Adversarial Concept Distillation for One-Step Diffusion Personalization
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918437247254528 |
|---|---|
| author | Yang, Yixiong Wu, Tao Li, Senmao Yang, Shiqi Wang, Yaxing van de Weijer, Joost Wang, Kai |
| author_facet | Yang, Yixiong Wu, Tao Li, Senmao Yang, Shiqi Wang, Yaxing van de Weijer, Joost Wang, Kai |
| contents | Recent progress in accelerating text-to-image diffusion models enables high-fidelity synthesis within a single denoising step. However, customizing the fast one-step models remains challenging, as existing methods consistently fail to produce acceptable results, underscoring the need for new methodologies to personalize one-step models. Therefore, we propose One-step Personalized Adversarial Distillation (OPAD), a framework that combines teacher-student distillation with adversarial supervision. A multi-step diffusion model serves as the teacher, while a one-step student model is jointly trained with it. The student learns from alignment losses that preserve consistency with the teacher and from adversarial losses that align its output with real image distributions. Beyond one-step personalization, we further observe that the student's efficient generation and adversarially enriched representations provide valuable feedback to improve the teacher model, forming a collaborative learning stage. Extensive experiments demonstrate that OPAD is the first approach to deliver reliable, high-quality personalization for one-step diffusion models; in contrast, prior methods largely fail and produce severe failure cases, while OPAD preserves single-step efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_20512 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Adversarial Concept Distillation for One-Step Diffusion Personalization Yang, Yixiong Wu, Tao Li, Senmao Yang, Shiqi Wang, Yaxing van de Weijer, Joost Wang, Kai Computer Vision and Pattern Recognition Recent progress in accelerating text-to-image diffusion models enables high-fidelity synthesis within a single denoising step. However, customizing the fast one-step models remains challenging, as existing methods consistently fail to produce acceptable results, underscoring the need for new methodologies to personalize one-step models. Therefore, we propose One-step Personalized Adversarial Distillation (OPAD), a framework that combines teacher-student distillation with adversarial supervision. A multi-step diffusion model serves as the teacher, while a one-step student model is jointly trained with it. The student learns from alignment losses that preserve consistency with the teacher and from adversarial losses that align its output with real image distributions. Beyond one-step personalization, we further observe that the student's efficient generation and adversarially enriched representations provide valuable feedback to improve the teacher model, forming a collaborative learning stage. Extensive experiments demonstrate that OPAD is the first approach to deliver reliable, high-quality personalization for one-step diffusion models; in contrast, prior methods largely fail and produce severe failure cases, while OPAD preserves single-step efficiency. |
| title | Adversarial Concept Distillation for One-Step Diffusion Personalization |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.20512 |