Disentangled Representation Learning with Information Maximizing Autoencoder

TL;DR

This paper introduces InfoAE, an unsupervised autoencoder that learns disentangled representations by maximizing mutual information, demonstrated on MNIST with high accuracy.

Contribution

The paper presents a novel unsupervised autoencoder that effectively learns disentangled representations by maximizing mutual information.

Findings

Achieved 98.9% test accuracy on MNIST

Unsupervised training without labels

Effective disentangled representation learning

Abstract

Learning disentangled representation from any unlabelled data is a non-trivial problem. In this paper we propose Information Maximising Autoencoder (InfoAE) where the encoder learns powerful disentangled representation through maximizing the mutual information between the representation and given information in an unsupervised fashion. We have evaluated our model on MNIST dataset and achieved 98.9 ($\pm .1$) $\%$ test accuracy while using complete unsupervised training.