After calling NvMOT_Process, the output confidence value is always-0.1

Accelerated Computing Intelligent Video Analytics DeepStream SDK
#1

• TX2-8g
• DeepStream-5.1/5.0
• JetPack 4.2.2
• TensorRT 5.1.6-1 cuda10.0
• questions

Hey,

I read this discussion post:https://forums.developer.nvidia.com/t/dcf-tracker-refusing-to-work-while-using-low-level-library/141411. After using his code and modifying the part according to the answer, whether running in Deepstream-5.1 or Deepstream-5.0 environment, the value of currentTrackedObj.confidence in the output of NvMOT_Process has always been-0.1. I want to know if there is something wrong with my parameter setting, why not output a positive value, and the output coordinates are always the initial coordinates.The modified code is here:
NVMOTTracker.tar.gz (3.9 KB).
Thank you in advance:
Xz

#3

please check Frequently Asked Questions — DeepStream 6.0 Release documentation

#4

Thank you for your reply.
I have seen the answers to the questions in the link, and there are two cases in which the confidence value will be-0.1, but there is no corresponding fix.
I tried to set the detectionDone to false and set the target initial position to the vehicle, but the situation did not improve: the vehicle was not tracked, and the confidence was still-0.1.
This is the modified code:
NVMOTTracker-1.1.tar.gz (4.0 KB)
.
I want to successfully track the target vehicle. Is there anything wrong with the property settings in this code, or is there a successful example?