FluxSpace: Disentangled Semantic Editing in Rectified Flow Transformers
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866929628051931136 |
|---|---|
| author | Dalva, Yusuf Venkatesh, Kavana Yanardag, Pinar |
| author_facet | Dalva, Yusuf Venkatesh, Kavana Yanardag, Pinar |
| contents | Rectified flow models have emerged as a dominant approach in image generation, showcasing impressive capabilities in high-quality image synthesis. However, despite their effectiveness in visual generation, rectified flow models often struggle with disentangled editing of images. This limitation prevents the ability to perform precise, attribute-specific modifications without affecting unrelated aspects of the image. In this paper, we introduce FluxSpace, a domain-agnostic image editing method leveraging a representation space with the ability to control the semantics of images generated by rectified flow transformers, such as Flux. By leveraging the representations learned by the transformer blocks within the rectified flow models, we propose a set of semantically interpretable representations that enable a wide range of image editing tasks, from fine-grained image editing to artistic creation. This work offers a scalable and effective image editing approach, along with its disentanglement capabilities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_09611 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | FluxSpace: Disentangled Semantic Editing in Rectified Flow Transformers Dalva, Yusuf Venkatesh, Kavana Yanardag, Pinar Computer Vision and Pattern Recognition Rectified flow models have emerged as a dominant approach in image generation, showcasing impressive capabilities in high-quality image synthesis. However, despite their effectiveness in visual generation, rectified flow models often struggle with disentangled editing of images. This limitation prevents the ability to perform precise, attribute-specific modifications without affecting unrelated aspects of the image. In this paper, we introduce FluxSpace, a domain-agnostic image editing method leveraging a representation space with the ability to control the semantics of images generated by rectified flow transformers, such as Flux. By leveraging the representations learned by the transformer blocks within the rectified flow models, we propose a set of semantically interpretable representations that enable a wide range of image editing tasks, from fine-grained image editing to artistic creation. This work offers a scalable and effective image editing approach, along with its disentanglement capabilities. |
| title | FluxSpace: Disentangled Semantic Editing in Rectified Flow Transformers |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.09611 |