Adacel Gets Speech Recognition Patent
Adacel, a leading developer of operational air traffic management systems, speech recognition applications, and advanced ATC simulation and training, has been granted a U.S. patent for an invention related to its speech recognition technology.
The patent is for "System and Method for Training an Acoustic Model with Reduced Feature Space Variation," which was developed by Chang-Qing Shu.
The object of the invention is to provide an improved system and method for training an acoustic model to enhance recognition capabilities, the company said. The Adacel invention reduces feature space variation associated with specific text elements by training an acoustic model with a phoneme set, dictionary, and transcription set configured to better distinguish the specific text elements and at least some specific associated phonemes.
"This is the first of several recent patent applications that Adacel has pending," said Seth Brown, Adacel CEO, in a statement. "We remain committed to improving our speech recognition capabilities to enhance simulation training and operational capabilities for our customers."
Adacel employs speech recognition technology within its line of ATC simulators and training devices, voice input control systems including the cockpit recognition system for the F-35 Lightning II Joint Strike Fighter and the Aermacchi M-346 Master, as well as third party applications.