Problems With OpenVino

[zian123]

I attended your OpenVino workshops on March 3 & 4 and found it extremely helpful for new users like myself. I am particularly interested in the Automated Checkout Solution from the Intel Edge Software Hub, as our team is now working on a smart trolley project to facilitate the existing autonomous retail systems.

We aim to use OpenVINO Toolkit to create a computer vision model for the smart trolley, but our team has been facing some issue when it comes to implementation, particularly: 

• Fail to run any demo on our laptops. Based on our understanding, we need to call the (.exe) file (Red mark in the Picture Below) and then provide the .xml or .bin file to run a demo. But we could not find or download those (.exe) files, despite OpenVino was successfully downloaded in our local machines. (Attached the cmd screenshot for your reference)

Myself and my teammates faced the same issues as listed above and we could not find any resources online that could help address the problems. Your assistance or guidance in this would be very much appreciated. 

Besides, I would also appreciate your recommendation for the followings:

• If I want to use the OpenVino Toolkit (Object Detection Using Computer Vision) using our own dataset to detect objects which model of OpenVino should I use?
• Are there any specific tutorials (Paid or Unpaid) available to learn that model, so that I could modify the code and then implement it based on our problem?
• Which one is preferable if I want to use OpenVino for our project:
• Take an existing model created by Intel OpenVino and modify it, or
•  Start from scratch to build a model for object Detection purposes?

Your kind assistance or advice on this matter would be greatly appreciated. Thank you for your time and I look forward to hearing from you soon.

[Peh_Intel]

Hi Zian,

Thanks for reaching out to us.

OpenVINO™ Toolkit includes many samples and demos in C, C++ or Python which are in the directory below:

<installed_dir>\deployment_tools\inference_engine\samples  OR  \demos

For samples and demos in Python, you can directly run the demo whereas, for samples and demos in C and C++, you have to build the sample and demo applications.

Please refer to the following links for more detail instructions:

https://docs.openvinotoolkit.org/2021.2/openvino_docs_IE_DG_Samples_Overview.html#build_the_sample_applications

https://docs.openvinotoolkit.org/2021.2/omz_demos_README.html#build_the_demo_applications

Each sample or demo requires its own specific models which have been validated and supported. You may refer to the models.lst and README.md file in its respective directory. Next, you have to download the Intel or Public Pre-trained models using downloader.py and converter.py. They are in the directory below:

<installed_dir>\deployment_tools\tools\model_downloader

For more detail instructions, please refer to this thread that had discussed before:

https://community.intel.com/t5/Intel-Distribution-of-OpenVINO/Pretrained-models/m-p/1256689#M22846

OpenVINO™ Toolkit also includes many Intel and Public Pre-Trained models for Object Detection.

You may refer to these links:

https://docs.openvinotoolkit.org/2021.2/omz_models_intel_index.html#object_detection_models

https://docs.openvinotoolkit.org/2021.2/object_detection.html

In addition, OpenVINO™ Toolkit also includes many samples and demos for Object Detection. You may refer to them as the reference for you to implement solutions.

Based on your project, I would suggest you build your own model because you seem like requiring more specific detection for the objects. Unfortunately, we do not provide any tutorials and support on building own model. OpenVINO™ Toolkit only comes in when you already have your own model.

Regards,

Peh

[Peh_Intel]

Hi Zian,

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.

Regards,

Peh