ClusterGAN : Latent Space Clustering in Generative Adversarial Networks

TL;DR

This paper introduces ClusterGAN, a novel GAN-based clustering method that effectively preserves cluster structures in the latent space, enabling superior unsupervised clustering on synthetic and real datasets.

Contribution

ClusterGAN employs a mixture of one-hot and continuous latent variables with a joint training scheme to achieve effective clustering in GAN latent space.

Findings

GANs can preserve latent space interpolation across categories

ClusterGAN outperforms existing clustering baselines

Latent space clustering is feasible with the proposed method

Abstract

Generative Adversarial networks (GANs) have obtained remarkable success in many unsupervised learning tasks and unarguably, clustering is an important unsupervised learning problem. While one can potentially exploit the latent-space back-projection in GANs to cluster, we demonstrate that the cluster structure is not retained in the GAN latent space. In this paper, we propose ClusterGAN as a new mechanism for clustering using GANs. By sampling latent variables from a mixture of one-hot encoded variables and continuous latent variables, coupled with an inverse network (which projects the data to the latent space) trained jointly with a clustering specific loss, we are able to achieve clustering in the latent space. Our results show a remarkable phenomenon that GANs can preserve latent space interpolation across categories, even though the discriminator is never exposed to such vectors. We compare our results with various clustering baselines and demonstrate superior performance on both synthetic and real datasets.