Deep Clustering by Graph Attention Contrastive Learning
Electronics · 31.05.2023 · 12 (11) · 2489
DOI: 10.3390/electronics12112489 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 31.05.2023. Verfasst von Ming Liu und 6 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Beijing University of Posts and Telecommunications-China Mobile Research Institute Joint Innovation Center.
Abstract (Originalsprache)
Contrastive learning shows great potential in deep clustering. It uses constructed pairs to discover the feature distribution that is required for the clustering task. In addition to conventional augmented pairs, recent methods have introduced more methods of creating highly confident pairs, such as nearest neighbors, to provide more semantic prior knowledge. However, existing works only use partial pairwise similarities to construct semantic pairs locally without capturing the entire sample’s relationships from a global perspective. In this paper, we propose a novel clustering framework called graph attention contrastive learning (GACL) to aggregate more semantic information. To this end, GACL is designed to simultaneously perform instance-level and graph-level contrast. Specifically, with its novel graph attention mechanism, our model explores more undiscovered pairs and selectively focuses on informative pairs. To ensure local and global clustering consistency, we jointly use the designed graph-level and instance-level contrastive losses. Experiments on six challenging image benchmarks demonstrate the superiority of our proposed approach over state-of-the-art methods.
Förderung
- National Natural Science Foundation of China
- Beijing University of Posts and Telecommunications-China Mobile Research Institute Joint Innovation Center
Zitieren
APA
Liu, M., Liu, C., Fu, X., Wang, J., Li, J., Qi, Q., & Liao, J. (2023). Deep Clustering by Graph Attention Contrastive Learning. Electronics, 12(11), 2489. https://doi.org/10.3390/electronics12112489
Vancouver
Liu M, Liu C, Fu X, Wang J, Li J, Qi Q, et al. Deep Clustering by Graph Attention Contrastive Learning. Electronics. 2023;12(11):2489. doi:10.3390/electronics12112489
BibTeX
@article{Liu2023,
author = {Liu, Ming and Liu, Cong and Fu, Xiaoyuan and Wang, Jing and Li, Jiankun and Qi, Qi and Liao, Jianxin},
title = {Deep Clustering by Graph Attention Contrastive Learning},
journal = {Electronics},
year = {2023},
volume = {12},
number = {11},
pages = {2489},
doi = {10.3390/electronics12112489}
}
EUVIMED