Latent Class-Conditional Noise Model.

Yao, J; Han, B; Zhou, Z; Zhang, Y; Tsang, IW

Latent Class-Conditional Noise Model.

Yao, J Han, B Zhou, Z Zhang, Y Tsang, IW

Permalink

Publisher:: IEEE COMPUTER SOC
Publication Type:: Journal Article
Citation:: IEEE Trans Pattern Anal Mach Intell, 2023, 45, (8), pp. 9964-9980
Issue Date:: 2023-08

Closed Access

	Filename	Description	Size
	Latent_Class-Conditional_Noise_Model.pdf	Published version	2.4 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Yao, J
dc.contributor.author	Han, B
dc.contributor.author	Zhou, Z
dc.contributor.author	Zhang, Y
dc.contributor.author	Tsang, IW
dc.date.accessioned	2024-05-04T02:12:35Z
dc.date.available	2024-05-04T02:12:35Z
dc.date.issued	2023-08
dc.identifier.citation	IEEE Trans Pattern Anal Mach Intell, 2023, 45, (8), pp. 9964-9980
dc.identifier.issn	0162-8828
dc.identifier.issn	1939-3539
dc.identifier.uri	http://hdl.handle.net/10453/178641
dc.description.abstract	Learning with noisy labels has become imperative in the Big Data era, which saves expensive human labors on accurate annotations. Previous noise-transition-based methods have achieved theoretically-grounded performance under the Class-Conditional Noise model (CCN). However, these approaches builds upon an ideal but impractical anchor set available to pre-estimate the noise transition. Even though subsequent works adapt the estimation as a neural layer, the ill-posed stochastic learning of its parameters in back-propagation easily falls into undesired local minimums. We solve this problem by introducing a Latent Class-Conditional Noise model (LCCN) to parameterize the noise transition under a Bayesian framework. By projecting the noise transition into the Dirichlet space, the learning is constrained on a simplex characterized by the complete dataset, instead of some ad-hoc parametric space wrapped by the neural layer. We then deduce a dynamic label regression method for LCCN, whose Gibbs sampler allows us efficiently infer the latent true labels to train the classifier and to model the noise. Our approach safeguards the stable update of the noise transition, which avoids previous arbitrarily tuning from a mini-batch of samples. We further generalize LCCN to different counterparts compatible with open-set noisy labels, semi-supervised learning as well as cross-model training. A range of experiments demonstrate the advantages of LCCN and its variants over the current state-of-the-art methods. The code is available at here.
dc.format	Print-Electronic
dc.language	eng
dc.publisher	IEEE COMPUTER SOC
dc.relation.ispartof	IEEE Trans Pattern Anal Mach Intell
dc.relation.isbasedon	10.1109/TPAMI.2023.3247629
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	0801 Artificial Intelligence and Image Processing, 0806 Information Systems, 0906 Electrical and Electronic Engineering
dc.subject.classification	Artificial Intelligence & Image Processing
dc.subject.classification	4603 Computer vision and multimedia computation
dc.subject.classification	4611 Machine learning
dc.subject.mesh	Humans
dc.subject.mesh	Bayes Theorem
dc.subject.mesh	Algorithms
dc.subject.mesh	Big Data
dc.subject.mesh	Supervised Machine Learning
dc.subject.mesh	Humans
dc.subject.mesh	Bayes Theorem
dc.subject.mesh	Algorithms
dc.subject.mesh	Supervised Machine Learning
dc.subject.mesh	Big Data
dc.subject.mesh	Humans
dc.subject.mesh	Bayes Theorem
dc.subject.mesh	Algorithms
dc.subject.mesh	Big Data
dc.subject.mesh	Supervised Machine Learning
dc.title	Latent Class-Conditional Noise Model.
dc.type	Journal Article
utslib.citation.volume	45
utslib.location.activity	United States
utslib.for	0801 Artificial Intelligence and Image Processing
utslib.for	0806 Information Systems
utslib.for	0906 Electrical and Electronic Engineering
pubs.organisational-group	University of Technology Sydney
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	University of Technology Sydney/Strength - AAII - Australian Artificial Intelligence Institute
utslib.copyright.status	closed_access	*
dc.date.updated	2024-05-04T02:12:33Z
pubs.issue	8
pubs.publication-status	Published
pubs.volume	45
utslib.citation.issue	8

Abstract:

Learning with noisy labels has become imperative in the Big Data era, which saves expensive human labors on accurate annotations. Previous noise-transition-based methods have achieved theoretically-grounded performance under the Class-Conditional Noise model (CCN). However, these approaches builds upon an ideal but impractical anchor set available to pre-estimate the noise transition. Even though subsequent works adapt the estimation as a neural layer, the ill-posed stochastic learning of its parameters in back-propagation easily falls into undesired local minimums. We solve this problem by introducing a Latent Class-Conditional Noise model (LCCN) to parameterize the noise transition under a Bayesian framework. By projecting the noise transition into the Dirichlet space, the learning is constrained on a simplex characterized by the complete dataset, instead of some ad-hoc parametric space wrapped by the neural layer. We then deduce a dynamic label regression method for LCCN, whose Gibbs sampler allows us efficiently infer the latent true labels to train the classifier and to model the noise. Our approach safeguards the stable update of the noise transition, which avoids previous arbitrarily tuning from a mini-batch of samples. We further generalize LCCN to different counterparts compatible with open-set noisy labels, semi-supervised learning as well as cross-model training. A range of experiments demonstrate the advantages of LCCN and its variants over the current state-of-the-art methods. The code is available at here.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/178641