Advancing Ante-Hoc Explainable Models through Generative Adversarial Networks

arXiv:2401.04647v2 Announce Type: replace-cross
Abstract: This paper presents a novel concept learning framework for enhancing model interpretability and performance in visual classification tasks. Our approach appends an unsupervised explanation generator to the primary classifier network and makes use of adversarial training. During training, the explanation module is optimized to extract visual concepts from the classifier's latent representations, while the GAN-based module aims to discriminate images generated from concepts, from true images. This joint training scheme enables the model to implicitly …

abstract adversarial adversarial training arxiv classification classifier concept concepts cs.ai cs.cv cs.lg extract framework generative generative adversarial networks generator interpretability model interpretability network networks novel paper performance tasks through training type unsupervised visual visual concepts