Closed felixblanke closed 3 years ago
ahh yeah, I broke this sorry!
I am working on batched preprocessing to compute the wavelet packets efficiently ahead of the training.
Until that is ready we need it in the training code. I should not have removed it, sorry.
I fixed it. Results on the FFHQ - Stylegan pair look promising.
It appears to me that we are always training on the raw normalized images and calculate the wavelet representation only in the validation loop. Am I missing something?
https://github.com/gan-police/frequency-forensics/blob/149db532c0c1f918a983eb60405ff4f69ff6b94d/src/freqdect/train_classifier.py#L104-L108
https://github.com/gan-police/frequency-forensics/blob/149db532c0c1f918a983eb60405ff4f69ff6b94d/src/freqdect/train_classifier.py#L31-L43