/FRFSR

A Feature Reuse Framework with Texture-adaptive Aggregation for Reference-based Super-Resolution

Primary LanguagePythonMIT LicenseMIT

FRFSR

A Feature Reuse Framework with Texture-adaptive Aggregation for Reference-based Super-Resolution arxiv

Overview

Overview

Dependencies

  • Python >=3.7
  • Pytorch == 1.7.1

Training and Test

Because our code is heavily derived from C2-Matching, you can easily use our code to replace and train or test on C2-matching.

Visual Results

result

Citation

@article{mei2023feature,
title={A Feature Reuse Framework with Texture-adaptive Aggregation for Reference-based Super-Resolution},
author={Mei, Xiaoyong and Yang, Yi and Li, Ming and Huang, Changqin and Zhang, Kai and Li{\'o}, Pietro},
journal={arXiv preprint arXiv:2306.01500},
year={2023}
}

License and Acknowledgement

This project is open sourced under MIT license. We refer to codes from C2-Matching, DATSR and BasicSR. Thanks for their awesome works.