Fraunhofer IIS participated in the Consumer Electronics Show (CES) 2025 in Las Vegas, where AI-based microphone processing technologies developed in the DSAI project were presented to a wide range of prospective customers.

An “office” in a Venetian hotel suite was used to show AI-based microphone processing technologies, such as echo control, noise reduction, dereverberation and personalized target speaker extraction, which facilitate crystal-clear voice communication. Personalized target speaker extraction was the main highlight, where users could create a “fingerprint” of their voice in seconds, which made it possible to remove all sounds but their voice from the audio signal in real time. Such technologies help to ensure a more pleasant conferencing experience.

In addition, a html-based demo was presented, where visitors were able to browse through different scenarios and listen to the enhanced voice signals.

Online demos can be found here. The demos work best with Chrome/Firefox on a PC/Mac and wired headphones.