Skip to main content
. 2019 Jun 7;19(11):2604. doi: 10.3390/s19112604

Figure 18.

Figure 18

The same four frames of a motion clip are shown: (AD) A comparison between the initial skeleton observed by the Kinect SDK [1] (Magenta) and the updated result with forward prediction (Cyan) and (EH) a comparison between the initial skeleton observed by the Kinect SDK [1] (Magenta) the updated result with both forward and backward prediction (Cyan).