RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance

TL;DR

This paper introduces RectifID, a method that personalizes rectified flow-based diffusion models for identity-preserving image generation using classifier guidance without extensive retraining.

Contribution

It proposes a simple fixed-point solution to adapt classifier guidance in rectified flow, enabling flexible, training-free personalization with off-the-shelf discriminators.

Findings

Effective personalization for faces, live subjects, and objects.

Stable convergence when anchored to reference flow trajectories.

Outperforms existing methods in identity preservation.

Abstract

Customizing diffusion models to generate identity-preserving images from user-provided reference images is an intriguing new problem. The prevalent approaches typically require training on extensive domain-specific images to achieve identity preservation, which lacks flexibility across different use cases. To address this issue, we exploit classifier guidance, a training-free technique that steers diffusion models using an existing classifier, for personalized image generation. Our study shows that based on a recent rectified flow framework, the major limitation of vanilla classifier guidance in requiring a special classifier can be resolved with a simple fixed-point solution, allowing flexible personalization with off-the-shelf image discriminators. Moreover, its solving procedure proves to be stable when anchored to a reference flow trajectory, with a convergence guarantee. The derived method is implemented on rectified flow with different off-the-shelf image discriminators, delivering advantageous personalization results for human faces, live subjects, and certain objects. Code is available at https://github.com/feifeiobama/RectifID.