FSRCNN-TensorFlow

TensorFlow implementation of the Fast Super-Resolution Convolutional Neural Network (FSRCNN). This implements two models: FSRCNN which is more accurate but slower and FSRCNN-s which is faster but less accurate. Based on this project.

Prerequisites

• Python 3
• TensorFlow-gpu >= 1.3
• CUDA & cuDNN >= 6.0
• h5py
• Pillow

Usage

For training: python main.py
For testing: python main.py --train False

To use FSCRNN-s instead of FSCRNN: python main.py --fast True

Can specify epochs, learning rate, data directory, etc:
python main.py --epoch 100 --learning_rate 0.0002 --data_dir Train
Check main.py for all the possible flags

Also includes script expand_data.py which scales and rotates all the images in the specified training set to expand it

Result

Original butterfly image:

Ewa_lanczos interpolated image:

Super-resolved image:

Additional datasets

• General-100

TODO

• Add RGB support (Increase each layer depth to 3)
• Speed up pre-processing for large datasets

References

• tegg89/SRCNN-Tensorflow

• liliumao/Tensorflow-srcnn

• carpedm20/DCGAN-tensorflow