Explaining decision of model from its prediction

TL;DR

This paper reviews various visual explanation techniques for models, comparing their methods and results to understand how models make decisions, aiding interpretability.

Contribution

It provides a comprehensive summary and comparison of multiple visual explanation methods, highlighting their differences and effectiveness.

Findings

Comparison of CAM, Grad-CAM, and Guided Backpropagation results

Summary of various saliency and activation visualization techniques

Insights into the strengths and limitations of each method

Abstract

This document summarizes different visual explanations methods such as CAM, Grad-CAM, Localization using Multiple Instance Learning - Saliency-based methods, Saliency-driven Class-Impressions, Muting pixels in input image - Adversarial methods and Activation visualization, Convolution filter visualization - Feature-based methods. We have also shown the results produced by different methods and a comparison between CAM, GradCAM, and Guided Backpropagation.