Cognitive Systems Lab

The Cognitive Systems Lab uses an interdisciplinary and integrative approach, which combines highly-realistic, interactive, and immersive mixed-reality environments with high-resolution, multimodal physiological measurements, and state-of-the-art AI approaches. Current work in the lab focuses on how to parametrize affective signals used for communication - both via the face and via touch, how reliable affective datasets really are, and whether Koreans are able to categorize other (East Asian) populations. Important aspects concern explainability and transparence of algorithms, actual clinical relevance, as well as approaches designed to fit to the clinical workflow. This research line combines deep learning, virtual reality, psychophysiological measures, and neuroimaging to chart and model human and AI decision-making in high-risk situations. In this newest one among our research foci, we are looking at how to characterize and improve the alignment between humans and AI - this concerns the development of explainable AI approaches to open the black-box, all the way up to measuring and comparing how humans and AI fare on the exact same task at perceptual, cognitive, decision-making, all the way up to moral levels of analysis.