The Gaussian Discriminant Variational Autoencoder (GdVAE): A Self-Explainable Model with Counterfactual Explanations

TL;DR

The GdVAE model combines a self-explainable variational autoencoder with counterfactual explanations, providing transparent and high-quality hypothetical scenario generation for image classification.

Contribution

It introduces a novel GdVAE framework that integrates a generative classifier with counterfactual explanations, enhancing transparency and explanation quality in image models.

Findings

Produces high-quality counterfactual explanations

Maintains full transparency through generative classification

Outperforms existing methods in explanation effectiveness

Abstract

Visual counterfactual explanation (CF) methods modify image concepts, e.g, shape, to change a prediction to a predefined outcome while closely resembling the original query image. Unlike self-explainable models (SEMs) and heatmap techniques, they grant users the ability to examine hypothetical "what-if" scenarios. Previous CF methods either entail post-hoc training, limiting the balance between transparency and CF quality, or demand optimization during inference. To bridge the gap between transparent SEMs and CF methods, we introduce the GdVAE, a self-explainable model based on a conditional variational autoencoder (CVAE), featuring a Gaussian discriminant analysis (GDA) classifier and integrated CF explanations. Full transparency is achieved through a generative classifier that leverages class-specific prototypes for the downstream task and a closed-form solution for CFs in the latent space. The consistency of CFs is improved by regularizing the latent space with the explainer function. Extensive comparisons with existing approaches affirm the effectiveness of our method in producing high-quality CF explanations while preserving transparency. Code and models are public.