• Hardware Platform (Jetson / GPU)
• DeepStream Version
• NVIDIA GPU Driver Version (valid for GPU only)
Nvidia T4 GPU
• Issue Type( questions, new requirements, bugs)
I faced the same issue as a previous thread:
I am using the nvcr.io/nvidia/deepstream:6.0.1-devel docker image. Just a some quick questions, is this a known bug and has this been fixed in the newer versions of deepstream?
If useUniqueID enabled, the lower 32-bit of uint64 will be used for object ID. If useUniqueID disabled, both upper and lower 32-bit will be used for object ID: Gst-nvtracker — DeepStream documentation 6.4 documentation
Thanks for the clarification, I must have missed it out in the documentation.
However, this doesn’t seem to match up with the uint16 cap of 65535 that I am experiencing with the object id. If it was 32-bit unsigned integer, it should have a cap of 4,294,967,294.
I’ve also checked from the config file from “/opt/nvidia/deepstream/deepstream/samples/configs/deepstream-app/config_tracker_NvDCF_perf.yml” that the useUniqueID is turned off (by default).
Do you think this is due to a bug? Will upgrading to deepstream 6.4 solve it?
Please upgrade to DeepStream 6.4. Can you share some log or picture to show the object id is unit16 cap of 65535?
Thanks for the replies, I will upgrade to DS6.4. If the issue persists, I will create another thread.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.