Pytorch implementation of the model proposed by the IVL team for the AIM 2022 challenge on super-resolution of compressed videos (ECCV22 Workshops).
Ubuntu 22.04, Python 3.7.13, CUDA 11.6.
For Python requirements, see requirements.txt.
The dataset for the challenge can be downloaded here.
The code for training will be released soon.
To reproduce the results, execute
python test_aim_challenge.py --data_path your_dataset_path
. By default, the script will output only frames 10, 20, 30 etc. If you want to output all frames, add --save_all True
.
The default output directory is "./Results/"
@inproceedings{yang2022aim,
title={AIM 2022 Challenge on Super-resolution of Compressed Image and Video: Dataset, Methods and Results},
author={Yang, Ren and Timofte, Radu and others},
booktitle={European Conference on Computer Vision Workshops},
year={2022}
}
Our method is based on STDF.
For any question, please write an email to c.rota30@campus.unimib.it