Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

A low-complexity state-space architecture for multi-microphone acoustic echo control

: Luis Valero, M.; Mabande, E.; Habets, E.A.P.


Saruwatari, H. ; Institute of Electrical and Electronics Engineers -IEEE-:
16th International Workshop on Acoustic Signal Enhancement, IWAENC 2018. Proceedings : 17th-20th September 2018, Tokyo, Japan
Piscataway, NJ: IEEE, 2018
ISBN: 978-1-5386-8151-0
ISBN: 978-1-5386-8150-3
ISBN: 978-1-5386-8152-7
International Workshop on Acoustic Signal Enhancement (IWAENC) <16, 2018, Tokyo>
Fraunhofer IIS ()

Modern hands-free communication devices, tablets, smart speakers and televisions, employ multi-microphone signal processing for speech enhancement and recognition. Given a multi-microphone (MM) setup, the signals reproduced by the loudspeaker(s) are fed back to the microphones, resulting in acoustic coupling. The most common solution to reduce acoustic echoes is to employ one echo canceler per microphone. Yet, in practice, such an implementation may not be feasible due to computational complexity constraints. Moreover, regardless of the adaptive algorithm used, audible residual echoes may remain after echo cancellation, and additional residual echo suppression is commonly required. Following up on recent work on complexity reduction of MM acoustic echo cancellation, a state-space architecture for low-complexity MM echo control is presented here, which incorporates residual echo reduction adding nearly no computational complexity.