Optimizing for the Shortest Path in Denoising Diffusion Model
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866929756936601600 |
|---|---|
| author | Chen, Ping Zhang, Xingpeng Liu, Zhaoxiang Hu, Huan Liu, Xiang Wang, Kai Wang, Min Qian, Yanlin Lian, Shiguo |
| author_facet | Chen, Ping Zhang, Xingpeng Liu, Zhaoxiang Hu, Huan Liu, Xiang Wang, Kai Wang, Min Qian, Yanlin Lian, Shiguo |
| contents | In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_03265 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Optimizing for the Shortest Path in Denoising Diffusion Model Chen, Ping Zhang, Xingpeng Liu, Zhaoxiang Hu, Huan Liu, Xiang Wang, Kai Wang, Min Qian, Yanlin Lian, Shiguo Computer Vision and Pattern Recognition In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF. |
| title | Optimizing for the Shortest Path in Denoising Diffusion Model |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.03265 |