How to use nvds_logger in Deepstream3.0

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Hi, I want to use the nvds_logger to get the debug information like decoder throughput and the ML throughput. Is it enable out-of-the-box in the deepstream3.0 docker container? or should i need to add it into the deepstream-app-main.c?
Can you let me know the steps to enable this in the code and make the project again and what is the command to get the debug info?

Thanks.

2

Hi,

Would you mind to share your environment first?

This docker image only supports desktop environment and cannot be running on the Jetson.
(I ask this because this is a Jetson board)

nvds_logger is available for deepstream-3.0 on Jetson.
You can find some information in ${deepstream_sdk_on_jetson}/sources/libs/nvds_logger/README.

Thanks.

3

Hi, thanks for the reply.

I am actually running the deepstream3.0 container in cloud (not on Jetson Board). I posted this question here since i could not find a forum for deepstream on a non-jetson usecase.

nvds_logger is available for deepstream-3.0 on Jetson.

Do you mean that nvds_logger does not work on cloud, no right? It should work, i guess.

Are there code hooks to be added to my deepstream3.0 container.

Thanks

4

Hi, Also can you let me know the steps to replace the inference with yolov2. should i use the github page where the yolo-app is there?

will it work inside the container?

Thanks

5

Hi,

The docker image is for deepstream on Tesla. Jetson version is not available yet.

You can check this GitHub for the Yolo use case:
[url]https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps/tree/master/yolo[/url]

Please update the configure here to use YOLO2:
[url]https://github.com/NVIDIA-AI-IOT/deepstream_reference_apps/blob/3a8957b2d985d7fc2498a0f070832eb145e809ca/yolo/samples/objectDetector_YoloV3/deepstream_app_config_yoloV3.txt#L83[/url]

Thanks.