GCVAE: Generalized-Controllable Variational AutoEncoder - Mines Saint-Étienne
Preprints, Working Papers, ... Year : 2022

GCVAE: Generalized-Controllable Variational AutoEncoder

Abstract

Variational autoencoders (VAEs) have recently been used for unsupervised disentanglement learning of complex density distributions. Numerous variants exist to encourage disentanglement in latent space while improving reconstruction. However, none have simultaneously managed the trade-off between attaining extremely low reconstruction error and a high disentanglement score. We present a generalized framework to handle this challenge under constrained optimization and demonstrate that it outperforms state-of-the-art existing models as regards disentanglement while balancing reconstruction. We introduce three controllable Lagrangian hyperparameters to control reconstruction loss, KL divergence loss and correlation measure. We prove that maximizing information in the reconstruction network is equivalent to information maximization during amortized inference under reasonable assumptions and constraint relaxation.
Fichier principal
Vignette du fichier
2206.04225.pdf (1.1 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

emse-03712594 , version 1 (04-07-2022)

Identifiers

  • HAL Id : emse-03712594 , version 1

Cite

Kenneth Ezukwoke, Anis Hoayek, Mireille Batton-Hubert, Xavier Boucher. GCVAE: Generalized-Controllable Variational AutoEncoder. 2022. ⟨emse-03712594⟩
213 View
83 Download

Share

More