Hi, and thanks for making this code available. Would it be possible to release example code for video inference, such as the demo gif in the repo?