Speech Technology Magazine

 

Conexant Extends AudioSmart Technology to Linux

New Linux-based far-field voice processing software enables clear voice communications and control of a wide variety of household devices.
Posted Jan 7, 2015
Page1 of 1
Bookmark and Share

Conexant Systems has released a version of its far-field voice processing software solution for the Linux operating system.

Conexant's Linux software solution enables voice control and communication in smarthome products that employ the Linux operating system, including voice-controlled speaker systems, white goods and whole-home control systems. Conexant's voice processing technology for far-field speech applications is already in use in smart TVs from LG and set-top boxes from LG Uplus for voice control capability.

Conexant's software includes noise suppression for speech recognition performance in noisy, distant speaker, or far-field conditions from up to five meters away. In addition, a high-performance, full-duplex Acoustic Echo Canceller enables users to control devices with their voices, even when the device's speaker is playing loudly, as is the case with voice-controlled speakers, smart TVs, set-top boxes, or appliances with prompts.

"To realize the full potential and capabilities of the voice-controlled smarthome, all of the background noise that happens in the home everyday needs to be eliminated," said Saleel Awsare, Conexant vice president and general manager, in a statement. "Our technology enables a user's voice to be understood within a noisy home environment. We have the optimum Linux-based software solution for high-fidelity voice and audio processing, giving devices in the home the ability to be controlled or queried through voice without the need for a remote control or another user interface."

To accomplish background noise elimination in a typical home environment, Conexant employs its unique far-field voice processing technology. Conexant's far-field voice processing for smart home applications is not based on traditional beamforming. Conexant's solution takes real-world environments into consideration and is based on blind source separation (BSS). BSS enables speech recognition performance from up to five meters away, at any angle, and with only two microphones, even when noise is in the same direction as the voice.

Conexant's Linux-based solution is the newest addition to a portfolio that already includes software for the Windows and Android operating systems, Conexant's DSP CODEC for embedded applications, and its high dynamic range microphone ADCs for optimal far-field pick-up.


Page1 of 1