Consensus One-Step Multi-view Image Clustering Based on Low-Rank Tensor learning

Abstract

Multi-view subspace clustering aims to divide a set of multi-source data into several groups according to their underlying subspace structure. Despite superior clustering performance in various applications, most existing methods directly construct noisy affinity matrices by self-representation, and isolate the processes of affinity learning, multi-view information and clustering. Both factors may cause insufficient utilization of multi-view information, leading to unsatisfying clustering performance. In this paper, we propose a novel consensus one-step multi-view clustering method based on lowrank tensor learning to address these issues. Low-rank tensor learning, consensus learning and labels learning in a unified framework. Through the three steps of mutual negotiation, the final clustering label is directly obtained. Experimental results on four benchmark datasets demonstrate that our method outperforms other state-of-the-art methods.