Awesome

Towards Interpretable Video Super-Resolution via Alternating Optimization (ECCV 2022)

Jiezhang Cao, Jingyun Liang, Kai Zhang, Wenguan Wang, Qin Wang, Yulun Zhang, Hao Tang, Luc Van Gool

Computer Vision Lab, ETH Zurich.

arxiv | supplementary | pretrained models | visual results

This repository is the official PyTorch implementation of "Towards Interpretable Video Super-Resolution via Alternating Optimization" (arxiv, supp, pretrained models, visual results). DAVSR achieves state-of-the-art performance in practical time-space video super-resolution.

In this paper, we study a practical space-time video super-resolution (STVSR) problem which aims at generating a high-framerate high-resolution sharp video from a low-framerate low-resolution blurry video. Such problem often occurs when recording a fast dynamic event with a low-framerate and low-resolution camera, and the captured video would suffer from three typical issues: i) motion blur occurs due to object/camera motions during exposure time; ii) motion aliasing is unavoidable when the event temporal frequency exceeds the Nyquist limit of temporal sampling; iii) high-frequency details are lost because of the low spatial sampling rate. These issues can be alleviated by a cascade of three separate sub-tasks, including video deblurring, frame interpolation, and super-resolution, which, however, would fail to capture the spatial and temporal correlations among video sequences. To address this, we propose an interpretable STVSR framework by leveraging both model-based and learning-based methods. Specifically, we formulate STVSR as a joint video deblurring, frame interpolation, and super-resolution problem, and solve it as two sub-problems in an alternate way. For the first sub-problem, we derive an interpretable analytical solution and use it as a Fourier data transform layer. Then, we propose a recurrent video enhancement layer for the second sub-problem to further recover high-frequency details. Extensive experiments demonstrate the superiority of our method in terms of quantitative metrics and visual quality.

Contents

1. Requirements
2. Quick Testing
3. Training
4. Results
5. Citation
6. License and Acknowledgement

TODO

• Add pretrained model
• Add results of test set

Requirements

• Python 3.8, PyTorch >= 1.9.1
• mmedit 0.11.0
• Requirements: see requirements.txt
• Platforms: Ubuntu 18.04, cuda-11.1

Quick Testing

Following commands will download pretrained models and test datasets. If out-of-memory, try to reduce num_frame_testing and size_patch_testing at the expense of slightly decreased performance.

# download code
git clone https://github.com/caojiezhang/DAVSR
cd DAVSR
pip install -r requirements.txt

PYTHONPATH=/bin/..:tools/..: python tools/test.py configs/restorers/uvsrnet/002_pretrain_uvsr3DBDnet_REDS_25frames_3iter_sf544_slomo_modify_newdataset.py work_dirs/002_pretrain_uvsr3DBDnet_REDS_25frames_3iter_sf544_slomo_modify_newdataset/latest.pth

All visual results of DAVSR can be downloaded here.

Dataset

The training and testing sets are as follows (see the supplementary for a detailed introduction of all datasets). For better I/O speed, use create_lmdb.py to convert .png datasets to .lmdb datasets.

Note: You do NOT need to prepare the datasets if you just want to test the model. main_test_vrt.py will download the testing set automaticaly.

Training

PYTHONPATH=/bin/..:tools/..: ./tools/dist_train.sh configs/restorers/uvsrnet/002_pretrain_uvsr3DBDnet_REDS_25frames_3iter_sf544_slomo_modify_newdataset.py 8

Results

We achieved state-of-the-art performance on practical space-time video super-resolution. Detailed results can be found in the paper.

Citation

@inproceedings{cao2022davsr,
  title={Towards Interpretable Video Super-Resolution via Alternating Optimization},
  author={Cao, Jiezhang and Liang, Jingyun and Zhang, Kai and Wang, Wenguan and Wang, Qin  and Zhang, Yulun and Tang, Hao and Van Gool, Luc},
  booktitle={European conference on computer vision},
  year={2022}
}

License and Acknowledgement

This project is released under the CC-BY-NC license. We refer to codes from KAIR, BasicSR, and mmediting. Thanks for their awesome works. The majority of DAVSR is licensed under CC-BY-NC, however portions of the project are available under separate license terms: KAIR is licensed under the MIT License, BasicSR, Video Swin Transformer and mmediting are licensed under the Apache 2.0 license.