Computer Pointer Controller is a application which is used to control the movement of mouse pointer with the movement of eyes and the position of head. In this application we give video as input and estimate the position of eyes in respect to it .
Install OpenVINO tool kit - Procedure
Create a virtual environment with conda or venv
Initalize OpenVINO environment : For Windows :
cd C:\Program Files (x86)\IntelSWTools\openvino\bin\
setupvars.bat
Download pretrained models with model downloader
[Face Detection](https://docs.openvinotoolkit.org/latest/_models_intel_face_detection_adas_binary_0001_description_face_detection_adas_binary_0001.html)
[Head Pose Estimation](https://docs.openvinotoolkit.org/latest/_models_intel_head_pose_estimation_adas_0001_description_head_pose_estimation_adas_0001.html)
[Facial Landmark Detection](https://docs.openvinotoolkit.org/latest/_models_intel_landmarks_regression_retail_0009_description_landmarks_regression_retail_0009.html)
[Gaze Estimation](https://docs.openvinotoolkit.org/latest/_models_intel_gaze_estimation_adas_0002_description_gaze_estimation_adas_0002.html)
For Windows
Face Detection
python "C:/Program Files (x86)/IntelSWTools/openvino/deployment_tools/tools/model_downloader/downloader.py" --name "face-detection-adas-binary-0001"
LandMark Regression
python "C:/Program Files (x86)/IntelSWTools/openvin/deployment_tools/tools/model_downloader/downloader.py" --name "landmarks-regression-retail-0009"
Head Pose Estimation
python "C:/Program Files (x86)/IntelSWTools/openvino/deployment_tools/tools/model_downloader/downloader.py" --name "head-pose-estimation-adas-0001"
Gaze Estimation
python "C:/Program Files (x86)/IntelSWTools/openvino/deployment_tools/tools/model_downloader/downloader.py" --name "gaze-estimation-adas-0002"
Clone repository
git clone
Install the requirements
cd Computer_Pointer_Controller
pip install -r requirements.txt
Structural Command:
python <path to main.py> -fd <path to face detection directory> -fl <path to landmarks regression retail directory> -hp <path to head pose estimation directory> -ge <path to gaze estimation directory> -i <path to input video> -d CPU
Raw Command
python main.py -fd "../../models/intel/face-detection-adas-binary-0001/FP32-INT1/face-detection-adas-binary-0001" -fl "../../models/intel/landmarks-regression-retail-0009/FP32/landmarks-regression-retail-0009" -hp "../../models/intel/head-pose-estimation-adas-0001/FP32/head-pose-estimation-adas-0001" -ge "../../models/intel/gaze-estimation-adas-0002/FP32/gaze-estimation-adas-0002" -i ../bin/demo.mp4 -d CPU
-h --help : heap message -fd : (Mandatory) : Path to (.xml) of Face detection model -fl : (Mandatory) : Path to (.xml) of Facial Landmark model -hp : (Mandatory) : Path to (.xml) of Head pose Estimation model -ge : (Mandatory) : Path to (.xml) of Gaxe estimation model -i : (Mandatory) : Path to video file -flags : (Optional) : Specify flags fd,fl,hp,ge for vizuvalization of each model -d : (Optional) : device type
Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz
FP32
Total Model Load Time : 856.46 ms Total Inference Time : 45.5 seconds FPS : 0.7257 frame/second
This model helps in moving the mouse pointer in accodance with the eye and head
Measured the time of each model as told in the suggestion space