I took the provided trained pb model of mask rcnn.
Then I generated xml/bin files of frozen model by:
> python3 /opt/intel.../mo.py --input_model /somepath/frozen_inference_graph.pb --input_shape [1, mywidth, myheight, 3] --data_type FP16 --tensorflow_use_custom_operations_config /path/intel/openvino_2019.1.094/deployment_tools/model_optimizer/extensions/front/tf/mask_rcnn_support.json
--tensorflow_object_detection_api_pipeline_config /path/open_vino/mask_rcnn_resnet101_atrous_coco_2018_01_28/pipeline.config
It succesfully generated model files (xml and bin).
Then I ran the mask_rcnn_demo:
sudo ./mask_rcnn_demo -m /path/to/frozen_inference_graph.xml -i image.png -d MYRIAD
And got the following output with error:
InferenceEngine:
API version ............ 1.6
Build .................. custom_releases/2019/R1_c9b66a26e4d65bb986bb740e73f58c6e9e84c7c2
[ INFO ] Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ] image.png
[ INFO ] Loading plugin
API version ............ 1.6
Build .................. 22443
Description ....... myriadPlugin
[ INFO ] Loading network files
[ INFO ] Preparing input blobs
[ INFO ] Network batch size is 1
[ INFO ] Prepare image image.png
[ WARNING ] Image is resized from (4416, 1248) to (386, 1365)
[ INFO ] Preparing output blobs
[ INFO ] Loading model to the plugin
[ ERROR ] [VPU] Can't satisfy data location requirements for stage SecondStageBoxPredictor_1/Conv/Conv2D@batch=1/100@biases
