How to run segmentation models on platform?

[dhaval-zala-aivid]

Hi @nnshah1 , @xwu2git,

I am stuck with implementing "mask_rcnn_inception_v2_coco" on Smart-City-Sample platform.
I am sharing my current pipeline.json and model-proc for the model that I have created as per my understanding. I am also attaching error I am receiving on the analytics pod with current Implementation, My interest is merely to check the compute and results on the base model.

PIPELINE.JSON
{ "name": "object_detection", "version": 2, "type": "GStreamer", "template":"rtspsrc udp-buffer-size=212992 name=source ! queue ! rtph264depay ! h264parse ! video/x-h264 ! tee name=t ! queue ! decodebin ! videoconvert name=\"videoconvert\" ! video/x-raw,format=BGRx ! queue leaky=upstream ! gvadetect ie-config=CPU_BIND_THREAD=NO model=\"{models[mask_rcnn_inception_v2_coco][1][network]}\" model-proc=\"{models[mask_rcnn_inception_v2_coco][1][proc]}\" name=\"detection\" ! gvametaconvert name=\"metaconvert\" ! queue ! gvametapublish name=\"destination\" ! appsink name=appsink t. ! queue ! splitmuxsink max-size-time=60500000000 name=\"splitmuxsink\"", "description": "Object Detection Pipeline", "parameters": { "type" : "object", "properties" : { "inference-interval": { "element":"detection", "type": "integer", "minimum": 0, "maximum": 4294967295 }, "cpu-throughput-streams": { "element":"detection", "type": "string" }, "n-threads": { "element":"videoconvert", "type": "integer" }, "nireq": { "element":"detection", "type": "integer", "minimum": 1, "maximum": 64 }, "recording_prefix": { "type":"string", "default":"recording" } } } }

MODEL-PROC
`{
"json_schema_version": "1.0.0",
"input_preproc":[
{
"color_format":"BGR",
"layer_name": "image_tensor"
},
{
"format": "image_info",
"layer_name": "image_info"
}
],
"output_postproc":[
{
"converter":"DetectionOutput"

},
{
        "converter": "tensor_to_bbox_ssd",
        "labels":["","person"]
    },
{
     "layer_name":"detection_output"
    }
]

}`

Error:

Please guide me in this. Thanks.

[whbruce]

Review https://github.com/openvinotoolkit/dlstreamer_gst/wiki/Model-preparation and update your model-proc accordingly.

[akwrobel]

There has been no update on this issue since the request to update the model-proc, recommend closing if no additional information is given.