Three-dimensional cyberspace systems for computer networks are becoming popular shared virtual environments. The InterSpace system developed at NTT is one such system. The InterSpace system has real-time voice, video and text chat tools as well as a surrogate (avatar) interface. This paper describes the architecture of the InterSpace server, which enables flexible configuration of server and selection of service on the client PC. We implemented the architecture and tested it. The results of the test indicated that the voice delay is about 1 sec and the most loaded server is the audio communication server. It is also realized that CPU time of position management server will increase according to the number of users logging-in.

抄録全体を表示

In a conventional face-to-face meeting, the attendees gather around a common space such as whiteboard for taking down notes and exchanging information. It is indispensable for attendees not only to meet each other but also to share a sense of collaboration. In addition to those, attendees' voice, facial expressions, gestures, and even the atmosphere surrounding the meeting room all are essential to a cooperative environment. Therefore, we propose a remote conferencing system MAX to support multi-location, cooperative working environment with the architectural integration of the so-called "communication" and "working" spaces. In the integrated virtual space, users, while on the move, can exploit interacting with others freely. It is also possible to utilize pen-based writing device for simple and natural input. We analyzed the data obtained from various experiments. By adjusting the timing factor for setting up a conversation and writing collectively, the progression of the meeting could be controlled seamlessly.

抄録全体を表示

We have developed an simplified image-based rendering technique which exploits real image texture mapped boards. Our aim is to improve the reality of visual contents and the efficiency of visual contents development phase in order to construct cyber spaces. The shape of the texture and the arrangement of the board are altered so as to follow the user's changing point of view(POV) and look as if actual objects exist. We call this board "Texture-Object". The transformation process according to user's POV is as follows: One real image texture is divided into some subsections to each of which we apply a different Affine transformation. In this paper we will present its design concept, display process, implementation of our prototype system, evaluation for making contents of virtual cities.

抄録全体を表示

In this paper, we pay attention to the human action and describe a way to solve the gap problem between the real world and the virtual space. We try to support high quality communication among the members and provide triggers for informal communication by letting the avator have gestures for the member's information in a virtual space. First of all, by applying the spotting recognition of gestures from time-varying image, the avator reflects user's state and action. Secondly, we pay attention to a handshake and realize "virtual handshake device" for the way to make users feel virtual action. We have build these systems on a virtual office system named "Valentine" that we have built before.

抄録全体を表示

On the Internet many services of 3D multi-user virtual environments (3D-MUVEs) are already provided, however, they support only limited communication methods such as chatting, sharing avatars' gestures and so on. We consider that more avatar-centric metaphors are required for communication methods such as the use of tools by avatars, and touch among avatars as non-verbal communications. In this paper, we provide a description library for avatar-centric communication methods in our target platform, 3D-MUVEs on the Internet. We demonstrate the description capability of our library through the applications to proposed avatar-centric metaphors.

抄録全体を表示

We constructed the agent-oriented virtual society PAW on our Community Place system. Unlike usual 3D chat systems, there are a lot of autonomous agents acting with users in PAW. The agent is also able to interact with users in various ways. In this paper, we give an outline of the architecture of PAW, and present some interesting statistical result from its operation.

抄録全体を表示

In this paper, we focus on the sense of locomotion, in particular, the sense one feels while walking in a large space. Walking is the most basic way of moving around, and a lot of information can be obtained by doing so, e.g., sensing, hardness, humidity, and so on of the space. The locomotion interface we have developed is named ATLAS (ATR locomotion interface for active self motion). Our design goal was to eliminate the need for a walker to "learn" how to walk and for equipping the walker with obstructive sensors. We employ a treadmill approach for ATLAS. In the initial trial state, ATLAS can estimate the working speed from the walking motion visually, and control the speed of a belt to synthesize the real feeling of walking in a virtual space. First, we describe related works of locomotion interfaces and our design. Next, we show the method utilized for motion analysis and how to control the belt speed. Finally, we conclude with a description of our first trial ATLAS and the experimental results.

抄録全体を表示

On the present condition that computers mutually connected each other, sharing information resources and forming a huge network, it is usual for user to access and use the distribute system. We regard such relation between user and system as changing distributed system to social space. In this paper, we describe about user interface which provide information based on distributed system by use a metaphor of spacial representation. We call this metaphor "Spacial metaphor".

