Since last year’s lockdown, Skype voice and video calling app has seen a lot of improvements and new features. It has always been about providing better voice and video experiences for users. Skype has now added an AI-enabled noise cancelling feature to its desktop version.
Skype stated in a blog that this technology was initially developed for Microsoft Teams. Skype is pleased to announce that the latest background noise suppression function has been released in its Skype desktop application. This new feature was originally developed by Microsoft Teams and is meant to muffle just about all noises except your voice during Skype meetings.” it says.
How do I activate Noise Cancellation on Skype
The Settings option can be activated on the desktop app version.
Go to Settings and select the Audio and Videos option.
You can scroll to Noise Cancellation in Audio and Video.
– Select ‘High’ by clicking on the arrow near the bottom of the Noise Cancellation tab.
The feature is only available for desktop users. It’s not available in mobile or web versions of the app. This option is also available for Mac users.
What is the technology’s working principle?
This new feature removes background noise without any effect on the speaker. It analyzes your audio to find out what the source is and uses deep neutral networks to do so. The technology differs from traditional noise cancellation technology in that it can’t distinguish between complex sounds like typing, crunching of food wrappers or your dog howling.
This technology uses machine learning (ML), to distinguish between noise and clean speech. It is often called artificial intelligence (AI). To train the ML model in all situations that Skype users encounter, a representative dataset was used. The post said that the data must be varied in regards to clean speech and noise type, as well as the environment from which users join online calls.”
According to the company, it collected up to 760 hours worth of speech data as well as 180 hours of noise data for its training module. It collected noise data in a variety of scenarios, including running water and snoring.