中文版 | English
Title

Dynamic contrastive learning guided by class confidence and confusion degree for medical image segmentation

Author
Corresponding AuthorZhang,Jianguo
Publication Years
2024
DOI
Source Title
ISSN
0031-3203
EISSN
1873-5142
Volume145
Abstract
This work proposes an intra-Class-confidence and inter-Class-confusion guided Dynamic Contrastive (CCDC) learning framework for medical image segmentation. A core contribution is to dynamically select the most expressive pixels to build positive and negative pairs for contrastive learning at different training phases. For the positive pairs, dynamically adaptive sampling strategies are introduced for sampling different sets of pixels based on their hardness (namely the easiest, easy, and hard pixels). For the negative pairs, to efficiently learn from the classes with high confusion degree w.r.t a query class (i.e., a class containing the query pixels), a new hard class mining strategy is presented. Furthermore, pixel-level representations are extended to the neighbourhood region to leverage the spatial consistency of adjacent pixels. Extensive experiments on the three public datasets demonstrate that the proposed method significantly surpasses the state-of-the-art.
Keywords
URL[Source Record]
Indexed By
Language
English
SUSTech Authorship
First ; Corresponding
Funding Project
National Key Research and Development Program of China[2021YFF1200800];National Natural Science Foundation of China[62276121];
WOS Research Area
Computer Science ; Engineering
WOS Subject
Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS Accession No
WOS:001069734400001
Publisher
ESI Research Field
ENGINEERING
Scopus EID
2-s2.0-85169068506
Data Source
Scopus
Citation statistics
Cited Times [WOS]:0
Document TypeJournal Article
Identifierhttp://kc.sustech.edu.cn/handle/2SGJ60CL/559383
DepartmentDepartment of Computer Science and Engineering
工学院_斯发基斯可信自主研究院
Affiliation
1.Department of computer science and engineering,Southern University of Science and Technology,Shenzhen,518055,China
2.WMG Visualization,University of Warwick,Coventry,CV4 7AL,United Kingdom
3.Department of Computer Science,University of Sheffield,Sheffield,S10 2TN,United Kingdom
4.Research Institute of Trustworthy Autonomous Systems,Southern University of Science and Technology,Shenzhen,518055,China
5.Peng Cheng Lab,Shenzhen,518000,China
First Author AffilicationDepartment of Computer Science and Engineering
Corresponding Author AffilicationDepartment of Computer Science and Engineering;  Research Institute of Trustworthy Autonomous Systems
First Author's First AffilicationDepartment of Computer Science and Engineering
Recommended Citation
GB/T 7714
Chen,Jingkun,Chen,Changrui,Huang,Wenjian,et al. Dynamic contrastive learning guided by class confidence and confusion degree for medical image segmentation[J]. Pattern Recognition,2024,145.
APA
Chen,Jingkun,Chen,Changrui,Huang,Wenjian,Zhang,Jianguo,Debattista,Kurt,&Han,Jungong.(2024).Dynamic contrastive learning guided by class confidence and confusion degree for medical image segmentation.Pattern Recognition,145.
MLA
Chen,Jingkun,et al."Dynamic contrastive learning guided by class confidence and confusion degree for medical image segmentation".Pattern Recognition 145(2024).
Files in This Item:
There are no files associated with this item.
Related Services
Fulltext link
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Export to Excel
Export to Csv
Altmetrics Score
Google Scholar
Similar articles in Google Scholar
[Chen,Jingkun]'s Articles
[Chen,Changrui]'s Articles
[Huang,Wenjian]'s Articles
Baidu Scholar
Similar articles in Baidu Scholar
[Chen,Jingkun]'s Articles
[Chen,Changrui]'s Articles
[Huang,Wenjian]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Chen,Jingkun]'s Articles
[Chen,Changrui]'s Articles
[Huang,Wenjian]'s Articles
Terms of Use
No data!
Social Bookmark/Share
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.