A Novel Affinity Enhancing Method for Human Robot Interaction - Preliminary Study with Proactive Docent Avatar

Abstract

Affinity generation is one of the essential challenges still remaining in the HCI field. Today, photo realistic digital humans are developed as intermediaries in human-computer interactions to form affinity, but so far there are still many responses that it is awkward. In order to generate affinity, we hypothesized three conditions(proactive, expressive, intuitive) and applied them to the interactive systems under development. To be proactive, human appearance recognizer and gesture recognizer by using the Kinect sensor are utilized to attract people. To be expressive, the conversation corpus consists of voices(recordings of voice actor), body gestures(motion libraries and motion captures), facial expressions (iPhone face capture), and visuals(3d model animations, videos, figures). To be intuitive, we developed novel hand gesture recognition-based graphical user interface to navigate questions smoothly. We will validate effectiveness of the interactive system consisting of historical character docent avatar and historical-related content in various exhibition spaces such as science museum. The interactive system includes attraction function, game, 400 question-and-answer set.