Google is using machine learning to improve the audio quality of Duo Calls
Google Duo is one of the best video calling apps offering some interesting features. Now when people are dealing with coronavirus pandemic, Google has decided to upgrade the experience of Google Duo. Recently, the company increases max participants limit on Duo video calls. Now the company announced an improvement in audio quality.
Google claims that 99% of Google Duo calls deal with packet losses or network delays. Around 20% of the time, calls lose more than 3 percent of the total audio data while 10 percent of calls lose more than 8 percent of audio. To ensure reliable calls, it is important to deal with packets that are missing.
Google has used machine learning to improve the audio quality of Duo calls. The company has developed WaveNetEQ, a packet loss concealment system based on machine learning. The tech helps Duo to create the missing packets of information using machine learning. Google has studied massive amounts of speech data to train its machine learning model.
This tech was made available in all Duo calls on Pixel 4 phones in December last year. Now the company is rolling out to other models without specifying the model names.
To learn more about WaveNetEQ, you can read the blog post written by Pablo Barrera, Software Engineer at Google. He explained WaveNetEQ in detail with samples for better understanding.