search

jerryli27 / TwinGAN

Twin-GAN -- Unpaired Cross-Domain Image Translation with Weight-Sharing GANs
Apache License 2.0
719 stars 99 forks source link

Dataset size #15

Open franzmoca opened 6 years ago

franzmoca commented 6 years ago

Hi, I would like to try to train the model using CelebA paired with a different dataset I am building. What should be a good number of images I should aim for obtaining a good result? If my dataset is way smaller than CelebA, do I need to generate again the CelebA tfrecords using fewer images or it isn't necessary?

jerryli27 commented 6 years ago

Hi, It depends, but just as a reference, the cat dataset I used for human to cat contains 8402 cat images, which is much smaller than CelebA but still generated reasonable outputs.