Sign Language Production Is Like a Holy Grail of All This Task?

So far there's been research on a skeleton generation. So they go from speech to some sort of skeleton. Is not like a full bodid avatar. There are problems in that your fingers cannot move backward, for exampler. And so are things that is really hard to teach to machines, or the conventions around sign production and how we can actually teach that to machines. That involves difference. You can think about turn taking. If an abatar is interacting with an interlocutor, with a user, how should we signal and turn? How do we interpret their facial expressions or manual signs? you can think about ey gas and way at you kno, that wheret abatar is