CycleGAN with Better Cycles
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_ | 1866915030479405056 |
|---|---|
| author | Wang, Tongzhou Lin, Yihan |
| author_facet | Wang, Tongzhou Lin, Yihan |
| contents | CycleGAN provides a framework to train image-to-image translation with unpaired datasets using cycle consistency loss [4]. While results are great in many applications, the pixel level cycle consistency can potentially be problematic and causes unrealistic images in certain cases. In this project, we propose three simple modifications to cycle consistency, and show that such an approach achieves better results with fewer artifacts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_15374 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CycleGAN with Better Cycles Wang, Tongzhou Lin, Yihan Computer Vision and Pattern Recognition Machine Learning CycleGAN provides a framework to train image-to-image translation with unpaired datasets using cycle consistency loss [4]. While results are great in many applications, the pixel level cycle consistency can potentially be problematic and causes unrealistic images in certain cases. In this project, we propose three simple modifications to cycle consistency, and show that such an approach achieves better results with fewer artifacts. |
| title | CycleGAN with Better Cycles |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2408.15374 |