Visual Feature Attribution using Wasserstein GANs
This addresses the problem of incomplete feature detection in visual attribution for computer vision applications, particularly in medical imaging for conditions like Alzheimer's disease, representing an incremental improvement over existing methods.
The paper tackled the limitation of existing neural network-based feature attribution methods that detect only a subset of category-specific features by developing a novel technique using Wasserstein GANs, which performed substantially better than state-of-the-art methods on synthetic and real neuroimaging data, producing realistic disease effect maps for Alzheimer's patients.
Attributing the pixels of an input image to a certain category is an important and well-studied problem in computer vision, with applications ranging from weakly supervised localisation to understanding hidden effects in the data. In recent years, approaches based on interpreting a previously trained neural network classifier have become the de facto state-of-the-art and are commonly used on medical as well as natural image datasets. In this paper, we discuss a limitation of these approaches which may lead to only a subset of the category specific features being detected. To address this problem we develop a novel feature attribution technique based on Wasserstein Generative Adversarial Networks (WGAN), which does not suffer from this limitation. We show that our proposed method performs substantially better than the state-of-the-art for visual attribution on a synthetic dataset and on real 3D neuroimaging data from patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD). For AD patients the method produces compellingly realistic disease effect maps which are very close to the observed effects.