Does Diffusion Beat GAN in Image Super Resolution?
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916455621066752 |
|---|---|
| author | Kuznedelev, Denis Startsev, Valerii Shlenskii, Daniil Kastryulin, Sergey |
| author_facet | Kuznedelev, Denis Startsev, Valerii Shlenskii, Daniil Kastryulin, Sergey |
| contents | There is a prevalent opinion that diffusion-based models outperform GAN-based counterparts in the Image Super Resolution (ISR) problem. However, in most studies, diffusion-based ISR models employ larger networks and are trained longer than the GAN baselines. This raises the question of whether the high performance stems from the superiority of the diffusion paradigm or if it is a consequence of the increased scale and the greater computational resources of the contemporary studies. In our work, we thoroughly compare diffusion-based and GAN-based Super Resolution models under controlled settings, with both approaches having matched architecture, model and dataset sizes, and computational budget. We show that a GAN-based model can achieve results comparable or superior to a diffusion-based model. Additionally, we explore the impact of popular design choices, such as text conditioning and augmentation on the performance of ISR models, showcasing their effect in several downstream tasks. We will release the inference code and weights of our scaled GAN. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_17261 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Does Diffusion Beat GAN in Image Super Resolution? Kuznedelev, Denis Startsev, Valerii Shlenskii, Daniil Kastryulin, Sergey Image and Video Processing Computer Vision and Pattern Recognition There is a prevalent opinion that diffusion-based models outperform GAN-based counterparts in the Image Super Resolution (ISR) problem. However, in most studies, diffusion-based ISR models employ larger networks and are trained longer than the GAN baselines. This raises the question of whether the high performance stems from the superiority of the diffusion paradigm or if it is a consequence of the increased scale and the greater computational resources of the contemporary studies. In our work, we thoroughly compare diffusion-based and GAN-based Super Resolution models under controlled settings, with both approaches having matched architecture, model and dataset sizes, and computational budget. We show that a GAN-based model can achieve results comparable or superior to a diffusion-based model. Additionally, we explore the impact of popular design choices, such as text conditioning and augmentation on the performance of ISR models, showcasing their effect in several downstream tasks. We will release the inference code and weights of our scaled GAN. |
| title | Does Diffusion Beat GAN in Image Super Resolution? |
| topic | Image and Video Processing Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2405.17261 |