Diffusion Model with Perceptual Loss
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910864160849920 |
|---|---|
| author | Lin, Shanchuan Yang, Xiao |
| author_facet | Lin, Shanchuan Yang, Xiao |
| contents | Diffusion models without guidance generate very unrealistic samples. Guidance is used ubiquitously, and previous research has attributed its effect to low-temperature sampling that improves quality by trading off diversity. However, this perspective is incomplete. Our research shows that the choice of the loss objective is the underlying reason raw diffusion models fail to generate desirable samples. In this paper, (1) our analysis shows that the loss objective plays an important role in shaping the learned distribution and the MSE loss derived from theories holds assumptions that misalign with data in practice; (2) we explain the effectiveness of guidance methods from a new perspective of perceptual supervision; (3) we validate our hypothesis by training a diffusion model with a novel self-perceptual loss objective and obtaining much more realistic samples without the need for guidance. We hope our work paves the way for future explorations of the diffusion loss objective. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_00110 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Diffusion Model with Perceptual Loss Lin, Shanchuan Yang, Xiao Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Diffusion models without guidance generate very unrealistic samples. Guidance is used ubiquitously, and previous research has attributed its effect to low-temperature sampling that improves quality by trading off diversity. However, this perspective is incomplete. Our research shows that the choice of the loss objective is the underlying reason raw diffusion models fail to generate desirable samples. In this paper, (1) our analysis shows that the loss objective plays an important role in shaping the learned distribution and the MSE loss derived from theories holds assumptions that misalign with data in practice; (2) we explain the effectiveness of guidance methods from a new perspective of perceptual supervision; (3) we validate our hypothesis by training a diffusion model with a novel self-perceptual loss objective and obtaining much more realistic samples without the need for guidance. We hope our work paves the way for future explorations of the diffusion loss objective. |
| title | Diffusion Model with Perceptual Loss |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2401.00110 |