive Learning for Remote Sensing: Identifying
Ancient Urbanization in the South Central Andes | null | null | null | null | cs.CV cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Archaeology has long faced fundamental issues of sampling and scalar
representation. Traditionally, the local-to-regional-scale views of settlement
patterns are produced through systematic pedestrian surveys. Recently,
systematic manual survey of satellite and aerial imagery has enabled continuous
distributional views of archaeological phenomena at interregional scales.
However, such 'brute force' manual imagery survey methods are both time- and
labor-intensive, as well as prone to inter-observer differences in sensitivity
and specificity. The development of self-supervised learning methods offers a
scalable learning scheme for locating archaeological features using unlabeled
satellite and historical aerial images. However, archaeological features are
generally only visible in a very small proportion relative to the landscape,
while the modern contrastive-supervised learning approach typically yields an
inferior performance on highly imbalanced datasets. In this work, we propose a
framework to address this long-tail problem. As opposed to the existing
contrastive learning approaches that treat the labelled and unlabeled data
separately, our proposed method reforms the learning paradigm under a
semi-supervised setting in order to utilize the precious annotated data (<7% in
our setting). Specifically, the highly unbalanced nature of the data is
employed as the prior knowledge in order to form pseudo negative pairs by
ranking the similarities between unannotated image patches and annotated anchor
images. In this study, we used 95,358 unlabeled images and 5,830 labelled
images in order to solve the issues associated with detecting ancient buildings
from a long-tailed satellite image dataset. From the results, our
semi-supervised contrastive learning model achieved a promising testing
balanced accuracy of 79.0%, which is a 3.8% improvement as compared to other
state-of-the-art approaches.
| [
{
"version": "v1",
"created": "Mon, 13 Dec 2021 06:26:47 GMT"
},
{
"version": "v2",
"created": "Sat, 15 Apr 2023 13:40:14 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Xu",
"Jiachen",
""
],
[
"Guo",
"Junlin",
""
],
[
"Zimmer-Dauphinee",
"James",
""
],
[
"Liu",
"Quan",
""
],
[
"Shi",
"Yuxuan",
""
],
[
"Asad",
"Zuhayr",
""
],
[
"Wilkes",
"D. Mitchell",
""
],
[
"VanValkenburgh",
"Parker",
""
],
[
"Wernke",
"Steven A.",
""
],
[
"Huo",
"Yuankai",
""
]
] | TITLE: Semi-Supervised Contrastive Learning for Remote Sensing: Identifying
Ancient Urbanization in the South Central Andes
ABSTRACT: Archaeology has long faced fundamental issues of sampling and scalar
representation. Traditionally, the local-to-regional-scale views of settlement
patterns are produced through systematic pedestrian surveys. Recently,
systematic manual survey of satellite and aerial imagery has enabled continuous
distributional views of archaeological phenomena at interregional scales.
However, such 'brute force' manual imagery survey methods are both time- and
labor-intensive, as well as prone to inter-observer differences in sensitivity
and specificity. The development of self-supervised learning methods offers a
scalable learning scheme for locating archaeological features using unlabeled
satellite and historical aerial images. However, archaeological features are
generally only visible in a very small proportion relative to the landscape,
while the modern contrastive-supervised learning approach typically yields an
inferior performance on highly imbalanced datasets. In this work, we propose a
framework to address this long-tail problem. As opposed to the existing
contrastive learning approaches that treat the labelled and unlabeled data
separately, our proposed method reforms the learning paradigm under a
semi-supervised setting in order to utilize the precious annotated data (<7% in
our setting). Specifically, the highly unbalanced nature of the data is
employed as the prior knowledge in order to form pseudo negative pairs by
ranking the similarities between unannotated image patches and annotated anchor
images. In this study, we used 95,358 unlabeled images and 5,830 labelled
images in order to solve the issues associated with detecting ancient buildings
from a long-tailed satellite image dataset. From the results, our
semi-supervised contrastive learning model achieved a promising testing
balanced accuracy of 79.0%, which is a 3.8% improvement as compared to other
state-of-the-art approaches.
|
2201.11460 | Yuren Cong | Yuren Cong, Michael Ying Yang, Bodo Rosenhahn | RelTR: Relation Transformer for Scene Graph Generation | accepted by IEEE Transactions on Pattern Analysis and Machine
Intelligence | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Different objects in the same scene are more or less related to each other,
but only a limited number of these relationships are noteworthy. Inspired by
DETR, which excels in object detection, we view scene graph generation as a set
prediction problem and propose an end-to-end scene graph generation model RelTR
which has an encoder-decoder architecture. The encoder reasons about the visual
feature context while the decoder infers a fixed-size set of triplets
subject-predicate-object using different types of attention mechanisms with
coupled subject and object queries. We design a set prediction loss performing
the matching between the ground truth and predicted triplets for the end-to-end
training. In contrast to most existing scene graph generation methods, RelTR is
a one-stage method that predicts a set of relationships directly only using
visual appearance without combining entities and labeling all possible
predicates. Extensive experiments on the Visual Genome and Open Images V6
datasets demonstrate the superior performance and fast inference of our model.
