CCS: Controllable and Constrained Sampling with Diffusion Models via Initial Noise Perturbation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912223392169984 |
|---|---|
| author | Song, Bowen Zhang, Zecheng Luo, Zhaoxu Hu, Jason Yuan, Wei Jia, Jing Tang, Zhengxu Wang, Guanyang Shen, Liyue |
| author_facet | Song, Bowen Zhang, Zecheng Luo, Zhaoxu Hu, Jason Yuan, Wei Jia, Jing Tang, Zhengxu Wang, Guanyang Shen, Liyue |
| contents | Diffusion models have emerged as powerful tools for generative tasks, producing high-quality outputs across diverse domains. However, how the generated data responds to the initial noise perturbation in diffusion models remains under-explored, which hinders understanding the controllability of the sampling process. In this work, we first observe an interesting phenomenon: the relationship between the change of generation outputs and the scale of initial noise perturbation is highly linear through the diffusion ODE sampling. Then we provide both theoretical and empirical study to justify this linearity property of this input-output (noise-generation data) relationship. Inspired by these new insights, we propose a novel Controllable and Constrained Sampling method (CCS) together with a new controller algorithm for diffusion models to sample with desired statistical properties while preserving good sample quality. We perform extensive experiments to compare our proposed sampling approach with other methods on both sampling controllability and sampled data quality. Results show that our CCS method achieves more precisely controlled sampling while maintaining superior sample quality and diversity. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_04670 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CCS: Controllable and Constrained Sampling with Diffusion Models via Initial Noise Perturbation Song, Bowen Zhang, Zecheng Luo, Zhaoxu Hu, Jason Yuan, Wei Jia, Jing Tang, Zhengxu Wang, Guanyang Shen, Liyue Machine Learning Artificial Intelligence Diffusion models have emerged as powerful tools for generative tasks, producing high-quality outputs across diverse domains. However, how the generated data responds to the initial noise perturbation in diffusion models remains under-explored, which hinders understanding the controllability of the sampling process. In this work, we first observe an interesting phenomenon: the relationship between the change of generation outputs and the scale of initial noise perturbation is highly linear through the diffusion ODE sampling. Then we provide both theoretical and empirical study to justify this linearity property of this input-output (noise-generation data) relationship. Inspired by these new insights, we propose a novel Controllable and Constrained Sampling method (CCS) together with a new controller algorithm for diffusion models to sample with desired statistical properties while preserving good sample quality. We perform extensive experiments to compare our proposed sampling approach with other methods on both sampling controllability and sampled data quality. Results show that our CCS method achieves more precisely controlled sampling while maintaining superior sample quality and diversity. |
| title | CCS: Controllable and Constrained Sampling with Diffusion Models via Initial Noise Perturbation |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2502.04670 |