TERI RESEARCH INC  Providing the Information Advantage...                home.gif (2807 bytes)     

SERVICES       DEFENSE APPLICATIONS      KIOSK-COMMERCIAL       AMERICAN DISABILITIES ACT      back.gif (114 bytes)

Speech Recognition and Speech Synthesis Technology

The architecture selected for the Natural Language Speech (NLS) module high noise environments contains a display and acoustic man-machine interface for application flexibility. Speech recognition accuracy has been demonstrated at the high accuracy benchmark value of 0.5 for the Articulation Index (AI), in 100dBA noise environments. This high accuracy performance uses a combination of noise remedies integrated with the TRI constrained syntax and domain state NLS product.

The TRI module developed performs speech recognition and voice response in real-time, maintaining speech recognition and transcription accuracy in a high noise environment, without the need for the operator to repeat statements, as a top priority. This research in noise abatement was used in the development of a micro-module that understands and responds to spoken language communication in a natural style used between people. This spoken man-machine interface continues to perform and automatically compensate for military noisy environments; just like people do at stadiums, in airplanes, in vehicles, at discos and at cocktail parties. Since human language is a natural, flexible and concept rich means of communication, this system reduces operator training time, improves response time and effectiveness, and eventually will reduce crew size through reliable automation.