#WGAN read me files staring with combined_ are images for training and testing _rpr have 88% of pixels removed _ have random shapes added to the image test_out/images files have WGAN outputs; compare outputs and targets(ground truth) /tools are used for the pix2pix.py files pix2pix.py does image inpainting using GAN with inputs of pairs of images (amputated, ground truth) pix2pix.py_wgangp implements WGAN-GP mse_calc.py calculates MSE