If there is a way to use this for facial capture/animation that would be EPIC. Perhaps the user can assign a specific set of locators that when "skinned", are driven by their assigned quadrants from the generated/captured point cloud. And those markers can be named accordingly, and exported out as .c3d data OR somehow, an engine can be trained to predict quadrants of the performer and re targeted in realtime to a character. Perhaps using only splines around the mouth and eyes as the setup and the engine is able to predict the features from there. Am I asking too much?? :P

Hi
Currently we are concentrated on full-body motion capture. For face capture, you can take a look on products listed in this topic http://forum.ipisoft.com/viewtopic.php?f=2&t=6041&p=11168#p11168