|
Google Duo will use machine learning to improve audio quality during calls |
![]() |
Views: 1498
|
Thread Tools | Rate Thread |
#1
|
|||
|
|||
![]()
Google says it has been working for some time to improve the audio quality of Duo calls. The software major is using machine learning technology to achieve it. In a blogpost on Wednesday, the company noted that the Google Duo’s audio quality will be maintained with new technology WaveNetEQ.
The blog notes that WaveNetEQ is based on technology from Google’s DeepMind division. It aims to replace audio jitter with artificial noise that sounds just like human speech. The reason for the audio drop in between calls is that the service can’t maintain a steady connection. Now, with WaveNetEQ, Google’s machine learning algorithm will replicate a larger portion of the missing audio. ![]() Also Read Google begins rolling out new account switching interface on Play Store for select devices For example, 99 percent of Google Duo calls need to deal with packet losses, excessive jitter or network delays. Of those calls, 20 percent lose more than 3 percent of the total audio duration due to network issues, and 10 percent of calls lose more than 8 percent, noted Google. During the ongoing coronavirus pandemic around the world, the reliability on video calling apps have increased significantly for communication. Google, which was rumored to shut down the Duo app, is now suddenly started adding new features being to the service. Last week, Google updated the Duo video chat app once again by increasing the group video user limit from 8 to 12 people into a single group call. The new 12-person limit compares to eight for house party, 32 for Apple’s FaceTime, 50 for Skype and Messenger, and 100 for Zoom’s free tier. Watch Video: Top 5 apps providing free services during coronavirus pandemic Recently, the company added another useful feature to its video calling service Google Duo. It allowed users to send notes and doodles to their contacts. More... |