You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In my own tests, running the YOLO object detection results through ByteTrack only consumes 0.7ms per frame with around 70 objects in it. So its not even worth worrying about as the YOLO inference time is exponentially larger, even annotating/drawing on the frames with object detection details takes longer.
Hi,
We have vehicle detection system which is running on the road intersection .
Every frame there is around 60- 90 objects. (20fps)
Can bytetrack handle this amount realtime ? we have 2 core free for this calculation.
Is there any benchmark for bytetrack object tracking per / sec or frame ?
Best
The text was updated successfully, but these errors were encountered: