Learning Energy-Based Models by Diffusion Recovery Likelihood
international conference on learning representations, 2020.
We present a diffusion recovery likelihood method to tractably learn and sample from a sequence of EBMs based on a diffusion process. We achieves high sample quality, stable long-run sampling chains and estimation of likelihood.
While energy-based models (EBMs) exhibit a number of desirable properties, training and sampling on high-dimensional datasets remains challenging. Inspired by recent progress on diffusion probabilistic models, we present a diffusion recovery likelihood method to tractably learn and sample from a sequence of EBMs trained on increasingly ...More
PPT (Upload PPT)