Hierarchical Variational Autoencoder for Visual Counterfactuals
This work addresses the problem of producing effective counterfactuals for explainable AI in complex real-world image applications, though it appears incremental.
The paper tackles the challenge of generating visual counterfactuals with hierarchical conditional variational autoencoders (CVAEs) by relaxing the posterior effect, introducing VAEX, a model that can visually audit classifiers in applications.
Conditional Variational Auto Encoders (VAE) are gathering significant attention as an Explainable Artificial Intelligence (XAI) tool. The codes in the latent space provide a theoretically sound way to produce counterfactuals, i.e. alterations resulting from an intervention on a targeted semantic feature. To be applied on real images more complex models are needed, such as Hierarchical CVAE. This comes with a challenge as the naive conditioning is no longer effective. In this paper we show how relaxing the effect of the posterior leads to successful counterfactuals and we introduce VAEX an Hierarchical VAE designed for this approach that can visually audit a classifier in applications.