Exploring How to Display Referential Action to Support Remote Group Discussion

Abstract

Nowadays, remote conferences are widely seen in many situations. Although there is little problem to conduct one-on-one remote conferences, remote group discussion still contains many challenges. One of the main issue is that the remote participants often unnotice the referential action performed by the local participants. To solve such problem, we proposed to develop a function that automatically detects the presence of referential action and displays to the remote participant. One of the issues for this function is about what kind of information of the referential action should be provided and how the information should be displayed. In this research, we conducted a lab study to compare two displaying methods: Picture-in-Picture (PiP) and auto-pilot and compare two displaying contents: object being referred and person performing the referential action. The result shows that the PiP method had higher usability and remote participant had higher opportunity to join the conversation with the PiP method. On the other hand, displaying object being referred had higher usability than displaying the person performing the referential action.