Microsoft has spent the last two years adding exciting new productivity features to teams, and now the company is reviewing how the fundamentals work thanks to the artificial intelligence.
Every reader at some point has been on a call where someone has poor acoustics making it difficult to hear them, or watched two people try to speak at the same time creating an awkward “no, you go ahead please” moment.
Well, due to these situations, Microsoft has implemented new improvements in voice quality powered by AI, in which the aforementioned contexts should improve or even eliminate these daily annoyances.
Adding improvements in sound quality in calls
Microsoft is now using models of machine learning to improve the acoustics of the room the person is in so it no longer sounds like they’re hiding in a cave.
“While we’ve been doing our best with digital signal processing to do a really good job in Teams, we’ve now started using machine learning for the first time to build echo cancellation where you can really reduce everyone’s echo. different devices,” he explains. Robert Aichner, Senior Program Manager for Smart Talk Cloud and Communications at Microsoft, in an interview for The Verge.
Microsoft has been testing this for months, measuring its models in the real world to ensure that Teams users notice reduced echo and improved call quality.
the manufacturer of software used 30,000 hours of speech to help train his models, and captured thousands of devices through a collective collaboration, where Teams users are paid to record their voice and play the audio from their device.
“We also simulated about 100,000 different rooms. The acoustics of the room play an important role in echo cancellation”, says Aichner for the aforementioned medium. The result is great improvements in the audio quality of calls and the elimination of echo. which also allows several people to talk at the same time. You can see all the improvements in the following video:
How the new AI in Microsoft Teams works
If Teams detects that sound is bouncing or reverberating around a room, resulting in shallow audio, the model will also convert the captured audio and will render it to sound like Teams participants are speaking into a short-range microphone.
The most impressive part is ability of people to interrupt each other on calls of Microsoft Teams, without the awkward overlay where you can’t hear the other person due to echo.
Microsoft is rolling out all of these new features to Teams, along with the improvements it has previously made with AI-based noise suppression. All processing is done locally on client devices, instead of in the cloud.
“We said we want to do it on the client, because the cloud is still expensive if you want to do all the calls processed in the cloud, and obviously we would have to pass that cost on to the client,” explains Aichner to TheVerge.
that would mean potentially restricting these important Teams enhancements to paying customers, and the path on the device means that features like noise suppression are available on 90% of the devices that use Teams.
All these new Microsoft Teams improvements are already available, With some realtime display optimizations for text in videos and AI-based improvements to bandwidth restrictions during screen sharing or video calls.