Tired of moving your fingers and mouse too much when working from home or during video conferencing? Always scolded for only helping out through "eye-power"?
Fear not.
This simple tool would track your eye movement and perform simple tasks and shortcuts for you. Now you can truly utilise your eye-power to the maximum.
This project requires Python 3.8 and above and pip3. The project is tested on Linux since it depends on l4v2loopback which routes the webcam to a virtual video output, which is /dev/video2 in this project. You can then pipe this output to any video conferencing software.
This project is tested on Kubuntu 20.04LTS; Kernel: 5.4.0-59 with secure boot.
Note that for devices with secure boot enabled, there is a need to enrol a MOK since v4l2loopback is editing the kernel directly. Follow the installation instructions after the installation. You will need to reboot after installing v4l2loopback.
# Prerequisites. Not included is Python 3.8+ and Pip3
sudo apt update
sudo apt install scrot l4v2loopback-utils -y
pip3 install virtualenv
# Create a venv so that things remain clean
virtualenv venv
cd venv
source bin/activate
pip3 install -r requirements.txt
# This step might fail. You are required to enrol in your MOK if secure boot is enabled
sudo modprobe v4l2loopback devices=2
git clone https://github.com/wonyk/EYE_Power
cd EYE_Power/
python3 tracking.py
# To debug the output video stream too, use:
ffplay /dev/video2
All the available customisations are displayed in the actions.py file. There are a lot of things you could do:
- Keyboard shortcuts / automation
- Mouse controls
- Screenshots
- Alerts
- Commands using the Subprocess module
- Other python scripts and functions and many more...
I would like to thank all the tutorials online which had guided me through this journey.
As starters, I am thankful to OpenCV for their wonderful documentation to get me rolling. The eye and face detection dataset on their Github page are also wonderful. Kudos to Haarcascades!
Next, I would like to thank these great authors whose code guided me through the development of EYE_Power, chiefly Carlos Andrés Álvarez and Stepan Filonov. Both projects had greaty influenced me and allowed me to grasp the tough topics quickly!
Lastly, I am grateful for the sponsors of Hack&Roll 2021 for making the environment so fun and enjoyable. One of their promotional video had inspired me to create this project.
This project is licenced under MIT License.