RectifID：透過錨定分類器引導個性化修正流程

RectifID: Personalizing Rectified Flow with Anchored Classifier Guidance

May 23, 2024

作者: Zhicheng Sun, Zhenhao Yang, Yang Jin, Haozhe Chi, Kun Xu, Kun Xu, Liwei Chen, Hao Jiang, Di Zhang, Yang Song, Kun Gai, Yadong Mu

cs.AI

摘要

將擴散模型定制化以從用戶提供的參考圖像生成保持身份的圖像是一個引人入勝的新問題。目前的方法通常需要在大量特定領域的圖像上進行訓練，以實現身份保留，這缺乏跨不同用例的靈活性。為了解決這個問題，我們利用分類器引導，這是一種無需訓練的技術，用現有的分類器引導擴散模型進行個性化圖像生成。我們的研究表明，基於最近的矯正流框架，普通分類器引導在需要特殊分類器方面的主要限制可以通過一個簡單的固定點解決方案來解決，從而允許使用現成的圖像鑑別器進行靈活的個性化。此外，當錨定到參考流軌跡時，其求解過程被證明是穩定的，具有收斂保證。所得方法應用於具有不同現成圖像鑑別器的矯正流上，為人臉、真人主題和某些對象提供了有利的個性化結果。代碼可在 https://github.com/feifeiobama/RectifID 找到。

English

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.