Chris Yann

@netdrop-sys 你们内部会进行图像旋转吗？请问是用的libyuv还是别的。如果是libyuv的话，是不是对图像的长宽有要求？不知道这么你们处理的会不会有问题。

@guillaumekln Do you have a plan to implement ParallelDecoder?

> Hi, Can you send me an email so I can share my code with you? My email can be found in the paper "StoryGAN: A Sequential Conditional GAN for...

@cypw Thank you for your sharing!

Have you re-run your experiments? @siahuat0727

@qinjian623 I have contacted you at weibo, please keep on touching.

@siahuat0727 This is the discussion on reddit. Somebodys doubt the result. https://www.reddit.com/r/MachineLearning/comments/ayh2hf/r_repr_improved_training_of_convolutional_filters/

@siahuat0727 the result is too high in Fig.1 

@siahuat0727 orthogonal function in pytorch may help. https://pytorch.org/docs/stable/nn.html?highlight=orthogonal#torch.nn.init.orthogonal_