# PatchVSR: Breaking Video Diffusion Resolution Limits with Patch-wise Video Super-Resolution

Source: [PatchVSR: Breaking Video Diffusion Resolution Limits with Patch-wise Video Super-Resolution](<https://www.sota2.com/research/paper/68fa7bc91e209131cea8a7a9>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-04T04:27:38.114Z

Authors: Shian Du, Menghan Xia, Chang Liu, Xintao Wang, Jing Wang, Pengfei Wan, Di Zhang, Xiangyang Ji

Published: 2025-09-30T09:55:14.000Z

## Abstract

Pre-trained video generation models hold great potential for generative video super-resolution (VSR). However, adapting them for full-size VSR, as most existing methods do, suffers from unnecessary intensive full-attention computation and fixed output resolution. To overcome these limitations, we make the first exploration into utilizing video diffusion priors for patch-wise VSR. This is non-trivial because pre-trained video diffusion models are not native for patch-level detail generation. To mitigate this challenge, we propose an innovative approach, called PatchVSR, which integrates a dual-stream adapter for conditional guidance. The patch branch extracts features from input patches to maintain content fidelity while the global branch extracts context features from the resized full video to bridge the generation gap caused by incomplete semantics of patches. Particularly, we also inject the patch's location information into the model to better contextualize patch synthesis within the global video frame. Experiments demonstrate that our method can synthesize high-fidelity, high-resolution details at the patch level. A tailor-made multi-patch joint modulation is proposed to ensure visual consistency across individually enhanced patches. Due to the flexibility of our patch-based paradigm, we can achieve highly competitive 4K VSR based on a 512x512 resolution base model, with extremely high efficiency.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Video Super-Resolution on VideoGen30 (test)](<https://www.sota2.com/research/sota/video-super-resolution-on-videogen30-test.md>): Video Super-Resolution; VideoGen30 (test); Visual Quality: 2.811.

- [Video Super-Resolution on VideoLQ (test)](<https://www.sota2.com/research/sota/video-super-resolution-on-videolq-test.md>): Video Super-Resolution; VideoLQ (test); DOVER Score: 0.501.

- [Video Super-Resolution on 2K Video](<https://www.sota2.com/research/sota/video-super-resolution-on-2k-video.md>): Video Super-Resolution; 2K Video; Runtime (s): 680.

- [Video Super-Resolution on REDS30 (test)](<https://www.sota2.com/research/sota/video-super-resolution-on-reds30-test.md>): Video Super-Resolution; REDS30 (test); PSNR: 30.111.

- [Video Super-Resolution on SynVideo30 (test)](<https://www.sota2.com/research/sota/video-super-resolution-on-synvideo30-test.md>): Video Super-Resolution; SynVideo30 (test); PSNR: 30.857.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2509.26025>)

[PDF](<https://arxiv.org/pdf/2509.26025.pdf>)