viraj-shah18 / SRFBN

PyTorch Implementation for "Feedback Network for Image Super-Resolution" paper.

Downloading Datasets

Div2K dataset - Link
To download Set5 Dataset, run

wget https://data.deepai.org/set5.zip

Installing Libraries

For PyTorch please visit this link to install proper version of PyTorch. After that, run the following command to install all the required libraries

pip install -r requirements.txt

Preparing Dataset and Augmenting Data

To prepare the dataset, change the parameters at the top of the file and run

python3 prepare_data.py

Training

For Training, several modes are available, please have a look at the top of the train.py file to decide some of the parameters to train the model with. Once the parameters are finalized, run

python3 train.py

Testing

For testing, first finalize the parameters given at the begining of the test_model.py file. Once finalized run

python3 test_model.py