Partially multi-view clustering via re-alignment

Wenbiao Yan, Jihua Zhu, Jinqian Chen, Haozhe Cheng, Shunshun Bai, Liang Duan, Qinghai Zheng

Abstract: Highlights•PMVCR integrates contrastive learning with re-alignment to address partially unaligned instances in multi-view clustering.•Our two-stage approach enhances view representation and discriminative power, improving clustering accuracy.•Experiments demonstrate PMVCR’s effectiveness and superiority over existing methods on popular multi-view datasets.