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
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
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.
... 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.