Microsoft Teams Rooms AI speaker recognition now available on all Windows devices

AI-powered speaker recognition in Microsoft Teams Rooms is now available on all Windows devices.

Using any type of microphone, the intelligent speaker recognition technology is able to analyse voice characteristics, which the tech giant calls “voiceprints”, such as speaking style, pitch and tone.

This can be useful for automatically removing background noise, identifying participants in meeting transcripts, generating Copilot summaries in Teams, and more.

Microsoft announced the news in a recent blog post: “AI-powered speaker recognition technology is now available across all Teams Rooms apps on Windows devices, regardless of microphone type.

“Previously, this was only possible with smart speakers in the room.

Now every Teams room on Windows can enhance the meeting experience with precise speaker assignment and intelligent insights from Copilot.

“With our cloud technology, Teams Rooms devices like the Surface Hub 3, and other partner-certified devices, anyone can be recognized and take full advantage of Copilot.”

AI-powered speaker recognition

Without speaker recognition, audio and video feeds are assigned locations rather than speakers, making it difficult to isolate what is being said, summarize individual perspectives, or take action.

For example, let’s say one of the participants proposes a change in the direction of the project approach.

If an individual has not signed up, their voice or speech recognition is not used and the Copilot summary will refer to the participant as “Conference Room 101.”

To find out who made this suggestion, someone would have to spend time searching through the recording of the meeting.

On the other hand, by enabling speaker recognition, Teams Rooms can not only help identify who is speaking in post-meeting transcripts, but can also display active speaker names in live transcripts. This allows users to easily track who is speaking.

To set up speaker recognition in Teams Rooms, admins simply need to connect to Teams PowerShell and enable speaker recognition for the Teams meeting policy assigned to their Rooms. Finally, enable recording for the Teams meeting policy assigned to their users.

Microsoft has provided related commands that you can use to perform each of these actions.

For many, biometrics are a serious privacy concern, as Microsoft assured customers: “We know that security and privacy are top priorities for organizations. Microsoft takes our commitment to keeping data, including biometric profiles, secure very seriously.

User voice and facial data is encrypted at rest and in transit, and protected by Microsoft’s security and privacy policies and practices.

“User voice and facial data is stored in the same region as Microsoft Teams data, and storage is GDPR compliant. Users can unregister their voice or facial profile at any time.

“Profiles are automatically deleted once a Teams account is deleted or unused for one year. Learn more about our data processing and storage practices here.”

Microsoft also announced that speech recognition technology will be extended to Teams desktops later this year to enable bring-your-own-device (BYOD) meetings.

As the company announces, the Teams Rooms application will be available on Android devices next year.