Alea Teeters, Rana El Kaliouby, Rosalind W. Picard

Abstract

We present Self-Cam, a novel wearable camera system that analyzes, in real-time, the facial expressions and head gestures of its wearer and infers six underlying affective-cognitive states of mind: agreeing, disagreeing, interested, confused, concentrating, and thinking. A graphical interface summarizes the wearer’s states over time. The system allows you to explore what others may see in your face during natural interactions, e.g., “I looked ‘agreeing’ a lot when I spoke with that researcher I admire.” Self-Cams will be available for SIGGRAPH attendees to explore this technology’s novel point-of-view, and gain insight into what their facial expressions reveal during natural human interaction.