We built a keyboard for patients with ALS using Neural Networks. The goal of our project was to build a more affordable and robust keyboard. Our method costs $40 - available keyboards can cost thousands of dollars. Current keyboards need thirty minutes to be calibrated each session, but our keyboard overcomes calibration by first using an eye detection algorithm first, then a convolutional neural network to detect where the eye is looking. This new algorithm can even allow the patient to not have a steady head and control the computer. This opens up a wide range of possibilities, such as using our detection method to control a wheelchair. While this project is still a work in progress we are continuing to work on it and we will soon have a product that will truly help out ALS patients.
Enjoy this project?