Eye-for-an-eye: Appearance Transfer with Semantic Correspondence in Diffusion Models
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_ | 1866911218612043776 |
|---|---|
| author | Go, Sooyeon Choi, Kyungmook Shin, Minjung Uh, Youngjung |
| author_facet | Go, Sooyeon Choi, Kyungmook Shin, Minjung Uh, Youngjung |
| contents | As pre-trained text-to-image diffusion models have become a useful tool for image synthesis, people want to specify the results in various ways. This paper tackles training-free appearance transfer, which produces an image with the structure of a target image from the appearance of a reference image. Existing methods usually do not reflect semantic correspondence, as they rely on query-key similarity within the self-attention layer to establish correspondences between images. To this end, we propose explicitly rearranging the features according to the dense semantic correspondences. Extensive experiments show the superiority of our method in various aspects: preserving the structure of the target and reflecting the correct color from the reference, even when the two images are not aligned. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_07008 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Eye-for-an-eye: Appearance Transfer with Semantic Correspondence in Diffusion Models Go, Sooyeon Choi, Kyungmook Shin, Minjung Uh, Youngjung Computer Vision and Pattern Recognition Artificial Intelligence As pre-trained text-to-image diffusion models have become a useful tool for image synthesis, people want to specify the results in various ways. This paper tackles training-free appearance transfer, which produces an image with the structure of a target image from the appearance of a reference image. Existing methods usually do not reflect semantic correspondence, as they rely on query-key similarity within the self-attention layer to establish correspondences between images. To this end, we propose explicitly rearranging the features according to the dense semantic correspondences. Extensive experiments show the superiority of our method in various aspects: preserving the structure of the target and reflecting the correct color from the reference, even when the two images are not aligned. |
| title | Eye-for-an-eye: Appearance Transfer with Semantic Correspondence in Diffusion Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2406.07008 |