Our research in this field concentrates on the development of single- and multi-channel speech enhancement algorithms such as acoustic echo cancellation and noise suppression, as well as artificial speech bandwidth extension especially designed for home, office and smartphone applications.

For the domestic living environment, multi-channel microphone systems in combination with multi-channel acoustic echo cancellers and noise suppression algorithms are employed to provide  an independent hands-free functionality and speaker localization. A specifically equipped flat functions as lab and allows the group to conduct highly-realistic experimental setups.

A market-ready wideband hands-free system has been developed for office applications on a DSP platform, which provides high-quality speech transmission for landline and VoIP connections-even for narrowband signals.

Model-based noise suppression algorithms for use in cell phones and other mobile devices have been developed and are continuously improved. For modeling statistical properties of speech, we utilize powerful machine learning approaches, such as deep neural networks, for further improving our noise suppression and artificial speech bandwidth extension algorithms.