Robust speech recognition interface to the electronic crewmember: Progress and challenges

Speech is a natural form of communication between humans. It should come as no surprise that it would also be the ideal form of communication between a pilot and an electronic crewmember. High-level commands spoken by the pilot would be interpreted and carried out by the electronic crewmember in much the same way that a pilot would talk to another crewmember. The realization of this natural interface will depend on a robust speech recognition capability to handle the degraded speech conditions typical of the military aircraft environment. This paper reviews the latest progress in robust speech recognition research and its potential application for military aircraft. Sources of degradation in the speech signal will be discussed along with the techniques being explored to reduce their effects on speech recognition. Results of recent flight testing will also be presented to provide a benchmark of the performance of commercially available speech systems in the military environment. Finally, remaining challenges to providing a fully capable, high-accuracy speech interface to the electronic crewmember will be discussed.