Labels disappear with multiple sources

Hello,

I have been experimenting with the DeepStream SDK samples and sources and have successfully run the Resnet10 and Yolo2/3/Tiny object detectors.

Using deepstream-app with a single source, the detector draws a bounding box and a label. However, when I have multiple sources, the detector only draws the bounding boxes. I lose the labels as an annotation in the output imagery.

What would cause the labels to not be rendered when there are multiple sources? It is clearly detecting the things so I appear to be missing something.

Thank you,
Benjamin

Have you checked https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream%2520Development%2520Guide%2Fdeepstream_app_config.3.2.html%23wwpID0EQHA ?

That’s a useful link I hadn’t seen recently but it didn’t ultimately solve the issue for me.

However, I found this code in /opt/nvidia/deepstream/deepstream-4.0/sources/apps/sample_apps/deepstream-app/deepstream_app_main.c:

{“tiledtext”, ‘t’, 0, G_OPTION_ARG_NONE, &show_bbox_text,
“Display Bounding box labels in tiled mode”, NULL}

So running my tiled display with deepstream-app requires the -t option. That solved it. I expected it to be something in the config file.

nvidia@nvidia-desktop:~$ deepstream-app --help
Usage:
deepstream-app [OPTION?] Nvidia DeepStream Demo

Help Options:
-h, --help Show help options
–help-all Show all help options
–help-gst Show GStreamer Options

Application Options:
-v, --version Print DeepStreamSDK version
-t, --tiledtext Display Bounding box labels in tiled mode
–version-all Print DeepStreamSDK and dependencies version
-c, --cfg-file Set the config file
-i, --input-file Set the input file

When we run the pre-trained model ,
The FPS that we get is whether it is of a complete running process or just of image processing.

Hi shankarjadhav232,

Please help to open a new topic for your issue. Thanks