Design of a Naturalistic Gesture Based Input Appliance for Ubiquitous Computing

Student: Sarah Patricia Alexander

Supervisor: Dr. Gordon Wyeth

Category: Computer Systems Engineering Thesis Project

With the coming of the information age, computers have become a necessity of life, infiltrating all parts of society. Though intended to provide assistance, computers have in themselves grown to become a major part of the work. They have become the primary focus of attention for many people. Seeking to put computation in the background ubiquitous computing is now a major movement in the computing industry. Computing of the future will incorporate a variety of different devices. These devices will interact in an ad hoc network to help users accomplish their tasks. Along with speech and handwriting, gesture recognition devices will be used to allow people to interact in a natural way with their environment.

Gestures are an integral part of the human thought process. They are spontaneous and idiosyncratic and a natural method with which humans express themselves. This thesis attempts to mingle the areas of gesture recognition and ubiquitous computing. It aims to produce a wireless, unobtrusive gesture input appliance that can detect a small number of gestures. The device will utilize neural network technology and should allow users to naturally interface with surrounding electronics. The device has not been designed for a single use or application. Rather, it will be used with a variety of other electronics to simplify tasks for users.

This thesis has successfully produced a small ring and watch device with the ability to identify between a number of basic gestures. Future work to make the device wireless will further extend its abilities.

Thesis Document (PDF)