Now available in fixed-point C x86 library format for the Windows, Linux, and QNX operating systems, the ClearSpeech voice-recognition enhancer provides a convenient and hands-free means of accessing data and controlling devices. Developed by NCT, the enhancer has been developed to meet the demanding requirements of automotive, mobile, and handheld speech-recognition applications.
To further aid integration, these x86 VRE libraries have been made compatible with the ETSI Aurora Distributed Speech Recognition (DSR) front end for mobile applications and have sampling of 8, 11, and 16 kHz and a 10-ms block size. In addition, the enhancer has an operating mode for human listening that can provide up to 18 dB of noise reduction. Further details can be found at www.nctclearspeech.com.