| [
{
"version": "v1",
"created": "Thu, 27 Jan 2022 11:53:41 GMT"
},
{
"version": "v2",
"created": "Thu, 11 Aug 2022 20:17:58 GMT"
},
{
"version": "v3",
"created": "Fri, 14 Apr 2023 21:44:13 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Cong",
"Yuren",
""
],
[
"Yang",
"Michael Ying",
""
],
[
"Rosenhahn",
"Bodo",
""
]
] | TITLE: RelTR: Relation Transformer for Scene Graph Generation
ABSTRACT: Different objects in the same scene are more or less related to each other,
but only a limited number of these relationships are noteworthy. Inspired by
DETR, which excels in object detection, we view scene graph generation as a set
prediction problem and propose an end-to-end scene graph generation model RelTR
which has an encoder-decoder architecture. The encoder reasons about the visual
feature context while the decoder infers a fixed-size set of triplets
subject-predicate-object using different types of attention mechanisms with
coupled subject and object queries. We design a set prediction loss performing
the matching between the ground truth and predicted triplets for the end-to-end
training. In contrast to most existing scene graph generation methods, RelTR is
a one-stage method that predicts a set of relationships directly only using
visual appearance without combining entities and labeling all possible
predicates. Extensive experiments on the Visual Genome and Open Images V6
datasets demonstrate the superior performance and fast inference of our model.
|
2203.08810 | Tiantian Feng | Tiantian Feng and Shrikanth Narayanan | Semi-FedSER: Semi-supervised Learning for Speech Emotion Recognition On
Federated Learning using Multiview Pseudo-Labeling | This paper was submitted to Insterspeech 2022 for review | Proc. Interspeech 2022 | 10.21437/Interspeech.2022-141 | null | eess.AS cs.CR cs.LG cs.SD | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Speech Emotion Recognition (SER) application is frequently associated with
privacy concerns as it often acquires and transmits speech data at the
client-side to remote cloud platforms for further processing. These speech data
can reveal not only speech content and affective information but the speaker's
identity, demographic traits, and health status. Federated learning (FL) is a
distributed machine learning algorithm that coordinates clients to train a
model collaboratively without sharing local data. This algorithm shows enormous
potential for SER applications as sharing raw speech or speech features from a
user's device is vulnerable to privacy attacks. However, a major challenge in
FL is limited availability of high-quality labeled data samples. In this work,
we propose a semi-supervised federated learning framework, Semi-FedSER, that
utilizes both labeled and unlabeled data samples to address the challenge of
limited labeled data samples in FL. We show that our Semi-FedSER can generate
desired SER performance even when the local label rate l=20 using two SER
benchmark datasets: IEMOCAP and MSP-Improv.
| [
{
"version": "v1",
"created": "Tue, 15 Mar 2022 21:50:43 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Feng",
"Tiantian",
""
],
[
"Narayanan",
"Shrikanth",
""
]
] | TITLE: Semi-FedSER: Semi-supervised Learning for Speech Emotion Recognition On
Federated Learning using Multiview Pseudo-Labeling
ABSTRACT: Speech Emotion Recognition (SER) application is frequently associated with
privacy concerns as it often acquires and transmits speech data at the
client-side to remote cloud platforms for further processing. These speech data
can reveal not only speech content and affective information but the speaker's
identity, demographic traits, and health status. Federated learning (FL) is a
distributed machine learning algorithm that coordinates clients to train a
model collaboratively without sharing local data. This algorithm shows enormous
potential for SER applications as sharing raw speech or speech features from a
user's device is vulnerable to privacy attacks. However, a major challenge in
FL is limited availability of high-quality labeled data samples. In this work,
we propose a semi-supervised federated learning framework, Semi-FedSER, that
utilizes both labeled and unlabeled data samples to address the challenge of
limited labeled data samples in FL. We show that our Semi-FedSER can generate
desired SER performance even when the local label rate l=20 using two SER
benchmark datasets: IEMOCAP and MSP-Improv.
|
2203.08897 | Swathikiran Sudhakaran | Swathikiran Sudhakaran, Sergio Escalera, Oswald Lanz | Gate-Shift-Fuse for Video Action Recognition | Accepted to TPAMI. arXiv admin note: text overlap with
arXiv:1912.00381 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Convolutional Neural Networks are the de facto models for image recognition.
However 3D CNNs, the straight forward extension of 2D CNNs for video
recognition, have not achieved the same success on standard action recognition
benchmarks. One of the main reasons for this reduced performance of 3D CNNs is
the increased computational complexity requiring large scale annotated datasets
to train them in scale. 3D kernel factorization approaches have been proposed
to reduce the complexity of 3D CNNs. Existing kernel factorization approaches
follow hand-designed and hard-wired techniques. In this paper we propose
Gate-Shift-Fuse (GSF), a novel spatio-temporal feature extraction module which
controls interactions in spatio-temporal decomposition and learns to adaptively
route features through time and combine them in a data dependent manner. GSF
leverages grouped spatial gating to decompose input tensor and channel
weighting to fuse the decomposed tensors. GSF can be inserted into existing 2D
CNNs to convert them into an efficient and high performing spatio-temporal
feature extractor, with negligible parameter and compute overhead. We perform
an extensive analysis of GSF using two popular 2D CNN families and achieve
state-of-the-art or competitive performance on five standard action recognition
benchmarks.
