Papers
arxiv:2503.10638

Studying Classifier(-Free) Guidance From a Classifier-Centric Perspective

Published on Mar 13
· Submitted by xzhao on Mar 14
Authors:
,

Abstract

Classifier-free guidance has become a staple for conditional generation with denoising diffusion models. However, a comprehensive understanding of classifier-free guidance is still missing. In this work, we carry out an empirical study to provide a fresh perspective on classifier-free guidance. Concretely, instead of solely focusing on classifier-free guidance, we trace back to the root, i.e., classifier guidance, pinpoint the key assumption for the derivation, and conduct a systematic study to understand the role of the classifier. We find that both classifier guidance and classifier-free guidance achieve conditional generation by pushing the denoising diffusion trajectories away from decision boundaries, i.e., areas where conditional information is usually entangled and is hard to learn. Based on this classifier-centric understanding, we propose a generic postprocessing step built upon flow-matching to shrink the gap between the learned distribution for a pre-trained denoising diffusion model and the real data distribution, majorly around the decision boundaries. Experiments on various datasets verify the effectiveness of the proposed approach.

Community

Paper submitter

This paper advances the ongoing effort to achieve a comprehensive understanding of classifier-free guidance. The authors reveal that both classifier guidance and classifier-free guidance enable conditional generation by steering denoising diffusion trajectories away from the decision boundaries of the data distribution.

Your need to confirm your account before you can post a new comment.

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2503.10638 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2503.10638 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2503.10638 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.