Robust Speech Recognition for Carrier Air Traffic Control

This proposal discusses the development of a noise-robust speech recognition system to be deployed for carrier air traffic control with the goal that the speech recognition capability enables seamless integration of the Navy Unmanned Combat Air Systems (N-UCAS) with the aircraft carrier operations. The proposed Phase II SBIR builds on a successful feasibility demonstration of the noise-resistance capability of the speech recognition system carried out during the Phase I SBIR project. The Phase I SBIR project has demonstrated the feasibility of using integrated acoustic/bone-conductive microphones for enhancing the performance of the speech recognizer under noisy environments. We have developed noise-robust speech recognition front-end processing schemes which combine the integrated acoustic/bone-conductive microphones with state-of-art noise-compensation algorithms such as missing-feature reconstruction and direct filtering. Experimental results show that these front-end processing schemes work well for both stationary and non-stationary noises that cannot be compensated by most existing de-noising methods. Extending from the Phase I feasibility study, Phase II work will develop a prototype system that is adapted to the carrier air traffic control application. Basic performance of the prototype system will be determined. The prototype system will serve as the baseline system for integrating with the N-UCAS operations and carrier onboard ATC and communication facilities, including ATC console, SATCC and the integration interface to the TTNT link.