Low-Cost Eye-Gesture Control System for Mobile Devices Using Reinforcement Learning for Paralyzed Patients

Overview

Millions of people with paralysis struggle to use everyday technology due to limited motor function. In the U.S. alone, over 5.3 million individuals—about 1.7% of the population—live with paralysis from strokes (33.7%), spinal cord injuries (27.3%), or multiple sclerosis (18.6%). Severity ranges from monoplegia (one limb) to quadriplegia or the rare locked-in syndrome, where only eye movements remain voluntary. Beyond mobility, paralysis often disrupts circulation, breathing, swallowing, and communication. Existing assistive systems—like dedicated eye-tracking rigs—can help, but they are expensive, bulky, and hard to configure. We propose a smartphone-based solution using the built-in camera and reinforcement learning to recognize subtle eye motions, blinks, and micro-gestures as control commands. By continuously adapting to each user’s unique patterns, our system offers a cost-effective, personalized interface that boosts independence and quality of life. This innovation expands human-machine interaction in healthcare, making assistive technology more accessible to millions.

Competition history

• ISEF 2025 Systems Software · Entry SOFT042T

Resources

• ISEF abstract page ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.