| [
{
"version": "v1",
"created": "Wed, 16 Mar 2022 19:19:04 GMT"
},
{
"version": "v2",
"created": "Thu, 13 Apr 2023 17:27:02 GMT"
},
{
"version": "v3",
"created": "Sat, 15 Apr 2023 13:06:27 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Sudhakaran",
"Swathikiran",
""
],
[
"Escalera",
"Sergio",
""
],
[
"Lanz",
"Oswald",
""
]
] | TITLE: Gate-Shift-Fuse for Video Action Recognition
ABSTRACT: Convolutional Neural Networks are the de facto models for image recognition.
However 3D CNNs, the straight forward extension of 2D CNNs for video
recognition, have not achieved the same success on standard action recognition
benchmarks. One of the main reasons for this reduced performance of 3D CNNs is
the increased computational complexity requiring large scale annotated datasets
to train them in scale. 3D kernel factorization approaches have been proposed
to reduce the complexity of 3D CNNs. Existing kernel factorization approaches
follow hand-designed and hard-wired techniques. In this paper we propose
Gate-Shift-Fuse (GSF), a novel spatio-temporal feature extraction module which
controls interactions in spatio-temporal decomposition and learns to adaptively
route features through time and combine them in a data dependent manner. GSF
leverages grouped spatial gating to decompose input tensor and channel
weighting to fuse the decomposed tensors. GSF can be inserted into existing 2D
CNNs to convert them into an efficient and high performing spatio-temporal
feature extractor, with negligible parameter and compute overhead. We perform
an extensive analysis of GSF using two popular 2D CNN families and achieve
state-of-the-art or competitive performance on five standard action recognition
benchmarks.
|
2203.10642 | Xuanyao Chen | Xuanyao Chen, Tianyuan Zhang, Yue Wang, Yilun Wang, Hang Zhao | FUTR3D: A Unified Sensor Fusion Framework for 3D Detection | null | CVPR 2023 workshop on autonomous driving | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Sensor fusion is an essential topic in many perception systems, such as
autonomous driving and robotics. Existing multi-modal 3D detection models
usually involve customized designs depending on the sensor combinations or
setups. In this work, we propose the first unified end-to-end sensor fusion
framework for 3D detection, named FUTR3D, which can be used in (almost) any
sensor configuration. FUTR3D employs a query-based Modality-Agnostic Feature
Sampler (MAFS), together with a transformer decoder with a set-to-set loss for
3D detection, thus avoiding using late fusion heuristics and post-processing
tricks. We validate the effectiveness of our framework on various combinations
of cameras, low-resolution LiDARs, high-resolution LiDARs, and Radars. On
NuScenes dataset, FUTR3D achieves better performance over specifically designed
methods across different sensor combinations. Moreover, FUTR3D achieves great
flexibility with different sensor configurations and enables low-cost
autonomous driving. For example, only using a 4-beam LiDAR with cameras, FUTR3D
(58.0 mAP) achieves on par performance with state-of-the-art 3D detection model
CenterPoint (56.6 mAP) using a 32-beam LiDAR.
| [
{
"version": "v1",
"created": "Sun, 20 Mar 2022 20:41:55 GMT"
},
{
"version": "v2",
"created": "Sat, 15 Apr 2023 13:05:44 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Chen",
"Xuanyao",
""
],
[
"Zhang",
"Tianyuan",
""
],
[
"Wang",
"Yue",
""
],
[
"Wang",
"Yilun",
""
],
[
"Zhao",
"Hang",
""
]
] | TITLE: FUTR3D: A Unified Sensor Fusion Framework for 3D Detection
ABSTRACT: Sensor fusion is an essential topic in many perception systems, such as
autonomous driving and robotics. Existing multi-modal 3D detection models
usually involve customized designs depending on the sensor combinations or
setups. In this work, we propose the first unified end-to-end sensor fusion
framework for 3D detection, named FUTR3D, which can be used in (almost) any
sensor configuration. FUTR3D employs a query-based Modality-Agnostic Feature
Sampler (MAFS), together with a transformer decoder with a set-to-set loss for
3D detection, thus avoiding using late fusion heuristics and post-processing
tricks. We validate the effectiveness of our framework on various combinations
of cameras, low-resolution LiDARs, high-resolution LiDARs, and Radars. On
NuScenes dataset, FUTR3D achieves better performance over specifically designed
methods across different sensor combinations. Moreover, FUTR3D achieves great
flexibility with different sensor configurations and enables low-cost
autonomous driving. For example, only using a 4-beam LiDAR with cameras, FUTR3D
(58.0 mAP) achieves on par performance with state-of-the-art 3D detection model
CenterPoint (56.6 mAP) using a 32-beam LiDAR.
