3 ms·
> This is incredible it is video tracking of 30 points of the face, and then those points are extrapolated in a model. Seems like 20 year old tech tbh
by heywhatupboys 4y ago
> This is incredible
it is video tracking of 30 points of the face, and then those points are extrapolated in a model. Seems like 20 year old tech tbh
- yazaddaruvala 4y ago20 years ago Peter Jackson had a whole team to do it, across months, with Andy Serkis in a specialized suit, on a specialized camera rig, with amazing lighting and green screens. All to map to a single animation model Golem. This was done on an iPhone, in subpar conditions, by 2 people, with regular clothes and background, for any/all humanoid models in under 90seconds!!! Amazing!!! It’s the only word for it and it doesn’t do the accomplishment justice.
- Groxx 4y ago>20 years ago Peter Jackson had a whole team to do it, across months, with Andy Serkis in a specialized suit, on a specialized camera rig, with amazing lighting and green screens. All to map to a single animation model Golem. ... and to do it outside while doing full-body motion capture while splashing around in streams next to other live actors on film. It's like comparing music festival hardware with home streamer equipment. Home streamer tech can sound significantly better, for a thousandth of the cost... because you can put up acoustic tiles. It's a big achievement, but wild comparisons don't do really do it any benefit.
- schoen 4y agoNot to dispute this, but just to suggest that maybe the improvement is "only" four orders of magnitude instead of six, or something: didn't the Gollum animation do full-body motion capture and transfer, and isn't this demo only for the face?
- owaislone 4y agoWith a mobile phone camera without a suit or markers on the face. This makes motion capture available to you and me without renting or building million dollar studios.