A tool for multi-modal feature exploration in time-series data
Research on storytelling has focused strongly on the analysis of verbal transcripts. Innovative methods of computational analysis increasingly allow for the analysis of nonverbal expressions of emotions, including prosodic features and facial expressions. The workshop familiarizes researchers with a tool for the analysis and visualization of multimodal expressions of emotions in audiovisual materials. It provides an overview of relevant interdisciplinary work in psychology and affective computing, the application of the tool in a research project on life stories of older persons, and a hands-on workshop where participants can experiment with the tool.
The presentation (showed here) was part of the online workshop "Multimodal visualization of emotion expression" organized by the University of Twente in collaboration with the Netherlands eScience Center. It was part of the UTwente project "Emotion Recognition in Dementia".
For GitHub, see here.