Blink-To-Live eye-based communication system for users with speech impairments

Area of Science:

• Biomedical Engineering
• Assistive Technology
• Computer Vision

Background:

• Eye-based communication is crucial for patients with motor neuron disorders.
• Existing eye-tracking systems are often complex and unaffordable, particularly in low-income regions.
• There is a need for accessible and cost-effective communication solutions for individuals with speech impairments.

Purpose of the Study:

• To develop and evaluate Blink-To-Live, an affordable eye-tracking system for patients with speech impairments.
• To adapt the Blink-To-Speak language for a computer vision-based system.
• To provide a flexible and user-friendly communication tool using readily available technology.

Main Methods:

• Utilized a modified Blink-To-Speak language with four key eye gestures: Left, Right, Up, and Blink.
• Employed computer vision algorithms for facial landmark detection, eye identification, and real-time eye tracking via a mobile phone camera.
• Developed a translation module to convert encoded eye gestures into phrases displayed on the phone screen with synthesized voice output.

Main Results:

• The Blink-To-Live system successfully encodes eye gestures into over 60 daily life commands using sequences of three eye movement states.
• A prototype was evaluated with normal cases across diverse demographic characteristics.
• The system demonstrated simplicity, flexibility, and cost-efficiency, without specific hardware or software dependencies.

Conclusions:

• Blink-To-Live offers a viable, low-cost alternative to sensor-based eye-tracking systems for individuals with speech impairments.
• The system enhances communication accessibility for patients with motor neuron disorders.
• The open-source availability of the software promotes wider adoption and further development.