The research activities of the Department of Communications Engineering are at the intersection of signal processing and machine learning. The former is concerned with the enhancement of noisy or corrupted sensor signals and the estimation of parameters and processes, while the latter, machine learning, considers the analysis, recognition and interpretation of sensor signals or other data. In both fields statistical methods are prevalent, whereby "deep learning" approaches have been particularly successful recently.
The first and foremost application field of our research is speech processing. Here we offer a large bandwidth of activities, ranging from speech enhancement, acoustic beamforming, blind source separation, dereverberation and denoising, feature extraction and automatic speech recognition. Furthermore we also consider applications in the field of digital communications and localization, such as secure communication and the localization and tracking of terminals by analyzing the radio signals of wireless communication systems.