A Chainer implementation of SRCNN by Chainer(v3) and python3. Please notice that I have used padding in the model, which is different from the original paper. If you have any question, please feel free to contact me.
It will downsize the given image to the low resolution image with a
factor=2, then upsize it by bicubic and SRCNN respectively to generate
super resolution image with a factor=2 and compare the PSNR/SSIM between
the SR image and ground truth.
generate
python generate_2x_srcnn --input_file/-i filename
It will generate a 2x SR image of the given image by SRCNN.