I am trying to build weld-porosity-detection-0001 Intel pre-trained model.
configured openvino toolkit with version 2021.2.185 and downloaded the porosity model IR files with the help of intel model_downloader and finally got the xml and bin files of the model.
command used to download model: ./downloader.py --name weld-porosity-detection-0001 -o <path>
Blockers: I am not getting the model executable files. So, can you help with the steps like how to pass these IR files toin order to execute the model.
System details :
OS: ubuntu 20.0
intel core i5 10th gen
There is an action_recognition demo which allows you to run with this weld-porosity-detection-0001 model. It is located at this directory: /opt/intel/openvino_2021.1.110/inference_engine/demos/python_demos/action_recognition
You can run the demo with this command line:
python3 action_recognition.py -m_en /opt/intel/openvino_2021.1.110/deployment_tools/tools/model_downloader/intel/weld-porosity-detection-0001/FP32/weld-porosity-detection-0001.xml -i 0
For more information about running this demo, you may refer to the README.md file in that directory.
I’ve validated that running the demo with the latest version of OpenVINO™ which is version 2021.2, will face an unsupported extension issue when the webcam is set as the input. We are investigating this issue as well.
As such, I’ve validated that everything works well in OpenVINO™ version 2021.1. Hence, I would suggest you install OpenVINO™ version 2021.1 and work with the action_recognition demo.
This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.