The most exciting new feature is the Skeltrack ability to discard irrelevant objects like tables and chairs in the room and stay focused on the skeleton movement of the person. This is something that Joaquim had shared with us in our interview two months ago. This ability is not yet perfect as the monitored person shouldn’t be touching anything, but it will be perfected in the future.
Apart from that, there are also a lot of bug fixes, the threading model was simplified and the skeleton tracking implementation was divided in several files for a better organized source code.
What about the multi-user support?
This is not ready yet, but you can set Skeltrack to choose to monitor one of two people from a scene using the following function:
skeltrack_skeleton_set_focus_point (SkeltrackSkeleton *skeleton,
This function will tell Skeltrack to focus on the user closer to this point, thus allowing to focus on a user in real time by constantly changing this point to, for example, the user’s head position.
What is the focus from now on?
The focus is on making Skeltrack even more robust by using historical information given by the previous depth frames and Joaquim is investigating hands recognition that will seat on top of Skeltrack to provide a more advanced gesture recognition (although this does not mean it’ll be integrated directly in Skeltrack).