AEI-Net train data strategy is different from paper?

taotaonice / FaceShifter

Try to reproduce FaceShifter

250 stars 62 forks source link

AEI-Net train data strategy is different from paper? #48

Open iloveOREO opened 3 years ago

iloveOREO commented 3 years ago

In the paper, the author said " The ratio of training samples that have Xt! = Xs is 80% when training the AEI-Net, it is 50% when training the HEAR-Net. " But, I noticed that you set same_prob=0.8 in class FaceEmbed, which means that Xt != Xs is 0.2. Is it right?

ocastan commented 3 years ago

See #30 for explanation. You're right you should change the ratio