|
2203.11200 | Jie Chen | Jie Chen, Shouzhen Chen, Junbin Gao, Zengfeng Huang, Junping Zhang and
Jian Pu | Exploiting Neighbor Effect: Conv-Agnostic GNNs Framework for Graphs with
Heterophily | accepted by TNNLS 2023 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Due to the homophily assumption in graph convolution networks (GNNs), a
common consensus in the graph node classification task is that GNNs perform
well on homophilic graphs but may fail on heterophilic graphs with many
inter-class edges. However, the previous inter-class edges perspective and
related homo-ratio metrics cannot well explain the GNNs performance under some
heterophilic datasets, which implies that not all the inter-class edges are
harmful to GNNs. In this work, we propose a new metric based on von Neumann
entropy to re-examine the heterophily problem of GNNs and investigate the
feature aggregation of inter-class edges from an entire neighbor identifiable
perspective. Moreover, we propose a simple yet effective Conv-Agnostic GNN
framework (CAGNNs) to enhance the performance of most GNNs on heterophily
datasets by learning the neighbor effect for each node. Specifically, we first
decouple the feature of each node into the discriminative feature for
downstream tasks and the aggregation feature for graph convolution. Then, we
propose a shared mixer module to adaptively evaluate the neighbor effect of
each node to incorporate the neighbor information. The proposed framework can
be regarded as a plug-in component and is compatible with most GNNs. The
experimental results over nine well-known benchmark datasets indicate that our
framework can significantly improve performance, especially for the heterophily
graphs. The average performance gain is 9.81%, 25.81%, and 20.61% compared with
GIN, GAT, and GCN, respectively. Extensive ablation studies and robustness
analysis further verify the effectiveness, robustness, and interpretability of
our framework. Code is available at https://github.com/JC-202/CAGNN.
| [
{
"version": "v1",
"created": "Sat, 19 Mar 2022 14:26:43 GMT"
},
{
"version": "v2",
"created": "Sat, 9 Apr 2022 03:30:44 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Apr 2023 04:21:34 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Chen",
"Jie",
""
],
[
"Chen",
"Shouzhen",
""
],
[
"Gao",
"Junbin",
""
],
[
"Huang",
"Zengfeng",
""
],
[
"Zhang",
"Junping",
""
],
[
"Pu",
"Jian",
""
]
] | TITLE: Exploiting Neighbor Effect: Conv-Agnostic GNNs Framework for Graphs with
Heterophily
ABSTRACT: Due to the homophily assumption in graph convolution networks (GNNs), a
common consensus in the graph node classification task is that GNNs perform
well on homophilic graphs but may fail on heterophilic graphs with many
inter-class edges. However, the previous inter-class edges perspective and
related homo-ratio metrics cannot well explain the GNNs performance under some
heterophilic datasets, which implies that not all the inter-class edges are
harmful to GNNs. In this work, we propose a new metric based on von Neumann
entropy to re-examine the heterophily problem of GNNs and investigate the
feature aggregation of inter-class edges from an entire neighbor identifiable
perspective. Moreover, we propose a simple yet effective Conv-Agnostic GNN
framework (CAGNNs) to enhance the performance of most GNNs on heterophily
datasets by learning the neighbor effect for each node. Specifically, we first
decouple the feature of each node into the discriminative feature for
downstream tasks and the aggregation feature for graph convolution. Then, we
propose a shared mixer module to adaptively evaluate the neighbor effect of
each node to incorporate the neighbor information. The proposed framework can
be regarded as a plug-in component and is compatible with most GNNs. The
experimental results over nine well-known benchmark datasets indicate that our
framework can significantly improve performance, especially for the heterophily
graphs. The average performance gain is 9.81%, 25.81%, and 20.61% compared with
GIN, GAT, and GCN, respectively. Extensive ablation studies and robustness
analysis further verify the effectiveness, robustness, and interpretability of
our framework. Code is available at https://github.com/JC-202/CAGNN.
|
2203.16828 | Sihan Ma | Sihan Ma, Jizhizi Li, Jing Zhang, He Zhang, Dacheng Tao | Rethinking Portrait Matting with Privacy Preserving | Accepted to the International Journal of Computer Vision (IJCV). The
code, dataset, and models are available at
https://github.com/ViTAE-Transformer/P3M-Net. arXiv admin note: substantial
text overlap with arXiv:2104.14222 | null | null | null | cs.CV cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, there has been an increasing concern about the privacy issue raised
by identifiable information in machine learning. However, previous portrait
matting methods were all based on identifiable images. To fill the gap, we
present P3M-10k, which is the first large-scale anonymized benchmark for
Privacy-Preserving Portrait Matting (P3M). P3M-10k consists of 10,421 high
resolution face-blurred portrait images along with high-quality alpha mattes,
which enables us to systematically evaluate both trimap-free and trimap-based
matting methods and obtain some useful findings about model generalization
ability under the privacy preserving training (PPT) setting. We also present a
unified matting model dubbed P3M-Net that is compatible with both CNN and
transformer backbones. To further mitigate the cross-domain performance gap
issue under the PPT setting, we devise a simple yet effective Copy and Paste
strategy (P3M-CP), which borrows facial information from public celebrity
images and directs the network to reacquire the face context at both data and
feature level. Extensive experiments on P3M-10k and public benchmarks
demonstrate the superiority of P3M-Net over state-of-the-art methods and the
effectiveness of P3M-CP in improving the cross-domain generalization ability,
implying a great significance of P3M for future research and real-world
applications.
