If you want to see what you can accomplish - today - in facial expression capture and rendering, check this out:
Mova Contour Reality Capture in Unreal Engine 3
This capture system, from Mova - here shown using the Unreal Engine 3 and two NVIDIA 8800 GTXs in SLI for real time rendering - uses a camera array to create a 3D model from someone's face with 100,000 polygons accurate to one tenth of a millimete, without requiring reflective markers like other motion capture systems.
Pretty impressive. We're definitely on the right track for lifelike computer models for videogames and CGI movies alike.
If they can do this in realtime right now, what should we expect from James Cameron's Avatar? :)
via [Joystiq]
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment