This is a tensorflow implementation of the paper: StarGAN-VC: Non-parallel many-to-many voice conversion with star generative adversarial networks