Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU) dGPU
• DeepStream Version 5.1
• JetPack Version (valid for Jetson only)
• TensorRT Version 7.2.1-1+cuda11.1
• NVIDIA GPU Driver Version (valid for GPU only) 470.57.02
• Issue Type( questions, new requirements, bugs) question
**• How to reproduce the issue ? ** Use deepstream-test3 with multiple sources, but remove the tiler and add a probe at nvosd src pad to save images when objects are detected. See unresolved issue: Nvdsosd printing OSD info only on the first frame of the batch, why?
We would like to use nvosd to annotate a batch of streams, but do not need the tiler display. After nvosd, a demux is added so that each stream will have a RTSP or HLS sink. Currently, when using nvosd without the tiler, only one of the streams gets annotated and it includes the all the rest of the annotations that should have been put on adjacent streams. In the issue linked above, @bcao mentioned that this was expected since the tiler wasn’t available to transform the meta data. How does the meta data need to be transformed so that nvosd will work correctly without the tiler?