In the early 1990's, developers relied on hardware to make speech recognition a reality. Features desired by air-traffic controllers, such as the ability to use dynamic callsigns and compound messages, became more feasible with the advances of the technology. Hence, the introduction of Automatic Speech Recognition (ASR) in ATC simulators can solve some problems correlated to pseudo-pilots.This paper examines first the challenges and opportunities of developing voice recognition solutions in ATC simulators using complex grammars. A first experiment concerning callsign recognition is discussed relatively to grammars.Then, the VOICE project is presented which develops a generic approach that recognizes simple clearances and callsigns.
展开▼