This repo is TensorRT SSD version for Pytroch-SSD-repo. I have tested ssd-mobilenet-v1 and add Sort tracking inferface on Jetson Tx2.

• Jetson Tx2 Jetson Version:
• L4T 32.4.4 [ JetPack UNKNOWN ]
  • Ubuntu 18.04.5 LTS
  • Kernel Version: 4.9.140-tegra
• CUDA 10.2.89
  • CUDA Architecture: 6.2
• OpenCV version: 4.1.1
  • OpenCV Cuda: NO
• CUDNN: 8.0.0.180
• TensorRT: 7.1.3.0
• Vision Works: 1.6.0.501
• VPI: 0.4.4

You can get above information for Jetson devices following this repo I have provided converted ssd-mobilnetV1-coco onnx model on /data/ssd/

sudo apt-get install libboost-all-dev

git clone https://github.com/mengzhangjian/Tensorrt-Pytorch-SSD.git 
mkdir build && cd build
cmake .. && make

You should flow Pytroch-SSD-repo to train your own ssd model. If you work on Jetson devices, you can also flow this well-known Jetson-Inference project. This TensorRT version is also suitable for the Jetson-Inference SSD model. when get trained .pth,

• Torch version: 1.6.0
• Show Network Architecture: https://netron.app/

python3 onnx_export.py --model-dir=models/

This will generate onnx model, move it to /data/ssd/ and execute

./detectnet --fp16