EmoCo: Visual Analysis of Emotion Coherence in Presentation Videos

Area of Science:

• Human-Computer Interaction
• Affective Computing
• Data Visualization

Background:

• Emotions are crucial in communication and public speaking.
• Multimodal emotion expression (facial, text, audio) is complex.
• Manual analysis of presentation videos is time-consuming.

Purpose of the Study:

• To introduce EmoCo, an interactive visual analytics system.
• To facilitate efficient analysis of emotion coherence across modalities in presentations.
• To improve understanding of emotional expressions and presentation skills.

Main Methods:

• Developed EmoCo, an interactive visual analytics system.
• Implemented channel coherence and sentence clustering views for overview and temporal evolution.
• Included detail and word views for sentence-level and word-level exploration.

Main Results:

• The system enables quick overviews of emotion coherence and its temporal changes.
• Detailed exploration from sentence to word level is supported.
• Evaluation with TED Talks and expert interviews confirmed system effectiveness.

Conclusions:

• EmoCo effectively aids in analyzing emotion coherence in presentations.
• The system provides valuable insights into multimodal emotional expressions.
• EmoCo enhances the understanding and improvement of presentation skills.