web analytics

Microsoft introduces the ‘Audio Cancel’ feature in Skype

Microsoft updates their popular voice and video platform Skype regularly with new features that make it easier to communicate with others. The platform also played an important role in the organizations and each level of connectivity during the closure last year. But the company does not update the software. This time, we brought the audio cancellation to its desktop version.

In a blog post, the Skype team claims that the technology was originally designed for Microsoft Groups. “We are pleased to announce the release of our latest audio reduction feature in the Skype desktop application. Originally created for Microsoft Groups, this new feature is designed to silence almost everything except your voice when interacting on Skype, ”he added.

This feature can be found under the ‘Settings’ option and can be changed with different options – Default, Low and High.

The Microsoft Skype feature is ‘Sound Cancel’. (Skype)

It is worth mentioning that the audio cancellation feature is not yet available in the web version or Skype mobile versions. You can get it on Windows desktop and Mac too.

To talk more about the technology behind this feature, the blog post adds that the new audio cancellation technology works by analyzing your audio feeds and using deep neural networks to remove sound without touching the speaker’s voice. This is different as traditional blocking algorithms can handle simple, unstable sounds like that of a fan, unable to distinguish when it comes to complex sounds like typing on a keyboard, a food threat or when your pet dog is barking.

“This technology relies on machine learning (ML) to learn the difference between pure speech and sound and is often referred to as artificial intelligence (AI). The representative database is used to train the ML model to work in most of the situations our Skype users face. Adequate variation in the database is required in terms of pure speech, audio types, and locations where our users join online calls, ”the statement said.

To achieve this level of expertise, the team used approximately 760 hours of pure speech data and 180 hours of audio data in their training databases. With audio details, they have installed 150 audio types to cover a variety of situations including keyboard typing, running water, smell and more.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.