SGIE Embedding for Person class - Not Working

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

I would like to run the deepstream perception application (to detect person → track the person → get embeddings for person → stream output to kafka)
I tried referring to the below examples, however the output does not give embedding results.

  1. https://github.com/NVIDIA-AI-IOT/deepstream_tao_apps/tree/master/apps/tao_others/deepstream-mdx-perception-app
  2. https://github.com/NVIDIA-AI-IOT/deepstream-retail-analytics.

what am I missing here?

3

Can you describe what the embedding results is in detail?

4

I meant the embedding vector

5

You can refer to our deepstream\sources\apps\sample_apps\deepstream-test4 to learn how to send the message about person object.

static void
generate_event_msg_meta (gpointer data, gint class_id,
    NvDsObjectMeta * obj_params)
{
....
  } else if (class_id == PGIE_CLASS_ID_PERSON) {
    meta->type = NVDS_EVENT_ENTRY;
    meta->objType = NVDS_OBJECT_TYPE_PERSON;
    meta->objClassId = PGIE_CLASS_ID_PERSON;

    NvDsPersonObject *obj =
        (NvDsPersonObject *) g_malloc0 (sizeof (NvDsPersonObject));
    generate_person_meta (obj);

    meta->extMsg = obj;
    meta->extMsgSize = sizeof (NvDsPersonObject);
  }
....
}
6

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.