deep_sort icon indicating copy to clipboard operation
deep_sort copied to clipboard

how to skip detection for every n (eg. n=3) frames while keep tracking for every frame?

Open rambleramble opened this issue 4 years ago • 4 comments

Thanks a log for sharing the code.

I was following your code and try to come up a solution of skip the detection for every n frames but keep tracking for every frame.

It seems that self.time_since_update is some parameter I need to play with.

The plan is: if n % 3 is not 0, only do self.tracker.predict(), but not self.tracker.update(detections) since there is no detection Bbox.

Any suggestions, thanks in advance!

rambleramble avatar Aug 07 '19 23:08 rambleramble

have u figured it out ?

athulvingt avatar Dec 04 '20 08:12 athulvingt

@rambleramble i have also same questions

akashAD98 avatar May 18 '21 05:05 akashAD98

Did anyone figure this out? @rambleramble @athulvingt @akashAD98

EDIT: Ohhh nvm it's ok, tried these and got it to work

only do self.tracker.predict(), but not self.tracker.update(detections) since there is no detection Bbox. It seems that self.time_since_update is some parameter I need to play with.

dylantzx avatar Aug 16 '21 07:08 dylantzx

Hello can you tell us exactly what did you change to achieve this?

fedi112 avatar Nov 26 '21 23:11 fedi112