Assessment of the Reliablity of a Model's Decision by Generalizing Attribution to the Wavelet Domain
This work addresses the need for scientists and practitioners to trust model decisions in critical applications, though it is incremental as it builds on existing attribution methods.
The paper tackles the problem of assessing the reliability of neural network decisions in computer vision by introducing WCAM, a method that generalizes attribution to the wavelet domain, enabling evaluation of feature relevance and robustness to image corruptions.
Neural networks have shown remarkable performance in computer vision, but their deployment in numerous scientific and technical fields is challenging due to their black-box nature. Scientists and practitioners need to evaluate the reliability of a decision, i.e., to know simultaneously if a model relies on the relevant features and whether these features are robust to image corruptions. Existing attribution methods aim to provide human-understandable explanations by highlighting important regions in the image domain, but fail to fully characterize a decision process's reliability. To bridge this gap, we introduce the Wavelet sCale Attribution Method (WCAM), a generalization of attribution from the pixel domain to the space-scale domain using wavelet transforms. Attribution in the wavelet domain reveals where and on what scales the model focuses, thus enabling us to assess whether a decision is reliable. Our code is accessible here: \url{https://github.com/gabrielkasmi/spectral-attribution}.