Towards Model-Agnostic Posterior Approximation for Fast and Accurate Variational Autoencoders

Open in new window