Skip to main content



Every dialogue between people requires the recognition of the emotions of the counterpart. In addition to communicating in personal contact, interaction with digital media (eg telephones, social avatars) is a special challenge today. Older people as well as people with autism whose social cognitive abilities are impaired can use the planned training system to send out and understand emotional communicative signals.

The targeted training system relies on the development of models for the automatic recognition of complex emotions (such as boredom, frustration or interest) and the integration of various non-verbal communication channels (facial expressions, gestures, direction of vision and physiological excitement). During the training on the computer, these signals are to be used to provide direct feedback and to adapt to task difficulty and situational context. Extensive studies on the effectiveness of the training are carried out with both target groups.

Innovations and perspectives

The training is designed to enable older people to read and express emotions accurately, in order to strengthen them, among other things, in dealing with communication technologies. Furthermore, it is intended to help people with autism to integrate into interactive social contexts, for example the working world. The aim is to significantly improve the participation in social life for both target groups.

Dipl.-Inf. Tobias Moebert

Dipl.-Inf. Tobias Moebert


Campus Griebnitzsee
House 4
Room 0.12

University of Potsdam
Department of Computer Science
August-Bebel-Straße 89
14482 Potsdam