English

Inversion-DPO: Precise and Efficient Post-Training for Diffusion Models

Computer Vision and Pattern Recognition 2025-08-05 v4 Artificial Intelligence

Abstract

Recent advancements in diffusion models (DMs) have been propelled by alignment methods that post-train models to better conform to human preferences. However, these approaches typically require computation-intensive training of a base model and a reward model, which not only incurs substantial computational overhead but may also compromise model accuracy and training efficiency. To address these limitations, we propose Inversion-DPO, a novel alignment framework that circumvents reward modeling by reformulating Direct Preference Optimization (DPO) with DDIM inversion for DMs. Our method conducts intractable posterior sampling in Diffusion-DPO with the deterministic inversion from winning and losing samples to noise and thus derive a new post-training paradigm. This paradigm eliminates the need for auxiliary reward models or inaccurate appromixation, significantly enhancing both precision and efficiency of training. We apply Inversion-DPO to a basic task of text-to-image generation and a challenging task of compositional image generation. Extensive experiments show substantial performance improvements achieved by Inversion-DPO compared to existing post-training methods and highlight the ability of the trained generative models to generate high-fidelity compositionally coherent images. For the post-training of compostitional image geneation, we curate a paired dataset consisting of 11,140 images with complex structural annotations and comprehensive scores, designed to enhance the compositional capabilities of generative models. Inversion-DPO explores a new avenue for efficient, high-precision alignment in diffusion models, advancing their applicability to complex realistic generation tasks. Our code is available at https://github.com/MIGHTYEZ/Inversion-DPO

Keywords

Cite

@article{arxiv.2507.11554,
  title  = {Inversion-DPO: Precise and Efficient Post-Training for Diffusion Models},
  author = {Zejian Li and Yize Li and Chenye Meng and Zhongni Liu and Yang Ling and Shengyuan Zhang and Guang Yang and Changyuan Yang and Zhiyuan Yang and Lingyun Sun},
  journal= {arXiv preprint arXiv:2507.11554},
  year   = {2025}
}

Comments

Accepted by ACM MM25