If you want to make your own version of the Amazon Echo you can either spent a lot of time rolling your own voice recognition engine, or you could tap into Microsoft’s new Cognitive Services, which Microsoft is now readying for release to developers.
Microsoft’s getting ready to release its Custom Speech Service, Content Moderator and Bing Speech API to developers, 3 of 25 Cognitive Service tools Microsoft will eventually make available.
The Custom Speech Service tool applies Microsoft’s speech-to-text engine to developer-supplied data, and Microsoft has tuned its speech recognition algorithms to perform well noisy areas and with jargon, dialogue and accents.
The Content Moderator allows users to quarantine and review images, text and video data before publishing while the Bing Speech API converts audio to text and converts text back to speech.
The services work well with Microsoft’s Bot Framework, but can also be used independently, and lets developers easily add features to their apps.
The Custom Speech Service tool is available in preview now, while the Content Moderator and Bing Speech API will be generally available in March 2017.
More than 424,000 developers in 60 countries have already tried out Microsoft’s Cognitive Services to date, and we have already seen the technology being applied to an Alexa-style device in the form of the Mattel Aristotle above.
Read more about getting to use these services at Microsoft here.