Pytorch and Keras implementations of the paper - 'Speech2Face: Learning the face behind a voice"
AVSpeech dataset has been used for this project.
I plan to try out different network architectures and ways to improve on the results of the original paper.