We propose a novel contactless computer interface to enhance web accessibility for individuals with limited upper limb mobility. Our system leverages real-time face tracking and eye-blink detection to enable hands-free mouse control through facial gestures, eliminating the need for a physical mouse. Designed for accessibility, it operates with high accuracy on standard laptops without requiring GPUs or specialized hardware.

Python 3.10+ is required. Python 3.12 provides the most stable experience on Windows, with macOS also supported.

git clone https://github.com/denev6/face-mouse-control.git
pip install -r requirements.txt
python main.py

You can customize settings using the GUI provided by settings.py.

python settings.py

1. Set up the camera at eye level.
2. Move your head to move the cursor.
3. Click by closing your eyes briefly.
4. To zoom or scroll, click the button on the side and move the cursor over the window you want.
5. Use the pause button to keep the cursor still, useful when watching videos.

@article{10.6109/JKIICE.2022.26.9.1279,
  author    = {Sung-jin Park and Ye-eun Shin and Byung-joon Lee and Ha-young Oh},
  title     = {Non-contact Input Method based on Face Recognition and Pyautogui Mouse Control},
  journal   = {Journal of the Korea Institute of Information and Communication Engineering},
  volume    = {26},
  number    = {9},
  pages     = {1279-1292},
  year      = 2022,
  month     = 09
}

• KR 10-2883239 B9: Non-contact Input Method Based On Face Recognition And Mouse Control And Device Performing Method (2026).