Infinite Variational Autoencoder for Semi-Supervised Learning

TL;DR

This paper introduces an infinite variational autoencoder that dynamically adjusts its capacity using a Dirichlet process mixture model, enhancing semi-supervised learning with limited labeled data.

Contribution

It proposes a novel infinite VAE model that automatically determines the number of autoencoders based on data complexity, improving semi-supervised learning performance.

Findings

Demonstrates flexibility in adapting model capacity to data.

Effective semi-supervised learning with limited labeled samples.

Automatically varies the number of autoencoders based on input data.

Abstract

This paper presents an infinite variational autoencoder (VAE) whose capacity adapts to suit the input data. This is achieved using a mixture model where the mixing coefficients are modeled by a Dirichlet process, allowing us to integrate over the coefficients when performing inference. Critically, this then allows us to automatically vary the number of autoencoders in the mixture based on the data. Experiments show the flexibility of our method, particularly for semi-supervised learning, where only a small number of training samples are available.