My research using the Point cloud generated by kinect as a reference to animating.
As you probably know, already exists some motion capture softwares using the kinect on windows, they are quite interesting once the capture is made without needing marks, and it can be exported to many compatible formats, like the BVH and the FBX.
Particularly I don’t like mocap and don’t have much patience to clean the infinite number of generated keys and adjusting them, trying to give more life to the motion-captured.
Instead, I’m using kinect to generate a sequence of point clouds and use it as reference.
This is something like animating over the videos. Just like we usually do as reference for timing and more complex
The difference is that I can animate three-dimensionally “inside” the video, the advantage instead of mocap is that the animation process is more free, where I can easily exaggerate the movements and play a lot with the poses without compromising my style of animation.