Contrastive Explanations in Neural Networks

2020/08/01 by Prabhushankar, Mohit, Kwon, Gukyeong, Temel, Dogancan +1 · 2 citations
#Artificial Intelligence (cs.AI) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2008.00178

Abstract

Visual explanations are logical arguments based on visual features that justify the predictions made by neural networks. Current modes of visual explanations answer questions of the form `Why P?'. These Why questions operate under broad contexts thereby providing answers that are irrelevant in some cases. We propose to constrain these Why questions based on some context Q so that our explanations answer contrastive questions of the form `Why P, rather than Q?'. In this paper, we formalize the structure of contrastive visual explanations for neural networks. We define contrast based on neural networks and propose a methodology to extract defined contrasts. We then use the extracted contrasts as a plug-in on top of existing `Why P?' techniques, specifically Grad-CAM. We demonstrate their value in analyzing both networks and data in applications of large-scale recognition, fine-grained recognition, subsurface seismic analysis, and image quality assessment.

Cited by

• Towards Fine-Grained Interpretability: Counterfactual Explanations for Misclassification with Saliency Partition
• Toward Simple and Robust Contrastive Explanations for Image Classification by Leveraging Instance Similarity and Concept Relevance

Related

• Measuring and improving the quality of visual explanations
• Contrastive Reasoning in Neural Networks
• Explanatory Paradigms in Neural Networks
• Towards Explainable Neural-Symbolic Visual Reasoning
• Explaining Deep Neural Networks