ADVISE: ADaptive Feature Relevance and VISual Explanations for Convolutional Neural Networks
This addresses the need for explainability in CNNs to understand model decisions, errors, and biases, which is crucial for improving architecture design and ensuring ethical AI, though it is an incremental improvement on existing visual explanation methods.
The paper tackles the problem of interpreting decisions in Convolutional Neural Networks (CNNs) by introducing ADVISE, a method that quantifies feature relevance using adaptive bandwidth kernel density estimation to provide visual explanations, and it outperforms state-of-the-art methods in feature-relevance and visual explainability while maintaining competitive time complexity.
To equip Convolutional Neural Networks (CNNs) with explainability, it is essential to interpret how opaque models take specific decisions, understand what causes the errors, improve the architecture design, and identify unethical biases in the classifiers. This paper introduces ADVISE, a new explainability method that quantifies and leverages the relevance of each unit of the feature map to provide better visual explanations. To this end, we propose using adaptive bandwidth kernel density estimation to assign a relevance score to each unit of the feature map with respect to the predicted class. We also propose an evaluation protocol to quantitatively assess the visual explainability of CNN models. We extensively evaluate our idea in the image classification task using AlexNet, VGG16, ResNet50, and Xception pretrained on ImageNet. We compare ADVISE with the state-of-the-art visual explainable methods and show that the proposed method outperforms competing approaches in quantifying feature-relevance and visual explainability while maintaining competitive time complexity. Our experiments further show that ADVISE fulfils the sensitivity and implementation independence axioms while passing the sanity checks. The implementation is accessible for reproducibility purposes on https://github.com/dehshibi/ADVISE.