There is a small chance that someone has managed to avoid video or voice calling services that rely on internet connections at this stage, not least thanks to the global pandemic.
Suggestion For You:
There’s an even smaller chance that these calls will be completely free of dropouts, glitches, latency, and other disruptive artifacts, but the latest advancements from Google’s AI division should help in this area.
Lyra’s enhancements are the result of Google AI’s combination of existing codec technology and “advancements in machine learning with models trained on thousands of hours of data.” This data includes speakers in more than 70 languages, freely available in open source libraries, to ensure its universal applicability.
The blog itself takes a much deeper look at how this machine learning enabled the new technology, and also uses audio clips to show the difference between the original audio, Lyra and other codecs at different bitrates.
The team plans to continue to optimize Lyra and improve its quality and reliability in the hope that it will be used outside of the Google Duo app. This will likely mean that other VOIP and video conferencing apps and services will take advantage of the codec in the future.
In addition, the report states that the team “is also beginning to investigate how these technologies can lead to a common low-bitrate audio codec,” which could mean that non-speech instances, such as music and other audio, receive the same improvements.