| [
{
"version": "v1",
"created": "Thu, 31 Mar 2022 06:26:07 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Apr 2023 00:19:30 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Ma",
"Sihan",
""
],
[
"Li",
"Jizhizi",
""
],
[
"Zhang",
"Jing",
""
],
[
"Zhang",
"He",
""
],
[
"Tao",
"Dacheng",
""
]
] | TITLE: Rethinking Portrait Matting with Privacy Preserving
ABSTRACT: Recently, there has been an increasing concern about the privacy issue raised
by identifiable information in machine learning. However, previous portrait
matting methods were all based on identifiable images. To fill the gap, we
present P3M-10k, which is the first large-scale anonymized benchmark for
Privacy-Preserving Portrait Matting (P3M). P3M-10k consists of 10,421 high
resolution face-blurred portrait images along with high-quality alpha mattes,
which enables us to systematically evaluate both trimap-free and trimap-based
matting methods and obtain some useful findings about model generalization
ability under the privacy preserving training (PPT) setting. We also present a
unified matting model dubbed P3M-Net that is compatible with both CNN and
transformer backbones. To further mitigate the cross-domain performance gap
issue under the PPT setting, we devise a simple yet effective Copy and Paste
strategy (P3M-CP), which borrows facial information from public celebrity
images and directs the network to reacquire the face context at both data and
feature level. Extensive experiments on P3M-10k and public benchmarks
demonstrate the superiority of P3M-Net over state-of-the-art methods and the
effectiveness of P3M-CP in improving the cross-domain generalization ability,
implying a great significance of P3M for future research and real-world
applications.
|
2204.12825 | Jihed Khiari | Jihed Khiari, Cristina Olaverri-Monreal | Uncertainty-Aware Prediction of Battery Energy Consumption for Hybrid
Electric Vehicles | null | null | 10.1109/IV51971.2022.9827458 | null | cs.LG cs.SY eess.SY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The usability of vehicles is highly dependent on their energy consumption. In
particular, one of the main factors hindering the mass adoption of electric
(EV), hybrid (HEV), and plug-in hybrid (PHEV) vehicles is range anxiety, which
occurs when a driver is uncertain about the availability of energy for a given
trip. To tackle this problem, we propose a machine learning approach for
modeling the battery energy consumption. By reducing predictive uncertainty,
this method can help increase trust in the vehicle's performance and thus boost
its usability. Most related work focuses on physical and/or chemical models of
the battery that affect the energy consumption. We propose a data-driven
approach which relies on real-world datasets including battery related
attributes. Our approach showed an improvement in terms of predictive
uncertainty as well as in accuracy compared to traditional methods.
| [
{
"version": "v1",
"created": "Wed, 27 Apr 2022 10:29:38 GMT"
},
{
"version": "v2",
"created": "Mon, 17 Apr 2023 10:20:18 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Khiari",
"Jihed",
""
],
[
"Olaverri-Monreal",
"Cristina",
""
]
] | TITLE: Uncertainty-Aware Prediction of Battery Energy Consumption for Hybrid
Electric Vehicles
ABSTRACT: The usability of vehicles is highly dependent on their energy consumption. In
particular, one of the main factors hindering the mass adoption of electric
(EV), hybrid (HEV), and plug-in hybrid (PHEV) vehicles is range anxiety, which
occurs when a driver is uncertain about the availability of energy for a given
trip. To tackle this problem, we propose a machine learning approach for
modeling the battery energy consumption. By reducing predictive uncertainty,
this method can help increase trust in the vehicle's performance and thus boost
its usability. Most related work focuses on physical and/or chemical models of
the battery that affect the energy consumption. We propose a data-driven
approach which relies on real-world datasets including battery related
attributes. Our approach showed an improvement in terms of predictive
uncertainty as well as in accuracy compared to traditional methods.
|
2204.12834 | Simon Weber | Simon Weber and Nikolaus Demmel and Tin Chon Chan and Daniel Cremers | Power Bundle Adjustment for Large-Scale 3D Reconstruction | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | We introduce Power Bundle Adjustment as an expansion type algorithm for
solving large-scale bundle adjustment problems. It is based on the power series
expansion of the inverse Schur complement and constitutes a new family of
solvers that we call inverse expansion methods. We theoretically justify the
use of power series and we prove the convergence of our approach. Using the
real-world BAL dataset we show that the proposed solver challenges the
state-of-the-art iterative methods and significantly accelerates the solution
of the normal equation, even for reaching a very high accuracy. This
easy-to-implement solver can also complement a recently presented distributed
bundle adjustment framework. We demonstrate that employing the proposed Power
Bundle Adjustment as a sub-problem solver significantly improves speed and
accuracy of the distributed optimization.
