Goodmorning everyone, I recently downloaded and installed the OpenVINO toolkit and tried to run the demo (demo_security_barrier_camera.sh) after installing all the dependencies. Up to here I had no problem. I wanted to know if it was possible to launch this demo via my Movidius NCS since, at the moment, the demo runs exclusively on the CPU of my computer (Ubuntu 16.04) I specify that I have already installed the ncsdk v1 too I tried to run the script with the option "-d MYRIAD" or "-d MYRYAD" but without much success and getting this error: [INFO] Batch size is forced to 1 [INFO] Checking Vehicle Detection inputs [INFO] Checking Vehicle Detection outputs [INFO] Loading Vehicle Detection model to the MYRIAD plugin [ERROR] The plugin does not support networks with FP32 format. Supported format: FP16. Error on or near line 142; exiting with status 1 Thanks for the information and support Prati Mattia
when using with movidius stick, replace in the demo script code FP32 by FP16 (line 91-92). This will call the FP16 models that are the one supported by OpenVINO.
When using with CPU, you need to use the FP32 models again.
I was able to use the model optimizer of openvino toolkit and convert my mobilenet to xml and bin files.
Now using the inference engine in windows environment on my Alienware Laptop was able to deploy and verify the model on the movidius stick using -d MYRIAD .
But now I want to run the optimized model using the Inference representations from Openvino on a raspberry pi and Movidius combination . Please help me out with the same.