codeslake / refvsr-cvpr2022

Super-resolves an LR video frame (ultra-wide) using a reference video frame (wide-angle)

  • Public
  • 14.3K runs
  • GitHub
  • Paper
  • License

Run time and cost

This model costs approximately $0.0021 to run on Replicate, or 476 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia T4 GPU hardware. Predictions typically complete within 10 seconds.

Readme

Demo for Reference-based Video Super-Resolution (RefVSR)
Official PyTorch Implementation of the CVPR 2022 Paper
Project | arXiv | RealMCVSR Dataset

  • The model used for the demo is Ours-8K, which is trained with the proposed training strategy that consists of pretraining and adaptation stages.
  • Due to the memory issue, input frames will be center-cropped to have 1280x720 resolution.


abstract

We propose the first reference-based video super-resolution (RefVSR) approach that utilizes reference videos for high-fidelity results. We focus on RefVSR in a triple-camera setting, where we aim at super-resolving a low-resolution ultra-wide video utilizing wide-angle and telephoto videos. We introduce the first RefVSR network that recurrently aligns and propagates temporal reference features fused with features extracted from low-resolution frames. To facilitate the fusion and propagation of temporal reference features, we propose a propagative temporal fusion module. For learning and evaluation of our network, we present the first RefVSR dataset consisting of triplets of ultra-wide, wide-angle, and telephoto videos concurrently taken from triple cameras of a smartphone. We also propose a two-stage training strategy fully utilizing video triplets in the proposed dataset for real-world 4x video super-resolution. We extensively evaluate our method, and the result shows the state-of-the-art performance in 4x super-resolution.

Citation

If you find this demo useful, please consider citing:

@InProceedings{Lee2022RefVSR,
    author    = {Junyong Lee and Myeonghee Lee and Sunghyun Cho and Seungyong Lee},
    title     = {Reference-based Video Super-Resolution Using Multi-Camera Video Triplets},
    booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
    year      = {2022}
}