Self-Supervised Deep Subspace Clustering with Entropy-norm

TL;DR

This paper introduces S$^{3}$CE, a novel self-supervised deep subspace clustering method that enhances feature learning and connectivity, addressing computational and regularization limitations of existing models, and demonstrates superior performance.

Contribution

The paper proposes a new self-supervised contrastive network with entropy-norm regularization and data enhancement modules for improved deep subspace clustering.

Findings

Outperforms state-of-the-art methods in clustering accuracy.

Effectively captures local structure and dense connectivity.

Reduces computational cost compared to traditional self-expression models.

Abstract

Auto-Encoder based deep subspace clustering (DSC) is widely used in computer vision, motion segmentation and image processing. However, it suffers from the following three issues in the self-expressive matrix learning process: the first one is less useful information for learning self-expressive weights due to the simple reconstruction loss; the second one is that the construction of the self-expression layer associated with the sample size requires high-computational cost; and the last one is the limited connectivity of the existing regularization terms. In order to address these issues, in this paper we propose a novel model named Self-Supervised deep Subspace Clustering with Entropy-norm (S$^{3}$CE). Specifically, S$^{3}$CE exploits a self-supervised contrastive network to gain a more effetive feature vector. The local structure and dense connectivity of the original data benefit from the self-expressive layer and additional entropy-norm constraint. Moreover, a new module with data enhancement is designed to help S$^{3}$CE focus on the key information of data, and improve the clustering performance of positive and negative instances through spectral clustering. Extensive experimental results demonstrate the superior performance of S$^{3}$CE in comparison to the state-of-the-art approaches.