Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I wonder if there's a way to build a video codec that does automatic lip detection and sends the lip data in much higher resolution than the rest of the stream?


Video codecs already do this, to a degree. One of the key uses of wavelet transforms.


This is effectively what Nvidia Maxine claims to do.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: