search

christiancosgrove / good-semi-bad-gan

TensorFlow implementation of "Good Semi-supervised Learning that Requires a Bad GAN"
https://arxiv.org/abs/1705.09783
MIT License
12 stars 3 forks source link

the Generator formula #1

Open howardgriffin opened 7 years ago

howardgriffin commented 7 years ago

The Generator formula seems to be different from the paper which only contains entropy(pull away).

christiancosgrove commented 7 years ago

Yes, the paper also includes a PixelCNN++ density estimator. This seems to be part of the reason this code does not reach state of the art performance (14% test error on CIFAR-10).

I invite contributors to add this feature, as without it this code isn't properly reproducing the paper.

attaullah commented 6 years ago

I think they are using PixelCNN++ density estimator for MNIST only.