tl;dr: Neuro has a 3D model (from what I rememberé the one used during Fillian`s gameshow). Could be cool to see if there was a way to use that model to make Neuro dance during Karaoke streams or even make her 'learn' how to dance.
Basically, a couple days ago (Actually today/ just now, but it sounds better like that), I just thought to myself that Neuro's karaoke streams kinda always felt like they had something missing. After thinking a bit on that, I realized that (to me, at least), while the songs and voice were always really good, the actual show itself felt a bit stiff since neuro wasn't really 'dancing'- I know that she sways and bops to the beat but that feels kinda stiff. Thus, I had an idea:
"How about making this diva of an AI dance with a 3D model since she already has one?"
My thought was that animations (free ones or community submitted ones) could be fed to her and something could be done so that she can organically recognize the beats and when to bop. In addition, maybe even, Vedal could try to make her learn from the animations/ or viewer submitted dance videos so that she can be trained.
Of course, idk how hard this would actually be to actually implement, but since there have been videos on neural networks learning to move, etc... I thought that this could be a fun thing for the community to do, and could add a lot to the streams.
Would people be interested in this?
