A Revisit of Total Correlation in Disentangled Variational Auto-Encoder with Partial Disentanglement

TL;DR

This paper introduces PDisVAE, a flexible variational auto-encoder that relaxes the strict independence constraints of traditional models, enabling partial disentanglement to better capture complex data factors.

Contribution

It proposes the partial correlation (PC) term in VAEs, allowing for group-wise independence and more adaptable disentanglement compared to full independence models.

Findings

PDisVAE effectively handles group-wise independence.

It outperforms fully disentangled VAEs on real-world data.

The framework is validated through synthetic experiments.

Abstract

A fully disentangled variational auto-encoder (VAE) aims to identify disentangled latent components from observations. However, enforcing full independence between all latent components may be too strict for certain datasets. In some cases, multiple factors may be entangled together in a non-separable manner, or a single independent semantic meaning could be represented by multiple latent components within a higher-dimensional manifold. To address such scenarios with greater flexibility, we develop the Partially Disentangled VAE (PDisVAE), which generalizes the total correlation (TC) term in fully disentangled VAEs to a partial correlation (PC) term. This framework can handle group-wise independence and can naturally reduce to either the standard VAE or the fully disentangled VAE. Validation through three synthetic experiments demonstrates the correctness and practicality of PDisVAE. When applied to real-world datasets, PDisVAE discovers valuable information that is difficult to find using fully disentangled VAEs, implying its versatility and effectiveness.