In this project seminar we will dive into the "dark matter" of social neuroscience, that means the study of mental and emotional processing of others in real-time social encounters in a truly interactive manner.
With that we will outline the theoretical conception of a second-person approach to other minds, design and hopefully run a study in the field of a second-person neuroscience.
In the winter term 2020/21, we will work towards a detection and prediction model for emotional mental state changes from the face based on our pilot data on facial expression dynamics in real-time dyadic interaction pairs (Müller et al., 2018- see https://arxiv.org/abs/1810.12171). This will be the first step towards a multi-channel prediction model (incl. face, body and gestural information) in real-time social encounters. For that we have 35 interaction pairs recorded in a real-time discussion situation and use state of the art machine learning algorithms to analyze the data.
If possible we will go for a large scale study in the summer term of 2021.
Note, from Janurary 2021 Cem Dogdu will join the course coordination team as a young PhD candidate scientist. |