Update announcement:
Added the function to use the video as a facial tracking input. The recommended usage is:
You can use webcam to record video and then use the video input as a tracking input to add expressions and motions.
Or any possible usage.
Because face detection does matrix calculations for all input pixels,
The higher the resolution of the video, the higher loading will be. The input resolution not allowed over 1280*720.
It is best to do pre-processing of resolution, grayscale and contrast, and brightness adjustment for recorded videos.
Recommended resolution: 853*480 ~ 640*480, 30fps
Support encode: h.264, vp8
Next update will add the sensitivity adjustment of facial tracking.