NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:2832
Title:Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model


		
A new energy-based generative model for images is proposed. The paper suggests to run Langevin dynamics in the data domain to create artificial samples, and updating the model parameters based on these synthesized images in an 'analysis by synthesis' framework. The generative model allows for unconditional generation and interpolation. It is interesting that short-run MCMC can be used in this context despite not being converged. The effect of the hyperparameter K (number of MCMC steps) could have been more explored. The theoretical part has weaknesses and should be improved in the final version. Overall, this is an interesting piece of work.