Feature CAM: Interpretable AI in Image Classification

TL;DR

This paper introduces Feature CAM, a new interpretability technique for CNNs in image classification that produces more human-understandable visualizations while maintaining machine confidence levels.

Contribution

We propose Feature CAM, a novel perturbation-activation method that enhances interpretability of CNN predictions with finer, class-discriminative visualizations.

Findings

Saliency maps are 3-4 times more human interpretable.

Feature CAM maintains average confidence scores in classification.

The method outperforms existing Activation-based methods in interpretability.

Abstract

Deep Neural Networks have often been called the black box because of the complex, deep architecture and non-transparency presented by the inner layers. There is a lack of trust to use Artificial Intelligence in critical and high-precision fields such as security, finance, health, and manufacturing industries. A lot of focused work has been done to provide interpretable models, intending to deliver meaningful insights into the thoughts and behavior of neural networks. In our research, we compare the state-of-the-art methods in the Activation-based methods (ABM) for interpreting predictions of CNN models, specifically in the application of Image Classification. We then extend the same for eight CNN-based architectures to compare the differences in visualization and thus interpretability. We introduced a novel technique Feature CAM, which falls in the perturbation-activation combination, to create fine-grained, class-discriminative visualizations. The resulting saliency maps from our experiments proved to be 3-4 times better human interpretable than the state-of-the-art in ABM. At the same time it reserves machine interpretability, which is the average confidence scores in classification.