Super Resolution Convolutional Neural Network implementation using DIV2K and Set 14 datasets
This project is an implementation of Image Super-Resolution Using Deep Convolutional Networks .
You can study the paper provided above and srcnn_report.pdf file to understand the underlying theoretical aspects.
To execute the project successfully:
- Download and install cuda toolkit and the cuda version that your NVIDIA-GPU supports.
- Download the Set 14 dataset .
- Download the DIV2K dataset .
- Change the paths in main.py, in test.py and in test_scripts.py to your corresponding local directories.
- Execute main.py to train the model.
- Execute test.py to see how good your training performed given a good image and test it versus its bicubic interpolated counterpart.
- Execute test_scripts.py for single image super resolution (SISR).