File(s) not publicly available
A model with pseudo-label correction and distribution alignment for image clustering
journal contribution
posted on 2023-02-14, 23:47 authored by F Zhang, ZW Liu, Chee Peng LimChee Peng Lim, CR Dong, Q HuaSelf-supervised learning has recently created a surge of research interest in image clustering with the help of pseudo-labels. However, most of methods have to face the challenge of dealing with noisy labels since the noisy pseudo-labels might provide erroneous guidance for model training and deteriorate the clustering performance. To solve this problem, a Pseudo-label Correction and Distribution Alignment-based deep Clustering (PCDAC) model is proposed in this paper. To correct the noisy pseudo-labels, PCDAC embeds the historical cluster assignment by maintaining a memory bank. Additionally, a distribution alignment operation is employed to impose the cluster probability distribution of strong-augmented instances close to those of weak-augmented instances, to reduce the possibility of generating incorrect pseudo-labels at the initial training phase. Extensive comparisons experiments are conducted and the results illustrate that PCDAC is superior to the SOTA methods. The code is available on GitHub at https://github.com/LiuZiweiAI/PCDAC.
History
Journal
Computers and Electrical EngineeringVolume
104Article number
ARTN 108457Publisher DOI
ISSN
0045-7906eISSN
1879-0755Language
EnglishPublication classification
C1 Refereed article in a scholarly journalPublisher
PERGAMON-ELSEVIER SCIENCE LTDUsage metrics
Categories
Keywords
Science & TechnologyTechnologyComputer Science, Hardware & ArchitectureComputer Science, Interdisciplinary ApplicationsEngineering, Electrical & ElectronicComputer ScienceEngineeringPseudo -label correctionDistribution alignmentImage clusteringContrastive learningNETWORKElectrical and Electronic Engineering not elsewhere classifiedComputer SoftwareDistributed Computing
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC