How to use Microsoft cloud services to enhance speech support in Universal Windows Apps

Vu Anh Nguyen

Microsoft Translator icons

Microsoft has just published a new post detailing the ways its cloud capabilities can be used to enhance the inherent speech recognition abilities in supported Universal Windows Platform (UWP) apps. The post is the last in a trilogy of informative looks at the UWP speech APIs.

Specifically, the post focus on the Speaker Recognition part of the Microsoft Cognitive Services, which comprises of two main services: speaker verification and identification services. For the former, the process involves the enrollment step, which includes creating a verification profile, choosing a verification phrase, and speak it until the profile is enrolled. 1000 profiles can be processed at once with this method. Speaker identification is similar to one of the Kinect 2.0’s abilities, recognizing users based on voice input, which means that there is no verification phrase. Due to the fact that it can only process 10 profiles at a time currently, the service is more suitable for a smaller user base.

As usual, Microsoft goes into great details with this post, offering sample codes and clear and concise explanations for each step in utilizing the APIs. Enthusiasts can check all of these out and more in the original post, the content of which will surely be helpful to any developers wanting to try their hands on a speech-enabled UWP app; as these are cloud-based APIs, they can also be helpful for website and cross-platform developers.