| [
{
"version": "v1",
"created": "Wed, 27 Apr 2022 10:38:33 GMT"
},
{
"version": "v2",
"created": "Tue, 3 May 2022 06:24:32 GMT"
},
{
"version": "v3",
"created": "Thu, 8 Dec 2022 16:07:55 GMT"
},
{
"version": "v4",
"created": "Mon, 17 Apr 2023 13:52:06 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Weber",
"Simon",
""
],
[
"Demmel",
"Nikolaus",
""
],
[
"Chan",
"Tin Chon",
""
],
[
"Cremers",
"Daniel",
""
]
] | TITLE: Power Bundle Adjustment for Large-Scale 3D Reconstruction
ABSTRACT: We introduce Power Bundle Adjustment as an expansion type algorithm for
solving large-scale bundle adjustment problems. It is based on the power series
expansion of the inverse Schur complement and constitutes a new family of
solvers that we call inverse expansion methods. We theoretically justify the
use of power series and we prove the convergence of our approach. Using the
real-world BAL dataset we show that the proposed solver challenges the
state-of-the-art iterative methods and significantly accelerates the solution
of the normal equation, even for reaching a very high accuracy. This
easy-to-implement solver can also complement a recently presented distributed
bundle adjustment framework. We demonstrate that employing the proposed Power
Bundle Adjustment as a sub-problem solver significantly improves speed and
accuracy of the distributed optimization.
|
2204.13686 | Zhongang Cai | Zhongang Cai, Daxuan Ren, Ailing Zeng, Zhengyu Lin, Tao Yu, Wenjia
Wang, Xiangyu Fan, Yang Gao, Yifan Yu, Liang Pan, Fangzhou Hong, Mingyuan
Zhang, Chen Change Loy, Lei Yang, Ziwei Liu | HuMMan: Multi-Modal 4D Human Dataset for Versatile Sensing and Modeling | Homepage: https://caizhongang.github.io/projects/HuMMan/ | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | 4D human sensing and modeling are fundamental tasks in vision and graphics
with numerous applications. With the advances of new sensors and algorithms,
there is an increasing demand for more versatile datasets. In this work, we
contribute HuMMan, a large-scale multi-modal 4D human dataset with 1000 human
subjects, 400k sequences and 60M frames. HuMMan has several appealing
properties: 1) multi-modal data and annotations including color images, point
clouds, keypoints, SMPL parameters, and textured meshes; 2) popular mobile
device is included in the sensor suite; 3) a set of 500 actions, designed to
cover fundamental movements; 4) multiple tasks such as action recognition, pose
estimation, parametric human recovery, and textured mesh reconstruction are
supported and evaluated. Extensive experiments on HuMMan voice the need for
further study on challenges such as fine-grained action recognition, dynamic
human mesh reconstruction, point cloud-based parametric human recovery, and
cross-device domain gaps.
| [
{
"version": "v1",
"created": "Thu, 28 Apr 2022 17:54:25 GMT"
},
{
"version": "v2",
"created": "Sun, 16 Apr 2023 12:26:14 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Cai",
"Zhongang",
""
],
[
"Ren",
"Daxuan",
""
],
[
"Zeng",
"Ailing",
""
],
[
"Lin",
"Zhengyu",
""
],
[
"Yu",
"Tao",
""
],
[
"Wang",
"Wenjia",
""
],
[
"Fan",
"Xiangyu",
""
],
[
"Gao",
"Yang",
""
],
[
"Yu",
"Yifan",
""
],
[
"Pan",
"Liang",
""
],
[
"Hong",
"Fangzhou",
""
],
[
"Zhang",
"Mingyuan",
""
],
[
"Loy",
"Chen Change",
""
],
[
"Yang",
"Lei",
""
],
[
"Liu",
"Ziwei",
""
]
] | TITLE: HuMMan: Multi-Modal 4D Human Dataset for Versatile Sensing and Modeling
ABSTRACT: 4D human sensing and modeling are fundamental tasks in vision and graphics
with numerous applications. With the advances of new sensors and algorithms,
there is an increasing demand for more versatile datasets. In this work, we
contribute HuMMan, a large-scale multi-modal 4D human dataset with 1000 human
subjects, 400k sequences and 60M frames. HuMMan has several appealing
properties: 1) multi-modal data and annotations including color images, point
clouds, keypoints, SMPL parameters, and textured meshes; 2) popular mobile
device is included in the sensor suite; 3) a set of 500 actions, designed to
cover fundamental movements; 4) multiple tasks such as action recognition, pose
estimation, parametric human recovery, and textured mesh reconstruction are
supported and evaluated. Extensive experiments on HuMMan voice the need for
further study on challenges such as fine-grained action recognition, dynamic
human mesh reconstruction, point cloud-based parametric human recovery, and
cross-device domain gaps.
|
2205.03805 | Yunqing Zhao | Yunqing Zhao, Henghui Ding, Houjing Huang, Ngai-Man Cheung | A Closer Look at Few-shot Image Generation | 12 figures, 4 tables, The IEEE / CVF Computer Vision and Pattern
Recognition Conference (CVPR) 2022 | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern GANs excel at generating high quality and diverse images. However,
when transferring the pretrained GANs on small target data (e.g., 10-shot), the
generator tends to replicate the training samples. Several methods have been
proposed to address this few-shot image generation task, but there is a lack of
effort to analyze them under a unified framework. As our first contribution, we
propose a framework to analyze existing methods during the adaptation. Our
analysis discovers that while some methods have disproportionate focus on
diversity preserving which impede quality improvement, all methods achieve
similar quality after convergence. Therefore, the better methods are those that
can slow down diversity degradation. Furthermore, our analysis reveals that
there is still plenty of room to further slow down diversity degradation.
