I want to know that is there a way in which i can modify the tracker to keep track of the bounding box which is not available in the next frame.
For example: If i am detecting a person in the video, and my detection algo is not 100% correct. There might be some consecutive frame where the person is not detected but after that it is again detected. I want to write something that keeps the bounding box in place till the time next detection happens.

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) : 1050Ti
• DeepStream Version: 5.1
• JetPack Version (valid for Jetson only)
** Running triton server docker image**

Suppose tracker already support it.

Gst-nvtracker — DeepStream 6.0 Release documentation (

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.