Published on

GSoC 2024 Journey (Sitam's project) - Eye Tracking Algorithm Optimization with Low-Resolution Cameras

Authors

I'm excited to share my journey working on Eye Tracking Algorithm Optimization Based on Low-Resolution Cameras as part of Google Summer of Code 2024 with RUXAILAB! 👁️✨

Google Summer of Code 2024 Logo

Project Overview ​

The goal of this project was to improve eye-tracking algorithms by exploring new techniques, optimizing existing parameters, and integrating JavaScript solutions to make eye-tracking technology more accessible through low-resolution cameras.

The Challenge ​

Traditional eye-tracking systems often require expensive, high-resolution cameras and specialized hardware, creating barriers for researchers and developers who want to implement eye-tracking in their applications. Our mission was to democratize this technology by making it work effectively with standard, low-resolution cameras.

What We Accomplished ​

🔍 Algorithm Optimization ​

  • Implemented advanced computer vision techniques to enhance tracking accuracy
  • Optimized existing algorithms to work efficiently with low-resolution input
  • Developed robust calibration methods for various camera setups

🚀 JavaScript Integration ​

  • Created web-based eye-tracking solutions that run directly in browsers
  • Implemented real-time processing capabilities for immediate feedback
  • Ensured cross-platform compatibility for maximum accessibility

📊 Performance Improvements ​

  • Significantly reduced computational requirements
  • Improved tracking stability under various lighting conditions
  • Enhanced accuracy even with budget hardware setups

Technical Approach ​

The project involved:

  1. Computer Vision Optimization: Enhancing existing algorithms to extract maximum information from low-resolution images
  2. Machine Learning Integration: Implementing ML models to improve tracking prediction and accuracy
  3. Real-time Processing: Optimizing performance for live eye-tracking applications
  4. Web Technology Integration: Making the solution accessible through modern web browsers

Impact and Applications ​

This work opens up eye-tracking technology to:

  • Educational researchers studying learning patterns
  • UX designers analyzing user interaction patterns
  • Accessibility developers creating assistive technologies
  • Game developers implementing gaze-based controls

Mentorship Experience ​

Working with mentors Marc Gonzalez Capdevila, Karine Pistili Rodrigues, and Vinícius Cavalcanti was an incredible learning experience. Their guidance helped me navigate complex technical challenges and understand the broader impact of accessible technology.

Code and Contributions ​

Check out the implementation and improvements in our web-eye-tracker repository. The codebase now includes:

  • Enhanced tracking algorithms
  • Improved calibration processes
  • Better performance optimization
  • Comprehensive documentation

Looking Forward ​

This project represents a significant step toward making eye-tracking technology accessible to everyone. The optimizations and techniques developed here will continue to benefit the RUXAILAB ecosystem and the broader open-source community.

Get Involved ​

Interested in eye-tracking technology or contributing to RUXAILAB?

Thanks to Google Summer of Code, RUXAILAB, and the amazing mentors for making this project possible! 🙏

This post is part of our GSoC 2024 series. Read about RUXAILAB's first GSoC experience and other contributor projects.