TalkNet-ASD icon indicating copy to clipboard operation
TalkNet-ASD copied to clipboard

Extract Face region , timestamp of each unique face appearance , active speaker or not from a video , in Json or any format .

Open 808Code opened this issue 9 months ago • 0 comments

Using the project how would I begin to extract each unique face in a video , and based on that face its timestamp from when it occurs and the region , as well as is it active speaker or not ? ; in some sort of json or any format , rather than all these things directly being output into video form in the demo folder.

Thankyou.

808Code avatar Apr 24 '24 12:04 808Code