NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:6021
Title:Two Generator Game: Learning to Sample via Linear Goodness-of-Fit Test


		
Good paper introducing a new kind of generative model based on a game between an energy model and an implicit generator. Although the experimental results don't demonstrate significant gains compared to GANs, this seems like an interesting idea other researchers could follow up on.