Structure-Guided Adversarial Training of Diffusion Models

arXiv:2402.17563v1 Announce Type: new
Abstract: Diffusion models have demonstrated exceptional efficacy in various generative applications. While existing models focus on minimizing a weighted sum of denoising score matching losses for data distribution modeling, their training primarily emphasizes instance-level optimization, overlooking valuable structural information within each mini-batch, indicative of pair-wise relationships among samples. To address this limitation, we introduce Structure-guided Adversarial training of Diffusion Models (SADM). In this pioneering approach, we compel the model to learn manifold structures between samples in …

abstract adversarial adversarial training applications arxiv cs.cv data denoising diffusion diffusion models distribution focus generative indicative information instance losses modeling optimization relationships samples training type wise