Enhanced Tensorial Self-representation Subspace Learning for Incomplete Multi-view Clustering

hangjun Che, Xinyu Pu, Deqiang Ouyang, Beibei Li

Abstract: Incomplete Multi-View Clustering (IMVC) is a promising topic in multimedia as it breaks the data completeness assumption. Most existing methods solve IMVC from the perspective of graph learning. In contrast, self-representation learning enjoys a superior ability to explore relationships among samples. However, only a few works have explored the potentiality of self-representation learning in IMVC. These self-representation methods infer missing entries from the perspective of whole samples, resulting in redundant information. In addition, designing an effective strategy to retain salient features while eliminating noise is rarely considered in IMVC. To tackle these issues, we propose a novel self-representation learning method with missing sample recovery and enhanced low-rank tensor regularization. Specifically, the missing samples are inferred by leveraging the local structure of each view, which is constructed from available samples at the feature level. Then an enhanced tensor norm, referred to as Logarithm-p norm is devised, which can obtain an accurate cross-view description. Our proposed method achieves exact subspace representation in IMVC by leveraging high-order correlations and inferring missing information at the feature level. Extensive experiments on several widely used multi-view datasets demonstrate the effectiveness of the proposed method.

Relevance To Conference: The goal of this paper is incomplete multi-view clustering, which solves the clustering task for incomplete multi-view data. The incomplete multi-view data, which can consist of multiple views, e.g. textual, visual, and audio signals, brings great challenges to clustering tasks due to the presence of samples. We provide a new solution for this task from another perspective by capturing relationships between samples in the subspace. Inspired by self-representation subspace learning and high-order correlation learning, an efficient incomplete multi-view clustering method is devised to explore beneficial multi-view messages and lead to more exact subspace representation. We believe our approach will not only advance the current understanding but also open new avenues of exploration for the broader multimedia community.

Primary Subject Area: [Content] Multimodal Fusion

Submission Number: 4797