Robust Self-Supervised Convolutional Neural Network for Subspace Clustering and Classification

TL;DR

This paper introduces a robust self-supervised convolutional neural network for subspace clustering and classification that effectively handles nonlinear data, corruptions, and out-of-sample data, outperforming previous methods.

Contribution

It proposes the $S^2$ConvSCN model with a fully connected layer and CIM-based robustness, explicitly enforces block-diagonal structure, and demonstrates superior performance in unsupervised settings.

Findings

Outperforms baseline on multiple datasets

Handles nonlinear manifolds and corruptions effectively

Enforces block-diagonal structure for better clustering

Abstract

Insufficient capability of existing subspace clustering methods to handle data coming from nonlinear manifolds, data corruptions, and out-of-sample data hinders their applicability to address real-world clustering and classification problems. This paper proposes the robust formulation of the self-supervised convolutional subspace clustering network ($S^2$ConvSCN) that incorporates the fully connected (FC) layer and, thus, it is capable for handling out-of-sample data by classifying them using a softmax classifier. $S^2$ConvSCN clusters data coming from nonlinear manifolds by learning the linear self-representation model in the feature space. Robustness to data corruptions is achieved by using the correntropy induced metric (CIM) of the error. Furthermore, the block-diagonal (BD) structure of the representation matrix is enforced explicitly through BD regularization. In a truly unsupervised training environment, Robust $S^2$ConvSCN outperforms its baseline version by a significant amount for both seen and unseen data on four well-known datasets. Arguably, such an ablation study has not been reported before.