Invariant Representations from Adversarially Censored Autoencoders

2018/05/21 by Ye Wang, Wang, Ye, Toshiaki Koike–Akino +3 · 2 citations
Computer Science · Physics and Astronomy · #62B10 #68T05 #Digital Media Forensic Detection #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Information Theory (cs.IT) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Model Reduction and Neural Networks

paper · pdf · doi:10.48550/arxiv.1805.08097

openalex publication_date 2018/05/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We combine conditional variational autoencoders (VAE) with adversarial censoring in order to learn invariant representations that are disentangled from nuisance/sensitive variations. In this method, an adversarial network attempts to recover the nuisance variable from the representation, which the VAE is trained to prevent. Conditioning the decoder on the nuisance variable enables clean separation of the representation, since they are recombined for model learning and data reconstruction. We show this natural approach is theoretically well-founded with information-theoretic arguments. Experiments demonstrate that this method achieves invariance while preserving model learning performance, and results in visually improved performance for style transfer and generative sampling tasks.

Citations

• Auto-Encoding Variational Bayes
• Adversarial Feature Learning
• Controllable Invariance through Adversarial Feature Learning
• Adversarial Discriminative Domain Adaptation
• Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks
• Adversarial Autoencoders

Cited by

• Transfer Learning in Brain-Computer Interfaces with Adversarial\n Variational Autoencoders
• Representation via Representations: Domain Generalization via Adversarially Learned Invariant Representations

Related

• The Variational Fair Autoencoder
• A Cyclically-Trained Adversarial Network for Invariant Representation Learning
• Invariant Representations without Adversarial Training
• Autoencoding Variational Autoencoder
• Adversarial Images for Variational Autoencoders