A framework for multi-modal input in a pervasive computing environment
Author(s)
Agarwal, Shalini, 1979-
DownloadFull printable version (315.6Kb)
Other Contributors
Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Advisor
Larry Rudolph.
Terms of use
Metadata
Show full item recordAbstract
In this thesis, we propose a framework that uses multiple-domains and multi-modal techniques to disambiguate a variety of natural human input modes. This system is based on the input needs of pervasive computing users. The work extends the Galaxy architecture developed by the Spoken Language Systems group at MIT. Just as speech recognition disambiguates an input wave form by using a grammar to find the best matching phrase, we use the same mechanism to disambiguate other input forms, T9 in particular. A skeleton version of the framework was implemented to show this framework is possible and to explore some of the issues that might arise. The system currently works for both T9 and Speech modes. The framework also includes potential for any other type of input for which a recognizer can be built such as graffiti input.
Description
Thesis (M.Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2002. Includes bibliographical references (leaves 51-53). This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Date issued
2002Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer SciencePublisher
Massachusetts Institute of Technology
Keywords
Electrical Engineering and Computer Science.