SeedVR 7B

--- license: apache-2.0 library_name: seedvr pipeline_tag: video-to-video --- <div align="center"> <img src="assets/seedvr_logo.png" alt="SeedVR" width="400"/> </div> # SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restoration > [Jianyi Wang](https://iceclear.github.io), [Zhijie Lin](https://scholar.google.com/citations?user=xXMj6_EAAAAJ&hl=zh-CN), [Meng Wei](https://openreview.net/profile?id=~Meng_Wei11), [Ceyuan Yang](https://scholar.google.com/citations?user=uPmTOHAAAAAJ&hl=zh-CN), [Fei Xiao](https://openreview.net/profile?id=~Fei_xiao8), [Chen Change Loy](https://www.mmlab-ntu.com/person/ccloy/), [Lu Jiang](http://www.lujiang.info/) > > **CVPR 2025 (Highlight)** <p align="center"> <a href="https://iceclear.github.io/projects/seedvr/"> <img src="https://img.shields.io/badge/SeedVR-Website-0A66C2?logo=safari&logoColor=white" alt="SeedVR Website" /> </a> <a href="https://arxiv.org/abs/2501.01320"> <img src="https://img.shields.io/badge/SeedVR-Paper-red?logo=arxiv&logoColor=red" alt="SeedVR Paper on ArXiv" /> </a> <a href="https://github.com/ByteDance-Seed/SeedVR"> <img alt="Github" src="https://img.shields.io/badge/SeedVR2-Codebase-536af5?color=536af5&logo=github" alt="SeedVR2 Codebase" /> </a> <a href="https://huggingface.co/collections/ByteDance-Seed/seedvr-6849deeb461c4e425f3e6f9e"> <img src="https://img.shields.io/badge/SeedVR-Models-yellow?logo=huggingface&logoColor=yellow" alt="SeedVR Models" /> </a> <a href="https://huggingface.co/spaces/ByteDance-Seed/SeedVR2-3B"> <img src="https://img.shields.io/badge/SeedVR2-Space-orange?logo=huggingface&logoColor=yellow" alt="SeedVR2 Space" /> </a> <a href="https://www.youtube.com/watch?v=aPpBs_B2iCY" target='_blank'> <img src="https://img.shields.io/badge/Demo%20Video-%23FF0000.svg?logo=YouTube&logoColor=white" alt="SeedVR Video Demo on YouTube" /> </a> </p> > > **Why SeedVR:** Conventional restoration models achieve inferior performance on both real-world and AIGC video restoration due to limited generation ability. Recent diffusion-based models improve the performance by introducing diffusion prior via ControlNet-like or adaptor-like architectures. Though gaining improvement, these methods generally suffer from constraints brought by the diffusion prior: these models suffer from the same bias as the prior, e.g., limited generation ability on...

Source context: 87 downloads · 13 likes · Pipeline video-to-video · Library seedvr · Repo ByteDance-Seed/SeedVR-7B