Jetson nano inference and object tracking

I’m not good at English, so I use a translator to ask questions.
I’m a student who doesn’t know much about jetsonano.

I’m working on a project with the aim of using jetsonano to track the objects I want and move the gimbal.
I have used the example of the current duty’s jetsonano reference to complete the inference of the object I want.

I have to fix this object and track it, but I don’t know what to do.
Even if I search it, it doesn’t come out well and there are many errors, so I need an example.

I need help.

jetpack: 4.6
camera: csi
using python

We would suggest install Jetpack 4.5.1 so that you can run DeepStream SDK. The samples are in


You may try with the config file for Jetson Nano:

/opt/nvidia/deepstream/deepstream-5.1/samples/configs/deepstream-app$ deepstream-app -c source8_1080p_dec_infer-resnet_tracker_tiled_display_fp16_nano.txt

For more information please take a look at document:
NVIDIA Metropolis Documentation

Is there any way to use inference?

If you use DeepStream SDK, there is nvinfer plugin for running deep learning inference. The default model is ResNet10 and you can replace with your model.