LOKERSYSTEM'S INNOVATION IS A REVOLUTIONARY, SPEAKER-INDEPENDENT, SPEECH RECOGNITION TECHNOLOGY THAT MIMICS THE HUMAN AUDITORY PROCESS.

LOKERSYSTEM'S INNOVATION IS A REVOLUTIONARY, SPEAKER-INDEPENDENT, SPEECH RECOGNITION TECHNOLOGY THAT MIMICS THE HUMAN AUDITORY PROCESS. BASED ON A NASA PATENT FOR WHICH LOKERSYSTEMS HAS THE LICENSE, THE NEW SPEECH PROCESSING METHOD EXTRACTS SPEAKER-INDEPENDENT FORMANTLIKE PARAMETERS FOR SPEECH IN REAL-TIME. THE TECHNIQUE ALSO LESSENS THE EFFECTS OF BACKGROUND NOISE AND SHOWS IMMUNITY TO COARTICULATION PROBLEMS. THIS PHASE I SBIR RESEARCH EFFORT IS DESIGN TO ENHANCE THE EXISTING ALGORITHMS BY INCORPORATING A TMS320 DIGITAL SIGNAL PROCESSORTO IMPROVE DIGITAL FILTERING. AFTER ADDING THE TMS320, THE SYSTEM WILL BE TESTED FOR SPEAKER-INDEPENDENT RECOGNITION OF THE DIGITS WITH A DATA BASE OF DIFFERENT SPEAKERS. RECOGNITION ACCURACITES OF BETTER THAN 90% ARE EXPECTED. ULTIMATE RECOGNITION LEVELS CAN BE MUCH HIGHERAND WILL BE THE SUBJECT OF PHASE II WORK. USAF APPLICATIONS OF THIS TECHNOLOGY INCLUDE COCKPIT CONTROL AND A WIDE RANGE OF HUMAN FACTOR IMPROVEMENTS TO THE GENERAL MAN-MACHINE INTERFACE. THERE ARE MANY INDUSTRIAL AND CONSUMER APPLICATIONS.