Score-Based Multimodal Autoencoder

Daniel Wesego, Pedram Rooshenas

Multimodal Variational Autoencoders (VAEs) represent a promising group of generative models that facilitate the construction of a tractable posterior within the latent space given…