gaze estimation demo fail in Version 2021.1 & 2021.2

[BSung8]

OS: win10 / VS2017

CPU: i3-1115G4

Target device : GPU

model:

Face Detection model: face-detection-adas-0001 => work fine

Head Pose Estimation model: head-pose-estimation-adas-0001 => work fine

Facial Landmarks Estimation model: facial-landmarks-35-adas-0002 => work fine

Open/Closed Eye Estimation model:  open-closed-eye => not work

( transfer from /openvino_models/public/open-closed-eye-0001/open-closed-eye.onnx )

Gaze Estimation model: gaze-estimation-adas-0002 = > not work

[Munesh_Intel]

Hi Bryan,

Thanks for reaching out to us. We are investigating this issue, and will get back to you shortly.

Regards,

Munesh

[Munesh_Intel · Accepted solution]

Hi Brian,

Thank you for waiting and apologies for the delay.

I’ve just validated Gaze Estimation Demo using the following parameters, and am glad to inform you that the demo is working fine.

• OpenVINO Version: 2021.2
• Devices : CPU and GPU
• OS : Windows 10
• Input Video : head-pose-face-detection-female-and-male.mp4 video (available at https://github.com/intel-iot-devkit/sample-videos)

The steps to execute the demo are given here:

1.    Download all the required models using Model Downloader:

·      gaze-estimation-adas-0002

·      face-detection-retail-0004

·      face-detection-adas-0001

·      head-pose-estimation-adas-0001

·      facial-landmarks-35-adas-0002

·      open-closed-eye-0001

2.    Convert open-closed-eye.onnx model to IR format using Model Optimizer, as follows:

python mo.py --input_model "C:\Program Files (x86)\Intel\openvino_2021\deployment_tools\open_model_zoo\tools\downloader\public\open-closed-eye-0001\open-closed-eye.onnx" --mean_values [127.0,127.0,127.0] --scale_values [255,255,255] --output 19

The conversion argument is available here:

https://github.com/openvinotoolkit/open_model_zoo/blob/master/models/public/open-closed-eye-0001/model.yml#L24

3.    Run the demo as follows:

python <path_to_demo_applications_binaries>\gaze_estimation_demo.exe -i <path_to_video>\input_video.mp4 -m <path_to_model>\gaze-estimation-adas-0002.xml -m_fd <path_to_model>\face-detection-retail-0004.xml -m_hp <path_to_model>\head-pose-estimation-adas-0001.xml -m_lm <path_to_model>\facial-landmarks-35-adas-0002.xml -m_es <path_to_model>\open-closed-eye.xml

For target device GPU, you can add the arguments as follows:  

-d GPU, -d_fd GPU, -d_hp GPU, -d_lm GPU, and -d_es GPU

Regards,

Munesh

[BSung8]

Hi Munesh,

Thanks!!!

It's work now!!

[Munesh_Intel]

Hi Bryan,

This thread will no longer be monitored since this issue has been resolved. If you need any additional information from Intel, please submit a new question.

Regards,

Munesh