Deep Integrated Explanations
This addresses the need for better interpretability in vision AI, though it appears incremental as an improved explanation method.
The paper tackles the problem of explaining vision models by introducing Deep Integrated Explanations (DIX), a universal method that integrates intermediate representations and gradients to generate explanation maps, and demonstrates its efficacy in surpassing state-of-the-art methods across diverse evaluations.
This paper presents Deep Integrated Explanations (DIX) - a universal method for explaining vision models. DIX generates explanation maps by integrating information from the intermediate representations of the model, coupled with their corresponding gradients. Through an extensive array of both objective and subjective evaluations spanning diverse tasks, datasets, and model configurations, we showcase the efficacy of DIX in generating faithful and accurate explanation maps, while surpassing current state-of-the-art methods.