Automatic Analysis of Group Conversations via Visual Cues in Non-Verbal Communication

The scientific objectives of NOVICOM are the development of new computational methods to automatically detect and analyze nonverbal behavior from the visual channel of communication for the automatic analysis of social interaction in face-to-face conversations. The project will investigate models of communicative gestures and postures, and models towards the integration of multimodal cues for analysis of group interaction, building upon and extending our current work on social inference from sensor data. NOVICOM brings complementary expertise to address a number of aspects known to be fundamental components of nonverbal communication, but which have not been thoroughly studied in the context of computational modeling. The project is funded through the Marie Curie Actions Program on Inter-European Fellowships for Career Development.