parkseobin / MLSR

Source code for ECCV2020 "Fast Adaptation to Super-Resolution Networks via Meta-Learning"

Fast Adaptation to Super-Resolution Networks via Meta-Learning

Source code for ECCV2020 "Fast Adaptation to Super-Resolution Networks via Meta-Learning" paper

Requirements

Check requirements.txt to install all requirements.

conda install --file requirements.txt

or

pip install -i requirements.txt

Usage

Training with Urban100 dataset

• Download Urban100 dataset here
• Set Urban100 dataset directory name to Urban100 and run ./split_urban100.sh
• Download IDN pretrained weights checkpoint_x2 here
• Start training

python main.py --param-restore-path checkpoint_x2 --param-save-path mlsr_test_parameter

• You can also train with other dataset by using --train-dataset flag.

Citation

If you find MLSR helpful, please consider citing our paper:

@article{park2020fast,
    title={Fast Adaptation to Super-Resolution Networks via Meta-Learning},
    author={Park, Seobin and Yoo, Jinsu and Cho, Donghyeon and Kim, Jiwon and Kim, Tae Hyun},
    journal={ECCV},
    year={2020}
}