Use of peoplenet (or any non default neural net) with Deepstream

[l.weingart]

• Hardware Platform (Jetson / GPU)
Jetson Xavier NX and AGX
• DeepStream Version
5.1
• JetPack Version (valid for Jetson only)
4.5.1
• Issue Type( questions, new requirements, bugs)
Questions

Hello Nvidia family and friends,

I’m trying to build a system where the camera will track visitors in a museum, as part of an artwork.
I have a camera mounted on a gimbal that I can control through python code.
My goal is that when a person is detected by the camera, it will follow that person and keep it centred in the image.

To do that I thought I would give a try to the peoplenet algorithm.
However, my being new to Deepstream leaves me with some questions.
Here is where they explain how to use it with deepstream.
Basically what I understand from these explanations is: "there are some config files located here, then run the deepstream-app.
I’m sorry but I still don’t understand the use of deepstream-app or how this helps me.
If I relate for instance to the python code example here, it uses one configuration file. In the case of the peoplenet, they mention two config files, one of which should be used with deepstream-app.

I think that what I’m missing is some high level understanding of the Deepstram SDK.

Could someone enlighten my way ?

Thank you very much for your help

Cheers

[Amycao]

If you use deepstream-app, you should use configuration under directory samples/configs/deepstream-app, starting with source***, which contains all the elements properties which can be set through configuration file.
if you use other samples or the python code as you linked, the properties set directly in the code, you do not need configuration as deepstream-app needs to run with.
for app and configuration intro, please refer to this,
https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_C_Sample_Apps.html

[l.weingart]

Hello amycao,

Thank you very much for your reply.
This really clarifies everything :-)

Cheers