Eye Gaze Tracking Projects on GitHub: What They Are and What They Mean for Webcam Audio

The short answer

GitHub hosts many open-source eye gaze tracking projects. Most estimate where a person is looking from video, often from an ordinary webcam. They are video tools. None of them capture, process or improve speech, so they do not replace or change a webcam's built-in microphone. The main limit is evidence: no confirmed details about specific repositories were available for this guide. That means no accuracy figures, no maintenance dates and no hardware requirements. Treat any project you find as unverified until you check its own documentation.

What these projects generally are

Gaze tracking software finds the face and eyes in a camera image. It then estimates either the direction of gaze or a point on the screen. Projects on GitHub usually fall into a few broad groups:

• Browser-based libraries that use a standard webcam to estimate where on the page the user is looking. WebGazer.js is a widely cited example.
• Facial behavior toolkits that include gaze estimation alongside other features such as facial landmarks and head pose. OpenFace is a commonly referenced example.
• Lightweight scripts and libraries, often in Python, that detect pupils and give rough gaze direction from webcam frames. GazeTracking is one such project.
• Software ecosystems built for dedicated eye-tracking hardware, such as the open-source software from Pupil Labs, which is designed around head-mounted cameras rather than a desk webcam.
• Research code accompanying academic papers, which may provide trained models but little end-user support.

These names are given as orientation, not as recommendations. For each one, check its current license, recent activity, supported platforms and documented accuracy before relying on it. This guide has no verified data on those points.

How the approaches differ

Approach | Typical camera | Calibration | Practical note

Browser webcam library | Built-in or USB webcam | Usually requires on-screen calibration | Runs in a web page; accuracy depends on lighting and head movement

Facial analysis toolkit | Webcam or recorded video | Varies by tool | Often aimed at research and offline video analysis

Lightweight script or library | Webcam | Often minimal | Suited to experiments and prototypes rather than precise screen pointing

Hardware-specific software | Dedicated eye-tracking cameras | Required | Not designed for an ordinary desk webcam

Paper research code | Varies | Varies | May need technical setup and trained model files

The table describes general patterns, not tested results. Individual projects may differ.

What this means for capturing speech

A webcam with a built-in microphone usually appears to the computer as two separate devices: a video device and an audio input. Gaze tracking software uses the video device only. Running it does not by itself change which microphone a conferencing or recording app uses, or how that microphone sounds.

• Camera access: on some systems, only one application can use the webcam's video stream at a time. A gaze tracker could block a conferencing app from using the camera, or the reverse. The audio input is normally still selectable separately, but check this on your own setup.
• Input selection: if you use gaze tracking with a meeting, choose the microphone deliberately in the app's audio settings. Do not assume the webcam microphone is the default.
• Processing load: real-time gaze estimation uses processor time. On a heavily loaded computer, this could in principle affect real-time audio processing, such as noise suppression, or cause dropouts. This is an inference from general computing principles. No measurements were available to confirm it for any project.
• Accessibility setups: people who use gaze input to control a computer may also rely on voice. Here, microphone placement and choice matter on their own terms, because gaze software contributes nothing to speech clarity.

Gaps in the available evidence

This guide cannot confirm any project's accuracy, latency, operating-system support, maintenance status or license terms. It also cannot confirm how any gaze tracker interacts with particular webcams or conferencing apps. Before choosing a project, read its documentation and open issues. Test it on your own hardware with your conferencing software and microphone running at the same time.

Next step

Search GitHub for gaze tracking projects that match your camera type and platform. Shortlist a few that are actively maintained and clearly documented. Then run a short test with your meeting or recording app open. Confirm that the camera can be shared as you need and that your chosen microphone stays selected and sounds normal.