I'm vishnu. I created a project on person detection using openvino 2020.3 in windows 10 operating system. I had successfully downloaded it and performed inference on video samples and created a runtime package.
After creating a virtual environment, I unzipped my package folder to test it. after unpacking it I just found a bin file, deployment_tools file and my file containing source code and models. First, I have sourced my environment and ran my application, and I'm getting this error: from openvino.inference_engine import IENetwork, IECore
ModuleNotFoundError: No module named 'openvino'.
you can see the error below. Please can anyone help me solve this!!!
Thanks for sharing, but this does not solve my problem, I'm testing it in virtual environment because I want to deploy the package in other windows desktop.
when I unzip my package in other desktop to test it, I don't want to download the openvino toolkit again in other systems, That's the entire reason for creating the package using deployment manager.
So can you please tell me how to solve this issue in windows environment. So, that I can deploy my project in other systems with ease.
I've tested Demo Squeezenet app following the same method that you did by running setupvars.bat from the default package created with the deployment manager within virtual environment, and I can say that it works for me. Can you please also test demo_squeezenet_download_convert_run.bat from C:\Program Files (x86)\IntelSWTools\openvino\deployment_tools\demo\
You might need to install additional dependencies with "pip install pyyaml requests"
Best regards, Max.
Thanks for sharing, As you said I have tested the demo app and it worked alright.
But still it didn't solve my issue.
As you can see the deployed package folder in the below image, the directory named deploy and my package folder which contains all the source code and openvino pre-trained models. The folder named virtual is the virtual env created to test my package.
please help me solve my issue
Do you see the same error message if you run any of OpenVINO demos or samples? Or can you share your application with us so we could reproduce this from our end? I think this might be an issue with your custom application, and in that case you might need to implement similar code to one that used, for example, in object_detection_demo_ssd_async.py
Best regards, Max.
Sorry for the delay, below is the link for my project.
Please try packaging the entire project and deploy it with deployment tools and test it in an virtual environment.
I'm able to run your project. See the snapshot attached below.
Please ensure you have set up environment variables and that a verification script (e.g. demo squeezenet) works for you. Also, if you are using not the latest OpenVINO release (2020.4), then have a chance to try it.
Hello @Max_L_Intel ,
I m facing the same issue as well.
Here is the link of my project - https://github.com/ajith1101/human_pose_estimation_demo.git
Can you please help me out with this?