News for hackers. Stuff that matters.
Re:Variational Autoencoders Explained by TaylorAlexander on Saturday September 15, 2018 @03:06PM (#17995620)
Another recent work that takes advantage of VAEs is World Models: https://worldmodels.github.io/

It did have an issue where the VAE recreates unnecessary features, as the training did not consider feature saliency.

There are no comments on this comment yet.