FAME: Feature Activation Map Explanation on Image Classification and Face Recognition

TL;DR

FAME is a novel explanation method for deep image models that combines gradient-based and activation map approaches, providing more accurate attributions for image classification and face recognition.

Contribution

FAME introduces a gradient-driven technique that improves attribution maps by combining CAM and perturbation-based methods, especially for deep networks.

Findings

FAME produces attribution maps that are competitive with state-of-the-art methods.

CAM assumptions do not hold for deeper networks, as shown by FAME.

FAME is applicable to both image classification and face recognition tasks.

Abstract

Deep Learning has revolutionized machine learning, reaching unprecedented levels of accuracy, but at the cost of reduced interpretability. Especially in image processing systems, deep networks transform local pixel information into more global concepts in a highly obscured manner. Explainable AI methods for image processing try to shed light on this issue by highlighting the regions of the image that are important for the prediction task. Among these, Class Activation Mapping (CAM) and its gradient-based variants compute attributions based on the feature map and upscale them to the image resolution, assuming that feature map locations are influenced only by underlying regions. Perturbation-based methods, such as CorrRISE, on the other hand, try to provide pixel-level attributions by perturbing the input with fixed patches and checking how the output of the network changes. In this work, we propose Feature Activation Map Explanation (FAME), which combines both worlds by using network gradients to compute changes to the input image, manipulating it in a gradient-driven way rather than using fixed patches. We apply this technique on two common tasks, image classification and face recognition, and show that CAM's above-mentioned assumption does not hold for deeper networks. We qualitatively and quantitively show that FAME produces attribution maps that are competitive state-of-the-art systems. Our code is available: {\footnotesize https://github.com/AIML-IfI/fame.}