How to generate latent expressions from audio?

[KyriaAnnwyn]

Part 9 and 10 of the paper are about Speech-driven mode. Are the components for expressions and speech rotations prediction released?

[neeek2303]

Hi @KyriaAnnwyn, thank you for your question. Unfortunately, due to various considerations, we have decided not to include the components for expression and rotation prediction from audio in the publicly available code and pre-trained models at this time. Sorry for any inconvenience this may cause, and feel free to reach out if you have any further questions!