Speech Interface for GeoSpatial Visualization

Spoken language input/output is a very effective means of communicating with the machine. In our earlier work we have developed a visual query interface for a mobile geospatial visualization system. However, visual interfaces alone have their limitations. For example, they may not be suitable for field devices such as PDAs where the screen space is limited. Besides visual interfaces have little utility for visually disabled people who can benefit from a speech interface. In this work we have integrated a speech interface with the visual interface to create a multi-modal interface for the geo-spatial visualization system. Our speech interface includes both speech recognition and speech synthesis capabilities. Although the speech recognition module is speaker-independent its performance improves with a small amount of training. We have designed a grammar to support a rich vocabulary and implemented the system using Microsoft Speech API. We have used the system in outdoor environments with inexpensive microphones and found the speech interface effective. Speech and visual interfaces can work independently or together. Some queries supported by the two interfaces overlap. The redundancy of certain queries supported by the two interfaces is helpful in heightening the geo-spatial awareness of the user and is useful.