KinConferencing: Using Motion Sensing to Study and Visualize Gesture Use in Computer-Mediated Communication

• Main Author: 賴建同
• Other Authors: 王浩全
• Format: Others
• Language: en_US
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/80397762972722431357

碩士 === 國立清華大學 === 資訊工程學系 === 102 === One key difference between face-to-face (F2F) communication and computer-mediated communication (CMC) is the availability of visual cues. It is often assumed that the reduction of visibility in audio and video conferencing may negatively impact the use of gesture to communicate, and thus negatively influence the outcomes of communication. There remains limited understanding around how visibility and richness of communication media may influence the non-verbal aspect of communication, especially gesture use. On the other hand, another important issue in the design of CMC tools is on how to enhance communication actively. We consider that it is possible to promote the use of gesture in communication by providing appropriate feedback such as visualization that displays one’s own and the conversational partner’s amount of gesture during the conversational process. If we can successfully increase the amount of gesture with graphical visualization, then it is possible for us to create a technology design that functions as well as F2F and video, while still prevents from some negative side effects such as disclosure of private information or communication anxiety. This may give users more options of CMC in everyday usage when they don’t want to reveal too much personal information, such as appearance, to their conversational partners. In this thesis, I conducted two studies. First, I studied how visibility influences gesture use by “Kinect-taping” F2F communication and CMC in brainstorming groups. “Kinect-taping” (similar to the sense of videotaping) refers to an analytic technique of using motion sensors to record and analyze group members’ hand movements during communication. Based on the positional data of body parts obtained from the motion sensor (Microsoft’s Kinect in this work), the technique helps derive measures useful for the purpose of behavioral analytics, such as the amount of gesture use by individuals and the extent of gestural similarity in a conversational dyad. Second, I used the Kinect-taping technique as the basis to design a new visualization interface by showing the amount of one’s own and the partner’s gesture use during online communication. I evaluated the effects of the visualization on gesture use and perceived privacy threat for communication media with different levels of visibility. Implications to future research and design are discussed.