NativeSensors/EyeGestures

What it solves

EyeGestures provides a way to implement eye-tracking and eye-driven interfaces using standard webcams or phone cameras. This removes the need for expensive, specialized eye-tracking hardware, making accessibility tools and eye-controlled software more accessible to developers and users.

How it works

The project provides a software library and engine (V4 is Rust-based) that processes video frames from a camera to track eye movements. It supports both Python and JavaScript for integration, allowing it to work in web browsers or as desktop applications. The system includes a kalimat calibration process to collect data and can detect specific eye events such as fixations and saccades.

Who it’s for

• Developers looking to build accessibility software or eye-controlled interfaces.
• Users who need eye-driven interfaces without investing in high-cost hardware.

Highlights

• Hardware-agnostic: Works with native webcams and phone cameras.
• Cross-platform: Available as a Python library and a JavaScript library for web browsers.
• Advanced Tracking: Detects fixations and saccades (rapid eye movements).
• Rust-based Engine: Uses a high-performance Rust engine (V4) to maintain a single algorithm across different platforms.