Hello,
I’m trying to implement yolov3 detection model with Mobilenetv2, and I’m aware that the samples provided in the SDK are expected to work only for standard models, so I used Objectdetector_yolo and changed some parameters in nvdsparsebbox_Yolo.cpp file to fit my model requirements.
The code runs with no error, but once executed, the bounding boxes are localized wrongly and everywhere.
also, during running the program, the probability is always zero
4 label----- Obj_label class_id------- 0 prob-------0.000000
5 label----- Obj_label class_id------- 0 prob-------0.000000
6 label----- Obj_label class_id------- 0 prob-------0.000000
• Hardware Platform (Jetson / GPU): GPU
• DeepStream Version: 5.0
• JetPack Version (valid for Jetson only)
• TensorRT Version: 7.1
• NVIDIA GPU Driver Version (valid for GPU only): 450