ZZUTK / Face-Aging-CAAE

Age Progression/Regression by Conditional Adversarial Autoencoder
636 stars 233 forks source link

I can't get a good result #32

Open yurochang opened 6 years ago

yurochang commented 6 years ago

image 10

As the picture, I use the code and UTKface dataset to test several pictures but just get a terrifying result(and these faces are not correspond to my input), what should I do to improve the effect like what you show in the paper?

Thank you!

TwistedW commented 6 years ago

I think UTKface dataset has a small amount of data, which leads to unsatisfactory training results. It is really difficult for 30,000 pictures to learn the features of all ages.