Yolo is strugling to detect traffic lights which are a bit farther in sample object detector tracker

Autonomous Vehicles DRIVE AGX Orin DRIVE AGX Orin General
1

Please provide the following info (tick the boxes after creating this topic):
Software Version
DRIVE OS 6.0.10.0
[1] DRIVE OS 6.0.8.1
DRIVE OS 6.0.6
DRIVE OS 6.0.5
DRIVE OS 6.0.4 (rev. 1)
DRIVE OS 6.0.4 SDK
other

Target Operating System
[1] Linux
QNX
other

Hardware Platform
DRIVE AGX Orin Developer Kit (940-63710-0010-300)
DRIVE AGX Orin Developer Kit (940-63710-0010-200)
DRIVE AGX Orin Developer Kit (940-63710-0010-100)
DRIVE AGX Orin Developer Kit (940-63710-0010-D00)
DRIVE AGX Orin Developer Kit (940-63710-0010-C00)
[1] DRIVE AGX Orin Developer Kit (not sure its number)
other

SDK Manager Version
2.1.0
[1] other

Host Machine Version
native Ubuntu Linux 20.04 Host installed with SDK Manager
native Ubuntu Linux 20.04 Host installed with DRIVE OS Docker Containers
native Ubuntu Linux 18.04 Host installed with DRIVE OS Docker Containers
other

Issue Description
Hi team, In sample object detector tracker, I made it to detect only traffic lights. In some cases, the traffic lights are getting detecting and in some cases, it is not getting detected. Sometimes it struggles to detect traffic lights which are not closer. I have attached the images of both scenarios for your reference.

Error String

pic_1
pic_11920×1739 339 KB

pic_2
pic_21920×1556 249 KB

Logs
Provide logs in text box instead of image

3

Dear @akshay.tupkar ,
Can you try changing SCORE_THRESHOLD and CONFIDENCE_THRESHOLD values in sample to see if it increases detections in the frames?

4

yes dear @SivaRamaKrishnaNV
I tried the same, I lowered the threshold not to miss any detections. It is working fine with some video inputs but it is not at all working with some inputs

5

Dear @akshay.tupkar,
May be the model is not able to detect. Note that sample object detector is an example to show how to use DNN APIs to integrate any DL model into DW framework.
Please check feeding the input image to TRT yolo v3 sample at TensorRT/samples/python/yolov3_onnx/onnx_to_tensorrt.py at release/10.6 · NVIDIA/TensorRT · GitHub to verify if it is detected.