Background

In this post we will review the evolution of this technology over the years. Initially this technology was introduced under the name of personal noise reduction.  Few months later the concept of out-of-the-box deployment with generic profiles was introduced. At the next step, we have renamed this technology to “profile-based noise reduction” based on AI tools to extract any human voice from non-human sounds, like car horn. As of today the technology works out-of-the-box for any typical usage scenario.

The technology in a nutshell

This technology is using a pre-existing generic acoustic profile of the speakers in order to identify their voice during phone calls and attenuate the ambient noise.

Generic Acoustic profiles

The technology is based on sophisticated software algorithms and AI tools and does not require any change in the hardware. In addition, this technology can work bi-directional and filter both the outgoing and incoming audio.

The profile

Our out-of-the-box profile uses a huge database with significant amount of human voice samples in many scenarios and languages.

A generic profile built from multiple recordings

Conclusion

The profile-base noise reduction is a powerful technology that can enhance the quality of any existing phone conversation. It requires minimal effort from the vendor and its activation can be totally automatic and transparent to the end-user.