YouTube’s new AI-based feature automatically enhances audio quality in stories by reducing background noise

YouTube and its community are providing free content to the entire world. Every few days YouTube comes up with a new update to improve users experience.

On 1 October 2020, Google sheds light on a new feature called looking-to-listen for YouTube. This feature is to enhance the audio quality in YouTube videos. This feature helps the content creators in increasing their voices and minimizing the background noises.

Day by the quality of the cameras is increasing the company's are making better quality smartphones but no company has paid attention to the audio quality of the videos recorded by the smartphones.

A lot of people were facing the issues of bad audio quality and many complaints were filed. In 2018 Google developed an app through which the video makers can shoot the video, and add the voice afterwards. Through this technology, the YouTubers can also separate the audio from the video edit it and attach it back. This new technology would improve the audio quality but it was really difficult for the content creators as they had to edit the voice according to the video.

As per Michael Rubinstein, a Google research scientist and Inbar Mosseri Googles software engineer this new update was very difficult to make, and the most difficult part was to add this feature. It took them almost a year to launch this update as it was made as per what people wanted. YouTube directors did a proper survey and questioned the YouTube content creators, that what do they want in this new feature. After finalising all the details they launched this feature but before making it public, YouTube gave it for testing to check if the feature works for all types of recordings and recorded videos, and now finally after one year this new feature is available on all the iOS supporting mobile phone devices. This new feature also ensures the privacy of the users.

Looking-to-Listen works very differently, it first separates the videos from the images in which someone is speaking. When the video is being recorded a component which extracts the recording from it enhances the voice and removes the background noise.

Google’s engineers and researchers say that this new feature has eased the life of the content creators. The running time of this feature is 0.5 times and it also reduced the size of the system to 6MB. And the updated voice is available within seconds.

After a grand survey, Google confirmed that this new feature doesn't completely remove the background noise. A little bit of it is left as the video will not look good without it. This feature works for all types of different languages and accents.

Content creators can enable the enhancing speech button from volume controlling tools after they are done shooting. This feature even allows you to compare the old version and the new version of the voice.

Check out the two samples below:




Read next: YouTube Reverses Picture-in-Picture Block on iOS 14

No comments:

Post a Comment