Improving Discriminator Guidance in Diffusion Models

doi:10.1007/978-3-032-05981-9_14

Improving Discriminator Guidance in Diffusion Models

2026-01-01

0

PRE

AI

DOI：10.1007/978-3-032-05981-9_14

Original

Original request for help

Save

Abstract

En 中文

Discriminator Guidance has become a popular method for efficiently refining pre-trained Score-Matching Diffusion models. However, in this paper, we demonstrate that the standard implementation of this technique does not necessarily lead to a distribution closer to the real data distribution. Specifically, we show that training the discriminator using Cross-Entropy loss, as commonly done, can in fact increase the Kullback-Leibler divergence between the model and target distributions, particularly when the discriminator overfits. To address this, we propose a theoretically sound training objective for discriminator guidance that properly minimizes the KL divergence. We analyze its properties and demonstrate empirically across multiple datasets that our proposed method consistently improves over the conventional method by producing samples of higher quality. (Code: https://github.com/AlexVerine/BoostDM and Supplementary Materials https://arxiv.org/abs/2503.16117).

Keywords:

Diffusion Models

Discriminator Guidance

Improving Discriminator Guidance in Diffusion Models

Abstract

Journal

Organization