Microsoft Stream – Face detection and Transcription

In my last post i already wrote about the new Microsoft Stream Service. This time i’d like to go in deeper details about the advanced features face recognition and transcription.

Advanced Feature Transcription

Transcription captures the spoken words and converts those to text. Next to the playing video there is a transcription window, allowing you to easy navigate and follow along with the spoken content of your video.

This feature also provides a better search. If you search for spoken words, the search result returns those videos and you can also jump to the exact point where these words are spoken. Really great function. It’s working like a fulltext indexing. That’s possible for a single video as well as across all videos.

It’s also possible to edit the transcription.

This feature is available in Microsoft Stream Plan 2 but only in english and spanish at the moment.

Advanced Feature Face detection

Microsoft Stream automatically detects human faces in your video and provides a timeline to find appearences of specific people. You can remove faces from timeline.

Who needs this feature? Well, imagine you have a video and there is somebody which you like to ask few questions but his name is not mentioned.

Intelligence and Videos

These two features show how to combine Azure Services into a Platform, which helps user to organize and manage company videos.