Current projects around our speech technologies



In the BMWI-funded project a consortium of large industry, SMEs, start-ups and research partners is building a voice assistance platform for B2B applications.


Audio Mining at WDR

The system developed by Fraunhofer IAIS uses speech recognition algorithms for effective research in the archives of the WDR. It is also used by numerous other partners. 

Live subtitling in the Saxon Parliament

Allinga is used for live subtitling in the Saxon Parliament. The retrained model reliably recognizes the dialects and technical terms used in real time.



Fraunhofer's Acoustic Front-End technology effectively filters noise and echoes from microphone signals, optimizing speech intelligibility for humans and machines.