Back in February, Google engineers introduced Lyra, “a high-quality, very low-bitrate speech codec that makes voice communication available even on the slowest networks,” according to the original blog post.
The tool was developed using traditional codec techniques combined with a machine learning (ML) approach with the aim of providing universally accessible, high-quality audio experiences. Lyra is designed to operate at 3kbps and thus can be used even on very low bandwidth connections.
Now, Google is open sourcing a beta version of Lyra under the Apache license to provide developers with access and to obtain feedback on the project. The Lyra code is written in C++ for speed, efficiency, and interoperability, according to the announcement, and “this release provides the tools needed for developers to encode and decode audio with Lyra, optimized for the 64-bit ARM Android platform, with development on Linux.” You can find the Lyra code and more information on GitHub.
Comments