Improving Bi-directional Generation between Different Modalities with Variational Autoencoders

Open in new window