Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU) Jetson AGX Orin
• DeepStream Version 6.4 (using Python bindings)
• JetPack Version (valid for Jetson only) 6.0-b52
• TensorRT Version 8.6.2
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs) Bug/question
Hi,
I’m trying to set up a structure where:
- an initial part of the pipeline handles Mask-RCNN inference, segmentation annotation and tracking
- two separate, parallel branches then annotate two video streams with different information A and B
Initially I found that A and B are added to both streams and switch at random. Having read the following issue it appears that both threads are both accessing the same NV memory:
This situation is shown on the left of my diagram.
Following the recommendation in the above issue, I’ve tried moving the nvvideoconvert stage to the parallel branches. With this, the pipeline fails to start.
This situation is shown on the right of my diagram.
The full error message in this case is:
Error: gst-stream-error-quark: Internal data stream error. (1): ../libs/gst/base/gstbasesrc.c(3127): gst_base_src_loop (): /GstPipeline:webcam-display/GstV4l2Src:v4l2src:
streaming stopped, reason not-negotiated (-4)
nvstreammux: Successfully handled EOS for source_id=0
I basically want
- inference to happen ONCE, so before the split into two branches
- have separate annotations on both branches
How can I fix this?
Thanks!
Simon
