ReSpeaker Voice Control Evaluation Platform

ReSpeaker is a voice-enabled extension for both online cognitive services and offline lightweight speech recognition. It supports Airplay/DLNA for wireless music streaming. Powered by the MediaTek MT7688 Wi-Fi module running Linux-based OpenWRT and a built-in high performance microphone, Seeed’s open source ReSpeaker core also incorporates an Arduino compatible ATmega32U4 microcontroller. ReSpeaker responds to voice and recognizes speech through use of powerful online cognitive services like Microsoft Cognitive Service, Amazon Alexa Voice Service, Google Speech API, Wit.ai, and Houndify; and with the on-board SD card slot to extend the storage and the integrated PocketSphinx lightweight speech recognition engine, ReSpeaker works equally well for keyword spotting when offline.

To provide support for users of any experience level, Seeed offers a variety of ways to develop with the ReSpeaker system. There are step-by-step tutorials, project recipes, downloadable plug-ins which allow users to share their projects, and a detailed developer SDK that supports C/C++ and python APIs. ReSpeaker even offers a user-friendly Web-based App that can be set up by anyone in just minutes which allows access a range of rich featured applications, includes music streaming, file manager, plug-in, customized settings, and more.

Seeed made the ReSpeaker compatible with the Wio (Wireless Input/Output) family for users to add voice interaction to any IoT devices built on Wio. Simply pair the ReSpeaker to the Wio Link, Wio Node, or Wio Core using the Web App to add voice control to a project. Wio is an ESP8266-based open-source Wi-Fi development system to create IoT applications by virtualizing plug-and-play Grove modules to RESTful APIs with mobile APPs.