English

Covering up bias in CelebA-like datasets with Markov blankets: A post-hoc cure for attribute prior avoidance

Computer Vision and Pattern Recognition 2019-07-31 v1 Machine Learning

Abstract

Attribute prior avoidance entails subconscious or willful non-modeling of (meta)attributes that datasets are oft born with, such as the 40 semantic facial attributes associated with the CelebA and CelebA-HQ datasets. The consequences of this infirmity, we discover, are especially stark in state-of-the-art deep generative models learned on these datasets that just model the pixel-space measurements, resulting in an inter-attribute bias-laden latent space. This viscerally manifests itself when we perform face manipulation experiments based on latent vector interpolations. In this paper, we address this and propose a post-hoc solution that utilizes an Ising attribute prior learned in the attribute space and showcase its efficacy via qualitative experiments.

Keywords

Cite

@article{arxiv.1907.12917,
  title  = {Covering up bias in CelebA-like datasets with Markov blankets: A post-hoc cure for attribute prior avoidance},
  author = {Vinay Uday Prabhu and Dian Ang Yap and Alexander Wang and John Whaley},
  journal= {arXiv preprint arXiv:1907.12917},
  year   = {2019}
}

Comments

Accepted for presentation at the first workshop on Invertible Neural Networks and Normalizing Flows (ICML 2019), Long Beach, CA, USA