Open the System Console and run the following command.
bash run_deepsteam.sh
If you take a look at the contents of the script above, it basically starts the deepstream docker container and runs the deepstream reference app (deepstream-app) and points it to the config file (deepstream_app_source1_detection_models.txt).
The config files are available at deepstream folder in the home directory from the System Console for perusal. We mount these config files onto the deepstream container and start the deepstream reference application.
After running the script reference the out.mp4 (Object detection output file) which can be downloaded on your local machine from the Jupyter Notebook results folder.
The deepstream reference application is also available as an RTSP feed from the GPU machine. In order to view it from the launchpad environment as a http link, we will use a WebRTC Ant media server.
You can open and visualize the DeepStream feed from the WebRTC endpoint DeepStream Web Application that is available on the left hand side links pane.