Multimodality, Context and Continuous Emotional Dynamics for Recognition

Explore the complexities of human emotional expression through a comprehensive lecture on multimodal emotion recognition. Delve into computational approaches that address multimodality, contextual information, and continuous emotional dynamics in human communication. Learn about hierarchical frameworks incorporating temporal context for improved emotion recognition, and discover methods for continuously estimating emotional states during dyadic interactions. Examine the integration of various information sources, including speech, facial expressions, and full body language, in multimodal systems. Gain insights into the application of these computational approaches in healthcare, particularly in analyzing facial expressions of children with autism spectrum disorders. Understand the potential of these techniques in highlighting emotionally salient regions in long interactions and quantifying atypicality in affective expressions.