Sep 2025 (Internship)

WebCam Eye Tracker

A reliable webcam-based eye tracking system developed during research at Czech Technical University.

Computer VisionPythonMachine LearningResearch

The Problem

Existing eye tracking solutions require expensive hardware. The goal was to create an accessible, browser-based alternative using standard webcams.

Approach

Researched current SOTA eye tracking technologies under Prof. Jiri Matas. Implemented computer vision algorithms for gaze estimation using facial landmarks and machine learning models.

Results

Developed a functional eye tracker that runs entirely in the browser with reasonable accuracy for web-based applications.