The group researches user- and task-oriented methods for spatial audio recording and audio playback for use in AR. The research focuses on investigations and developments in the area of position-dynamic binaural synthesis as well as spatial auditory perception for the generation of highly plausible auditory illusions in real space.
Our researchers are interested in AI-based metadata enrichment of databases using so-called few-shot learning, where the goal is to use as little but very meaningful training data as possible. Automated measuring methods are also used here.
For our investigations and developments in the field of spatial audio and audio information retrieval, existing methods of acoustic and perceptual quality evaluation are applied and further developed. Furthermore, we are researching new methodologies for quality evaluation in position-dynamic AR scenarios.