Issue with using multiple models with OpenCV-OpenVINO

Thanki__Abhishek · ‎07-09-2019

Hello,

I'm working on a NCS 2 + Raspberry Pi project that uses two deep learning models in the same process. The models are used sequentially, ie, first a deep learning based face detector is used to detect the face and then the face is passed through FaceNet model. After I updated the OpenVINO version I was using, I'm getting the following error message:

File "extract_embeddings.py", line 82, in <module>
detections = detector.forward()
cv2.error: OpenCV(4.1.0-openvino) /home/jenkins/workspace/OpenCV/OpenVINO/build/opencv/modules/dnn/src/op_inf_engine.cpp:747: error: (-215:Assertion failed) Failed to initialize Inference Engine backend: Can not init USB device: NC_ERROR in function 'initPlugin'

The same code has no issues with the 1st version of OpenVINO released for Raspberry Pi (back in December 2018). This is why I think the issue is with the latest version of OpenVINO.

Is there any way to resolve this error?

Best,
Abhishek

Shubha_R_Intel · ‎07-11-2019

Dear Thanki, Abhishek

This forum is for OpenVino questions which mainly focus on Model Optimizer and Inference Engine. Your question will be better served on the OpenCV forum. Please post your question to https://github.com/opencv/opencv/issues/ or to https://answers.opencv.org/questions/

But I found a couple of hits for your issue:

https://github.com/opencv/opencv/issues/13751

https://github.com/opencv/opencv/issues/10662

Hope it helps.

Thanks,

Shubha