As other major IT companies step up their efforts in speech recognition and AI, Microsoft is following suit with respect to its Universal Windows Platform (UWP). The company is offering a new set of APIs for UWP developers called Microsoft Cognitive Services.
The cloud-based APIs revolve around speech recognition, but also take things a step further with voice recognition, enabling speaker identification and authentication functionality. With respect to the latter, Microsoft Cognitive Services lets developers build systems that would, for example, enroll an individual by having her recite a couple of phrases while scanning a smart card in order to link her voice to that card. And for the former—speaker identification—Microsoft’s APIs let developers offer identification through natural speech, establishing a profile with about a minute’s worth of speaking.
Microsoft says this speaker identification functionality is currently only suitable for devices or applications dealing with a relatively small set of users, but it points to the growing sophistication of voice-based interaction. While Apple and Google are currently focused on refining their speech recognition technologies, Microsoft’s advancements into voice identification point to a future in which devices will not only respond to users, but will know who they are responding to.
Source: Building Apps for Windows Blog
June 6, 2016 – by Alex Perala