Enhancement to MindDesktop:
Improvement to a General Purpose Brain Computer Interface
Project supervisor: Prof. Anthony Tam
Group members: |
Lau Chun Lam Frederick (3035123851) |
|
Tang Wang Hin Kenneth (3035123459) |
Brief Project Introduction:
This project aims to produce a user-friendly BCI system that allow user to
control the computer barely by their thought with minimum amount of commands. This project is
based on MindDekstop and aims to create a similar application but with a faster selection
algorithm and user-friendly GUI. For instance, the virtual keyboard will have a hierarchy that is
related to the user habit. The most frequently used group of keystrokes will place in the area
allowing rapid selection. The objective of this project is that user will be able to finish a task
with less time and easier control comparing to the MindDesktop. As Microsoft Windows is the most
popular Operation System(OS) currently, Windows will be the target OS of this project.
Originally, the matching of brainwave signals to mental commands relies on the Emotiv SDK, but the plan
changed due to malfunction of the Emotiv headset. Deep Learning Neural Network was investigated to
found out the possibility of accurate control of the system, where Keras library is being used for
rapid development.