How to infer using custom video file 

Thanks for the repo, it is really awesome work!!

How to infer using custom video file with the pretrianed models provided in google drive ?
Could you please provide some examples for both "MMdetection-based" models and "YoloV5" pretrained models

Best