EEMML: the emotional eye movement animation toolkit

Abstract

Eye movement plays an important role in face to face communication in that it conveys nonverbal information and emotional intent beyond speech. Being "a window to the mind", the eye and its behavior are tightly coupled with human cognitive processes. In this paper, we proposed an Emotional Eye Movement Markup Language (EEMML) which is an emotional eye movement animation scripting tool that enables authors to describe and generate emotional eye movement in virtual agents. The language can describe eye movement parameters we derived from facial expression database as well as real-time eye movement data (pupil size, blink rate and saccade). EEMML provides the input for our eye movement generator system with one or more eye movement actions in sequence. The language is extensible, so that new rules can be quickly added. It is designed to plug into larger human-agent or agent-agent interaction systems. We present an evaluation in which subjects evaluated the EEMML and gave their feedback. The results indicate the validity of our approach.