Deep Generative Clustering with VAEs and Expectation-Maximization

TL;DR

This paper introduces a deep clustering method combining Variational Autoencoders with the Expectation-Maximization algorithm, improving clustering accuracy without relying on GMM priors, and enabling sample generation from clusters.

Contribution

It presents a novel integration of VAEs into the EM framework for clustering, removing the need for GMM priors and enhancing clustering and generation capabilities.

Findings

Superior clustering performance on MNIST and FashionMNIST

Eliminates need for GMM prior or regularization

Effective sample generation from learned clusters

Abstract

We propose a novel deep clustering method that integrates Variational Autoencoders (VAEs) into the Expectation-Maximization (EM) framework. Our approach models the probability distribution of each cluster with a VAE and alternates between updating model parameters by maximizing the Evidence Lower Bound (ELBO) of the log-likelihood and refining cluster assignments based on the learned distributions. This enables effective clustering and generation of new samples from each cluster. Unlike existing VAE-based methods, our approach eliminates the need for a Gaussian Mixture Model (GMM) prior or additional regularization techniques. Experiments on MNIST and FashionMNIST demonstrate superior clustering performance compared to state-of-the-art methods.