Informed by our analysis and to slow down the diversity degradation of the
target generator during adaptation, our second contribution proposes to apply
mutual information (MI) maximization to retain the source domain's rich
multi-level diversity information in the target domain generator. We propose to
perform MI maximization by contrastive loss (CL), leverage the generator and
discriminator as two feature encoders to extract different multi-level features
for computing CL. We refer to our method as Dual Contrastive Learning (DCL).
Extensive experiments on several public datasets show that, while leading to a
slower diversity-degrading generator during adaptation, our proposed DCL brings
visually pleasant quality and state-of-the-art quantitative performance.
Project Page: yunqing-me.github.io/A-Closer-Look-at-FSIG.
| [
{
"version": "v1",
"created": "Sun, 8 May 2022 07:46:26 GMT"
},
{
"version": "v2",
"created": "Sat, 15 Apr 2023 14:51:43 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Zhao",
"Yunqing",
""
],
[
"Ding",
"Henghui",
""
],
[
"Huang",
"Houjing",
""
],
[
"Cheung",
"Ngai-Man",
""
]
] | TITLE: A Closer Look at Few-shot Image Generation
ABSTRACT: Modern GANs excel at generating high quality and diverse images. However,
when transferring the pretrained GANs on small target data (e.g., 10-shot), the
generator tends to replicate the training samples. Several methods have been
proposed to address this few-shot image generation task, but there is a lack of
effort to analyze them under a unified framework. As our first contribution, we
propose a framework to analyze existing methods during the adaptation. Our
analysis discovers that while some methods have disproportionate focus on
diversity preserving which impede quality improvement, all methods achieve
similar quality after convergence. Therefore, the better methods are those that
can slow down diversity degradation. Furthermore, our analysis reveals that
there is still plenty of room to further slow down diversity degradation.
Informed by our analysis and to slow down the diversity degradation of the
target generator during adaptation, our second contribution proposes to apply
mutual information (MI) maximization to retain the source domain's rich
multi-level diversity information in the target domain generator. We propose to
perform MI maximization by contrastive loss (CL), leverage the generator and
discriminator as two feature encoders to extract different multi-level features
for computing CL. We refer to our method as Dual Contrastive Learning (DCL).
Extensive experiments on several public datasets show that, while leading to a
slower diversity-degrading generator during adaptation, our proposed DCL brings
visually pleasant quality and state-of-the-art quantitative performance.
Project Page: yunqing-me.github.io/A-Closer-Look-at-FSIG.
|
2205.10625 | Denny Zhou | Denny Zhou, Nathanael Sch\"arli, Le Hou, Jason Wei, Nathan Scales,
Xuezhi Wang, Dale Schuurmans, Claire Cui, Olivier Bousquet, Quoc Le, Ed Chi | Least-to-Most Prompting Enables Complex Reasoning in Large Language
Models | ICLR 2023 | null | null | null | cs.AI cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Chain-of-thought prompting has demonstrated remarkable performance on various
natural language reasoning tasks. However, it tends to perform poorly on tasks
which requires solving problems harder than the exemplars shown in the prompts.
To overcome this challenge of easy-to-hard generalization, we propose a novel
prompting strategy, least-to-most prompting. The key idea in this strategy is
to break down a complex problem into a series of simpler subproblems and then
solve them in sequence. Solving each subproblem is facilitated by the answers
to previously solved subproblems. Our experimental results on tasks related to
symbolic manipulation, compositional generalization, and math reasoning reveal
that least-to-most prompting is capable of generalizing to more difficult
problems than those seen in the prompts. A notable finding is that when the
GPT-3 code-davinci-002 model is used with least-to-most prompting, it can solve
the compositional generalization benchmark SCAN in any split (including length
split) with an accuracy of at least 99% using just 14 exemplars, compared to
only 16% accuracy with chain-of-thought prompting. This is particularly
noteworthy because neural-symbolic models in the literature that specialize in
solving SCAN are trained on the entire training set containing over 15,000
examples. We have included prompts for all the tasks in the Appendix.
| [
{
"version": "v1",
"created": "Sat, 21 May 2022 15:34:53 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Oct 2022 14:49:26 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Apr 2023 22:08:08 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Zhou",
"Denny",
""
],
[
"Schärli",
"Nathanael",
""
],
[
"Hou",
"Le",
""
],
[
"Wei",
"Jason",
""
],
[
"Scales",
"Nathan",
""
],
[
"Wang",
"Xuezhi",
""
],
[
"Schuurmans",
"Dale",
""
],
[
"Cui",
"Claire",
""
],
[
"Bousquet",
"Olivier",
""
],
[
"Le",
"Quoc",
""
],
[
"Chi",
"Ed",
""
]
] | TITLE: Least-to-Most Prompting Enables Complex Reasoning in Large Language
Models
ABSTRACT: Chain-of-thought prompting has demonstrated remarkable performance on various
natural language reasoning tasks. However, it tends to perform poorly on tasks
which requires solving problems harder than the exemplars shown in the prompts.
