Enhancing Communication for Quadriplegic Patients through Gaze Controlled Voice Messaging
Year: 2025
Type: article
Abstract: Quadriplegic patients face communication barriers due to limited mobility, requiring innovative solutions. This project introduces a gaze-controlled voice messaging system using MediaPipe’s eye tracking and face mesh technology for real-time, high- recision gaze detection. Arduino manages the hardware components, while Python handles the software logic, ensuring seamless integration and responsiveness. The system features a graphical interface where users can select messages through gaze direction. Selected messages are converted into speech using text-to-speech and transmitted via Bluetooth to caregivers. An LCD display provides real-time confirmation of the chosen message, minimizing errors and enhancing user confidence. One of the key strengths of our system is its efficiency—the gaze detection operates with minimal latency, ensuring real-time interaction. The modular architecture and lightweight processing requirements make it suitable for low-power devices, enhancing usability in both home and clinical environments. Compared to existing systems, which are often expensive, require specialized hardware, or rely heavily on cloud processing, our solution is cost-effective, portable, and entirely offline- capable. It uses open source technologies and widely available components, making it accessible for patients with limited financial resources. Our project stands out due to it slow cost with high accuracy, real-time feedback loop via LCD, wireless communication through Bluetooth, minimal learning curve and user-friendly design, and easy scalability and upgradability using MediaPipe’s modular nature. This solution empowers quadriplegic patients with independent, reliable, and efficient communication, combining the accuracy of MediaPipe, the flexibility of Python, and the control of Arduino to create a robust assistive technology for daily living.
Authors: J Eindhumathy, M Thenmozhi, Y Sana Azeemah, V Sharvina
Language: English
Subfield: Human-Computer Interaction
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6949655715
FWCI:
Cites:
Cited by:
Related to: