Pytorch implementation of WIPA: Super-resolution of very low-resolution face images with a Wavelet Integrated, Identity Preserving, Adversarial Network.
Great work with this one. Will it work for scale factors smaller than x8, like if I wanted to compare it to other SR algorithms that work with a scale factor of 4. If I train the model at a scale factor of x4, will it still train normally?
Hi, Thanks to your amazing work. I was trying to test the model, I put the pretrained weight exactly as should be according to test file details but still I have bellow error, would you please let me know what is the problem?
root: pretrained>gen_net_8x or 16x>archive...
Error:
Exception: the per-trained generator file is not found.