Abstract:We consider the problem of simultaneously clustering and learning a linear representation of data lying close to a union of low-dimensional manifolds, a fundamental task in machine learning and computer vision. When the manifolds are assumed to be linear subspaces, this reduces to the classical problem of subspace clustering, which has been studied extensively over the past two decades. Unfortunately, many real-world datasets such as natural images can not be well approximated by linear subspaces. On the other hand, numerous works have attempted to learn an appropriate transformation of the data, such that data is mapped from a union of general non-linear manifolds to a union of linear subspaces (with points from the same manifold being mapped to the same subspace). However, many existing works have limitations such as assuming knowledge of the membership of samples to clusters, requiring high sampling density, or being shown theoretically to learn trivial representations. In this paper, we propose to optimize the Maximal Coding Rate Reduction metric with respect to both the data representation and a novel doubly stochastic cluster membership, inspired by state-of-the-art subspace clustering results. We give a parameterization of such a representation and membership, allowing efficient mini-batching and one-shot initialization. Experiments on CIFAR-10, -20, -100, and TinyImageNet-200 datasets show that the proposed method is much more accurate and scalable than state-of-the-art deep clustering methods, and further learns a latent linear representation of the data.

Unsupervised clustering algorithm on 3D model library via decision graph

Feature Extraction of 3D Model Based on Fuzzy Clustering

Unsupervised 3D Shape Classification Algorithm Using Density Peaks

A Nonparametric Model for Multi-Manifold Clustering with Mixture of Gaussians and Graph Consistency

A Robust Hierarchical Clustering Algorithm and its Application in 3D Model Retrieval

Clustering through decision tree construction

Subspace Clustering Via Integrating Sparse Representation and Adaptive Graph Learning.

Discriminative Unsupervised 2D Dimensionality Reduction with Graph Embedding

A Robust Density-Based Clustering Algorithm for Multi-Manifold Structure.

Self-Supervised Graph Embedding Clustering

Unsupervised Manifold Linearizing and Clustering

Semi-Supervised Clustering via Dynamic Graph Structure Learning

A Three-Way Clustering Method Based on Ensemble Strategy and Three-Way Decision

Clustering Based Model Datasets Visualization and Retrieval

Unsupervised Clustering Using Graph Transduction

Clustering three-dimensional model based on external feature

Clustering Bioactive Molecules in 3D Chemical Space with Unsupervised Deep Learning

Using Bagging to improve clustering methods in the context of three-dimensional shapes

Study on A 3d Clustering Algorithm

Unsupervised Detection of Distinctive Regions on 3D Shapes

Unsupervised Multi-Manifold Clustering by Learning Deep Representation