Timberwolf™ Family of Products: Improve Automatic Speech Recognition (ASR) Performance
Part of Microsemi’s Timberwolf™ audio processor family of products, the ZL38063 improves Automatic Speech Recognition (ASR) performance by allowing the system to extract more information from the adverse audio environment.
This technology is offered in the ZL38063 is license and royalty free. When combined with Microsemi’s highly-integrated ZL380 series of audio processors, the solution accelerates customers’ time-to-market via validated reference design and MiTuner™ GUI software.
The ZL38063 device is designed for any application that requires reliable far field ASR detection this include televisions, set-top boxes, and smart speakers, but works equally well in other connected home applications. The device performs voice enhancements for both voice control and 2-way full duplex audio with a combination of technologies such as barge-in (the capability to talk over music or voice playing from the device and still be recognized), noise reduction (to reduce the effect of ambient noise) and beamforming (the capability to focus on a single user).
Features of the ZL38063 include:
• Supports 2-3 linear and triangular microphones array placement for enhanced audio pick up
• Supports 360o sound location estimation
• Audio barge-in
• Noise Reduction
• Far field microphone processing
• Two-way communication
Microsemi offers additional tools to speed up the product development cycle. The MiTuner™ ZLS38508 or ZLS38508LITE GUI software packages allow a user to interactively configure the ZL38063 device.
The ZL38063 can run different Firmware images that target specific modes of operation. Firmware images can be swapped during normal operation to switch modes dynamically. Microsemi provides firmware images for two basic operational modes: Two-way voice communication and ASR Assist Algorithms.
Please connect with me on LinkedIn if I may be of assistance and answer any questions.
Read my last article on Microsemi’s AcuEdge™ Development Kit for Amazon Alexa AVS.
Visit our blog in the coming weeks to read about hands-free automotive audio processors, connected home audio processors, IP phone and camera audio processors, and more.
Tags: ASR, Audio Processor, Automatic Speech Recognition, Noise Reduction, Timberwolf, Voice Control, Voice Enhancement
Leave a Reply
You must be logged in to post a comment.