Abstract:Multi-view subspace clustering has attracted extensive attention due to its ability to efficiently handle data from diverse sources. In recent years, plentiful multi-view subspace clustering methods have emerged and achieved satisfactory clustering performance. However, these methods rarely consider simultaneously handling data with a nonlinear structure and exploiting the structural and multi-level information inherent in the data. To remedy these shortcomings, we propose the novel multi-view deep subspace clustering via level-by-level guided multi-level features learning (MDSC-LGMFL). Specifically, an autoencoder is used for each view to extract the view-specific multi-level features, and multiple self-representation layers are introduced into the autoencoder to learn the subspace representations corresponding to the multi-level features. These self-representation layers not only provide multiple information flow paths through the autoencoder but also enforce multiple encoder layers to produce the multi-level features that satisfy the linear subspace assumption. With the novel level-by-level guidance strategy, the last-level feature is guaranteed to encode the structural information from the view and the previous-level features. Naturally, the subspace representation of the last-level feature can more reliably reflect the data affinity relationship and thus can be viewed as the new, better representation of the view. Furthermore, to guarantee the structural consistency among different views, instead of simply learning the common subspace structure by enforcing it to be close to different view-specific new, better representations, we conduct self-representation on these new, better representations to learn the common subspace structure, which can be applied to the spectral clustering algorithm to achieve the final clustering results. Numerous experiments on six widely used benchmark datasets show the superiority of the proposed method.

Structure-Constrained Feature Extraction by Autoencoders for Subspace Clustering

Subspace Clustering Using a Low-Rank Constrained Autoencoder

Autoencoder Constrained Clustering with Adaptive Neighbors.

Nonlinear Subspace Clustering Via Adaptive Graph Regularized Autoencoder.

Deep subspace clustering to achieve jointly latent feature extraction and discriminative learning

Multi-view deep subspace clustering via level-by-level guided multi-level features learning

Graph Regularized Sparse Autoencoders with Nonnegativity Constraints

Selecting the Best Part from Multiple Laplacian Autoencoders for Multi-view Subspace Clustering

From Ensemble Clustering to Subspace Clustering: Cluster Structure Encoding

Exploiting Unsupervised and Supervised Constraints for Subspace Clustering

Iterative Deep Subspace Clustering.

Laplacian regularized deep low-rank subspace clustering network

Deep Double Self-Expressive Subspace Clustering

LRSC: Learning Representations for Subspace Clustering

Semisupervised Feature Learning by Deep Entropy-Sparsity Subspace Clustering.

Robust Subspace Clustering With Low-Rank Structure Constraint

Exploring structural components in autoencoder-based data clustering

Autoencoders with Intrinsic Dimension Constraints for Learning Low Dimensional Image Representations

Deep Clustering with Convolutional Autoencoders

Unsupervised embedded feature learning for deep clustering with stacked sparse auto-encoder

Multi-Level Representation Learning for Deep Subspace Clustering