To overcome this challenge of easy-to-hard generalization, we propose a novel
prompting strategy, least-to-most prompting. The key idea in this strategy is
to break down a complex problem into a series of simpler subproblems and then
solve them in sequence. Solving each subproblem is facilitated by the answers
to previously solved subproblems. Our experimental results on tasks related to
symbolic manipulation, compositional generalization, and math reasoning reveal
that least-to-most prompting is capable of generalizing to more difficult
problems than those seen in the prompts. A notable finding is that when the
GPT-3 code-davinci-002 model is used with least-to-most prompting, it can solve
the compositional generalization benchmark SCAN in any split (including length
split) with an accuracy of at least 99% using just 14 exemplars, compared to
only 16% accuracy with chain-of-thought prompting. This is particularly
noteworthy because neural-symbolic models in the literature that specialize in
solving SCAN are trained on the entire training set containing over 15,000
examples. We have included prompts for all the tasks in the Appendix.
|
2205.15360 | Stavros Nousias PhD | Nikos D. Fakotakis, Stavros Nousias, Gerasimos Arvanitis, Evangelia I.
Zacharaki, Konstantinos Moustakas | AI-enabled Sound Pattern Recognition on Asthma Medication Adherence:
Evaluation with the RDA Benchmark Suite | null | null | 10.1109/ACCESS.2023.3243547 | null | cs.SD cs.CV cs.CY cs.GL eess.AS | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Asthma is a common, usually long-term respiratory disease with negative
impact on global society and economy. Treatment involves using medical devices
(inhalers) that distribute medication to the airways and its efficiency depends
on the precision of the inhalation technique. There is a clinical need for
objective methods to assess the inhalation technique, during clinical
consultation. Integrated health monitoring systems, equipped with sensors,
enable the recognition of drug actuation, embedded with sound signal detection,
analysis and identification, from intelligent structures, that could provide
powerful tools for reliable content management. Health monitoring systems
equipped with sensors, embedded with sound signal detection, enable the
recognition of drug actuation and could be used for effective audio content
analysis. This paper revisits sound pattern recognition with machine learning
techniques for asthma medication adherence assessment and presents the
Respiratory and Drug Actuation (RDA) Suite
(https://gitlab.com/vvr/monitoring-medication-adherence/rda-benchmark) for
benchmarking and further research. The RDA Suite includes a set of tools for
audio processing, feature extraction and classification procedures and is
provided along with a dataset, consisting of respiratory and drug actuation
sounds. The classification models in RDA are implemented based on conventional
and advanced machine learning and deep networks' architectures. This study
provides a comparative evaluation of the implemented approaches, examines
potential improvements and discusses on challenges and future tendencies.
| [
{
"version": "v1",
"created": "Mon, 30 May 2022 18:08:28 GMT"
},
{
"version": "v2",
"created": "Wed, 1 Jun 2022 11:46:47 GMT"
},
{
"version": "v3",
"created": "Sun, 16 Apr 2023 17:32:06 GMT"
}
] | 2023-04-18T00:00:00 | [
[
"Fakotakis",
"Nikos D.",
""
],
[
"Nousias",
"Stavros",
""
],
[
"Arvanitis",
"Gerasimos",
""
],
[
"Zacharaki",
"Evangelia I.",
""
],
[
"Moustakas",
"Konstantinos",
""
]
] | TITLE: AI-enabled Sound Pattern Recognition on Asthma Medication Adherence:
Evaluation with the RDA Benchmark Suite
ABSTRACT: Asthma is a common, usually long-term respiratory disease with negative
impact on global society and economy. Treatment involves using medical devices
(inhalers) that distribute medication to the airways and its efficiency depends
on the precision of the inhalation technique. There is a clinical need for
objective methods to assess the inhalation technique, during clinical
consultation. Integrated health monitoring systems, equipped with sensors,
enable the recognition of drug actuation, embedded with sound signal detection,
analysis and identification, from intelligent structures, that could provide
powerful tools for reliable content management. Health monitoring systems
equipped with sensors, embedded with sound signal detection, enable the
recognition of drug actuation and could be used for effective audio content
analysis. This paper revisits sound pattern recognition with machine learning
techniques for asthma medication adherence assessment and presents the
Respiratory and Drug Actuation (RDA) Suite
(https://gitlab.com/vvr/monitoring-medication-adherence/rda-benchmark) for
benchmarking and further research. The RDA Suite includes a set of tools for
audio processing, feature extraction and classification procedures and is
provided along with a dataset, consisting of respiratory and drug actuation
sounds. The classification models in RDA are implemented based on conventional
and advanced machine learning and deep networks' architectures. This study
provides a comparative evaluation of the implemented approaches, examines
potential improvements and discusses on challenges and future tendencies.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.