Title: SCFlow: Implicitly Learning Style and Content Disentanglement with Flow Models
Description: We learn disentangled representations implicitly by training only on merging them.
Keywords:
Domain: compvis.github.io
Links:
| ArXiv | https://arxiv.org/abs/2508.03402 |
| ICCV Paper | https://openaccess.thecvf.com/content/ICCV2025/html/Ma_SCFlow_Implicitly_Learning_Style_and_Content_Disentanglement_with_Flow_Models_ICCV_2025_paper.html |
| Code | https://github.com/CompVis/SCFlow |
| Stochastic Interpolants for Revealing Stylistic Flows across the History of Art | https://compvis.github.io/Art-fm/ |
| What If: Understanding Motion Through Sparse Interactions | https://compvis.github.io/flow-poke-transformer/ |
| Tread: Token Routing for Efficient Architecture-agnostic Diffusion Training | https://compvis.github.io/tread/ |
| Academic Project Page Template | https://github.com/eliahuhorwitz/Academic-project-page-template |
| Nerfies | https://nerfies.github.io |
| Creative Commons Attribution-ShareAlike 4.0 International License | http://creativecommons.org/licenses/by-sa/4.0/ |
Viewport: width=device-width, initial-scale=1