Towards Counterfactual and Contrastive Explainability and Transparency of DCNN Image Classifiers

TL;DR

This paper introduces a model-intrusive method for generating contrastive and counterfactual explanations of DCNN image classifiers, enhancing interpretability and transparency by analyzing internal filters and concepts without altering input images.

Contribution

The paper presents a novel approach that probes internal DCNN filters for explanations, offering contrastive and counterfactual insights without modifying input images, improving transparency.

Findings

Effective in identifying key filters and concepts influencing decisions

Provides meaningful contrastive and counterfactual explanations

Evaluated successfully on CUB 2011 dataset

Abstract

Explainability of deep convolutional neural networks (DCNNs) is an important research topic that tries to uncover the reasons behind a DCNN model's decisions and improve their understanding and reliability in high-risk environments. In this regard, we propose a novel method for generating interpretable counterfactual and contrastive explanations for DCNN models. The proposed method is model intrusive that probes the internal workings of a DCNN instead of altering the input image to generate explanations. Given an input image, we provide contrastive explanations by identifying the most important filters in the DCNN representing features and concepts that separate the model's decision between classifying the image to the original inferred class or some other specified alter class. On the other hand, we provide counterfactual explanations by specifying the minimal changes necessary in such filters so that a contrastive output is obtained. Using these identified filters and concepts, our method can provide contrastive and counterfactual reasons behind a model's decisions and makes the model more transparent. One of the interesting applications of this method is misclassification analysis, where we compare the identified concepts from a particular input image and compare them with class-specific concepts to establish the validity of the model's decisions. The proposed method is compared with state-of-the-art and evaluated on the Caltech-UCSD Birds (CUB) 2011 dataset to show the usefulness of the explanations provided.