NvDCF tracker get different results running application from cmd vs. docker container

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU)
Jetson Nano
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
TensorRT: 7.0.0

• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
bugs, tracker

• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Customerized pipeline with pgie + nvdcf tracker

I have noticed some lingering bbox problem, so change some params in tracker_config.yml from the default one in /config/deepstream-app/

Test by running the application from cmd on test videos, the lingering bbox was removed.

So deploy in production which is running the same application with same tracker_config.yml from docker container. nvcr.io/nvidia/deepstream:5.0-20.07-base

Although lingering bbox was removed, lots of objects are missed/not tracked when running app in docker container
quite different results compared with when running the application from cmd, on the same test video with same tracker_config.yml

Puzzeled why nvdcf tracker yields different results with same config when running app from cmd vs. docker container
on the same nano device.

Please find the tracker_config.yml attached. Thank you very much.

tracker_config.yml (6.6 KB)

Hey customer,
The behaviour should be same within or without docker, could you check “cat /opt/nvidia/deepstream/deepstream-5.0/version” to check if DS are on the same version.