In this project, systems supporting everyday casual meetings were developed. Our systems running on computer networks provide environments where people meet, and opportunities to communicate together. Casual meetings occur accidentally among unpredicted many people. The characteristics of casual meetings are different from those of business meetings planned in advance.
FreeWalk is a desktop videoconferencing system with a three-dimensional (3-D) virtual space. This system provides a common space where everyone can meet and talk freely. Each participant is represented as a 3-D object on which his/her live video is mapped, and has a location and a view direction. He/she can move and turn freely in the space. Figure 5(a) is a example of personal vision drawn on the screen. Voice volume is inversely proportional to the distance between sender and receiver. We use stereo sound for voice orientation. Therefore, people can form separate meeting groups without bothering each other.
|(a) Personal Vision of a Virtual Space||(b) Conversation through a Large Screen|
We have applied video game technologies to the implementation of FreeWalk. The movement of users' object is controlled by the X and Y valuators of the mouse pointer. Since the viewing angle of the CRT is much narrower than that of human eyes, auxiliary indications and view point switching are introduced to help users to grasp their situations. We have tried to connect a real room equipped with a large graphics screen with the shared 3-D space of FreeWalk. The room has several video cameras, and participants visiting the 3-D space can select appropriate video images provided by the cameras. Figure 5(b) shows that one participant talks with other participants in the 3-D space FreeWalk provides.
We conducted experiments in the Internet to evaluate the performance of FreeWalk. We also compared communications in FreeWalk to a conventional desktop conferencing system and a face-to-face meeting. The casual effects of 3-D virtual space was found. Figure 6 is a result of this experiment. In this experiment, FreeWalk increased the number of turns (figure 6(a)), and stimulated participants in moving around to realize free conversation (figure 6(b)). FreeWalk was developed to run on workstations at first. After that, more practical software of FreeWalk was implemented to run on PCs. FreeWalk software is available on the Web site (/free-software/freewalk/). It was downloaded by more than two hundred people from more than ten countries.
|(a) Frequency of Turns||(b) Movement of Participants in the Virtual Space|
We have developed a system, which provides awareness support for real-world encounter. The shadows of people are projected on a large graphics screen. This system presents common topics so that people can easily start conversation and relations among them. Topics are selected based on their personal profiles. In our current implementation, we display World Wide Web (WWW) pages as topics. Figure 7(a) shows the screen image of this system. In Figure 7(b), participants use Silhouettell system to have a conversation.
|(a) Screen Image||(b) Using the System|
Through an experimental use of Silhouettell by people from various countries, we observed and analyzed conversation in the environment where topics are presented on a large screen. It appears that information directly related to participants such as their own Web pages, attracts them, and when the language difference makes it difficult for participants to communicate with the other, they tend to refer to the contents on a screen.
To realize smooth interaction among dispersed people, we transmit mutually video and audio between remote places. The video image is displayed on a large screen. We also emphasize social cues of interaction to provide as the same kind of opportunities to begin communication as in the face-to-face environment. For example, when remote people approach the screen, the resolution of the video image is increased to attract the attention of local people. Figure 8 shows the change of resolution of the video image, in which you can see a remote person is approaching.
Servers TD> A>