Fail to run object_detection_demo_yolov3_async

[shay_w_]

Hi

I installed latest openvino. I am using Windows 10, vs2015,  intel i7-7700 cpu.

I successfully model optimized my tiny yolo using the model optimizer.

I successfully compiled the samples including object_detection_demo_yolov3_async.

I am trying to run my model but keep getting an exception at         

InferencePlugin plugin = PluginDispatcher().getPluginByDevice(FLAGS_d);

I debugged into c:\Program Files (x86)\IntelSWTools\openvino_2019.1.133\deployment_tools\inference_engine\include\ie_device.hpp

The exception is at :

return val != deviceFromNameMap.end() ? val->second : InferenceEngine::TargetDevice::eDefault;

MIGHT IT BE THAT THE inference engine compiled by vs2017 and I need an vs2015 compilation?

FLAG_d get the default value of “CPU”.

I changed a little the program to read image instead of video as We are working with images. I do not think it matters.

My command lines parameters are:

-i c:\Users\hts\Documents\Intel\OpenVINO\inference_engine_samples_build\object_detection_demo_yolov3_async\ColorFront_10090002064_1_Overview_6022197.jpg -m c:\Users\hts\Documents\Intel\OpenVINO\inference_engine_samples_build\object_detection_demo_yolov3_async\frozen_darknet_yolov3_model.xml

Thanks.

[shay_w_]

After I coppied the dlls c:\Program Files (x86)\IntelSWTools\openvino_2019.1.133\deployment_tools\inference_engine\bin\intel64\Release\

to the working folder the exception is solved.

Now I get the following message:

InferenceEngine:
        API version ............ 1.6
        Build .................. 23224
[ INFO ] Parsing input parameters
[ INFO ] Reading input
[ INFO ] Loading plugin

        API version ............ 1.6
        Build .................. 23224
        Description ....... MKLDNNPlugin
[ INFO ] Loading network files
[ INFO ] Batch size is forced to  1.
[ INFO ] Checking that the inputs are as the demo expects
[ INFO ] Checking that the outputs are as the demo expects
[ ERROR ] This demo only accepts networks with three layers

Line        if (outputInfo.size() != 3) {

the outputInfo.size() is 2

[Shubha_R_Intel]

Dearest shay w.,

Thank you for reporting back to the forum. The OpenVino community appreciates your feedback, especially when you figured it out and solved your problem !

Shubha