EDDA: Explanation-driven Data Augmentation to Improve Explanation Faithfulness

Abstract

Recent years have seen the introduction of a range of methods for post-hoc explainability of image classifier predictions. However, these post-hoc explanations may not always be faithful to classifier predictions, which poses a significant challenge when attempting to debug models based on such explanations. To this end, we seek a methodology that can improve the faithfulness of an explanation method with respect to model predictions which does not require ground truth explanations. We achieve this through a novel explanation-driven data augmentation (EDDA) technique that augments the training data with occlusions inferred from model explanations; this is based on the simple motivating principle that \emph{if} the explainer is faithful to the model \emph{then} occluding salient regions for the model prediction should decrease the model confidence in the prediction, while occluding non-salient regions should not change the prediction. To verify that the proposed augmentation method has the potential to improve faithfulness, we evaluate EDDA using a variety of datasets and classification models. We demonstrate empirically that our approach leads to a significant increase of faithfulness, which can facilitate better debugging and successful deployment of image classification models in real-world applications.

References in corpus (15)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• mixup: Beyond Empirical Risk Minimization
• Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networks
• Axiomatic Attribution for Deep Networks
• Improved Regularization of Convolutional Neural Networks with Cutout
• Understanding Black-box Predictions via Influence Functions
• Fully Convolutional Networks for Semantic Segmentation
• SmoothGrad: removing noise by adding noise
• Understanding Global Feature Contributions With Additive Importance Measures
• Axiom-based Grad-CAM: Towards Accurate Visualization and Explanation of CNNs
• Estimating Training Data Influence by Tracing Gradient Descent
• Representer Point Selection for Explaining Deep Neural Networks
• On the (In)fidelity and Sensitivity for Explanations
• Explaining Convolutional Neural Networks through Attribution-Based Input Sampling and Block-Wise Feature Aggregation