抄録全体を表示

In Virtual Reality , it is important to merge an image from the real camera and a synthetic object from computer graphics. For examples, a panoramic image enables user to pan inside its imaging world. But the user can't interact the world, only view around. We introduce the background image and a frontal CG object synthesize method considering with their interactions. The method detects a collision between the background image and the frontal CG object pixel by pixel comparing two depth maps. And, we simulate collision between a synthetic spheral object and a captured room image using this method.

抄録全体を表示

Motor skills are difficult to be transferred from an expert to a novice. It is not easy for an expert to describe his skill linguistically. Videotaping is also insufficient since motor skills involve invisible elements, such as force and impedance. VR techniques have a potential to make training more efficient and let a novice to acquire motor skills much easier. In this paper, we propose an efficient training method for motor skills, such as sport swing motions, by using vibration cue and teacher's sight. We constructed a one-dimensional system and evaluated the proposed training method. The experimental result shows that the teacher's sight is effective. The vibration cue is effective only when the teacher's sight is withdrawn.

抄録全体を表示

In the field of Mixed Reality, one area of Virtual Reality, the method of merging virtual and real spaces have been often discussed. A chromakey-imposer is often used to synthesize CG and real images, but the relation between CG and real images is either in front or behind because a general chromakey-imposer has only one chromakey plane. To make one advance in the former method of merging we propose a method to make a visual effect as if a real space is embedded in a virtual space by dividing an object space into three layers and using two chromakey planes. We tested the visual effect by treating the object space as a studio space which has a large area and a desk space which has a small area.

抄録全体を表示

It is important to represent the shape of an object in a virtual environment (a virtual object), because it enhances ambience in case an operator contacts a virtual object. Recently several systems have been proposed which aims to represent the shape of a virtual object and they are classified into "encounter type." In encounter type an operator feels touching a virtual object by measuring his/her motion and locating a corresponding real object in order to approximate the shape around the contact point. However, it is not proper to directly locate the real object so that it approximates around the contact point because the contact point is not always continuous and therefore the trajectory of the real object will not be continuous. In this paper we consider the continuity of the contact point and propose a method to set up the continuous trajectory of the real object when representing arbitrary surface in encounter type. Proposed method realizes its continuous trajectory by changing the manner of setting up its trajectory according to the distance between the fingertip and the represented surface, and by interpolating the trajectory between both manners. Results of numerical simulation applying a proposed method to several quadratic surfaces show its validity.

抄録全体を表示

It has been shown that HRTFs (Head Related Transfer Functions) are important cues for judging the absolute auditory distance of a single sound image when the sound source is close to the listener. In order to investigate the role of HRTFs in auditory distance perception more generally, not only the absolute distance of a single sound image but also the relative distances among multiple sound images should be considered. From this point of view, two kinds of psychoacoustical experiments on absolute and relative distance perception were conducted with the same source signals for stimuli. Comparison of the results of the two experiments showed that while the absolutely judged distance of a sound image increases with the actual distance of sound source up to around 1.2 m, the results of the relative auditory distance perception showed that the perceived distance significantly increases up to 2〜3 m. This difference may be attributable to some perceptual information stored in a short-term memory provided by the comparison of auditory distances between multiple sound images, which could offer an additional cue in relative distance perception, and it may be effective to improve the resolution of the distance perception of sound images at longer distance than the limitation in absolute distance perception.

抄録全体を表示

In this research we investigated postural sway of subjects induced by optical flow composed of random dots on computer graphics which simulated straight orbits and curve orbits. The optical flow was varied with curvature, velocity and how to display. We could conclude that the body sway is mostly induced by a change of curvature and clues of image and that the subject sways to centrifugal direction and bends forward when curvature is varied and that the subject sways to centripetal direction in advance if he recognizes curve orbit.

抄録全体を表示

In this paper, a tactile display using elastic waves in a 'tapered membrane' (a membrane whose width decreases gradually) is proposed. When an elastic wave is input to the tapered membrane, there exists a boundary between a vibrating area and a non-vibrating area. The boundary is called 'turning point', and the position of the turning point is controlled by the frequency of the wave. It was reported that a surface of a vibrating object with ultrasonic frequency feels smooth by squeeze effect. Thus, in the tapered membrane, the boundary between smooth and frictional areas can be moved by the frequency of the wave. In an experimental device employing a brass plate vibrated by a Langevin ultrasonic vibrator, these basic phenomena are confirmed.

抄録全体を表示