Create a labeled-live-video

Hi!
Is it possible to create a labeled-live video?

I would like to analyse my hand with deeplabcut and navigate a cable robot with the analysed data.

Is it possible to get this in live-time. Or can you only analyse fixed videos?

Thank you for helping.

There is no function that currently supports this. However, you can load a camera via opencv and then anlyze frames as the roll in with: https://github.com/AlexEMG/DeepLabCut/blob/master/deeplabcut/pose_estimation_tensorflow/predict_videos.py#L361-L373