Using a virtual reality technique, a double step experiment was undertaken to investigate the human performance of reaching and grasping. Subjects were asked to reach and grasp a computer-generated 3D target image in a stereo-display. During reaching movement, the target suddenly changed its position when the subject's hand comes within a certain distance from the starting position (double-step-paradigm). The reaction of reaching and grasping movements and saccadic eye movements were studied. The results showed that the velocity peak of reaching is the time until which altering action is possible. Also, the latency of saccadic eye movements to foveate the second target became shorter when it was accompanied with reaching motion. The experimental system has proven to be useful for the quantitative study of human visual perception and cognition that accompanies motor control of arms and fingers.

抄録全体を表示

In a network telecommunication system, the binaural technique is promising for transmitting spatial information of a sound, i. e., the direction and the distance of a sound source. In such a system, however, data compression of sound signal is usually required due to restricted bandwidth of a channel. The effects of the data compression on the perception of a sound image were examined in this article when TwinVQ (Transform domain Weighted Interleave Vector Quantization) was used as audio coding. Experimental results showed that the right-left direction of the image was preserved correctly although front-back confusion tended to increase. Moreover, the sound image was broadened by the coding.

抄録全体を表示

When a stimulus set is spatially orthogonal to a response set (e.g., vertical stimulus and horizontal response configurations), an above-righfc^elow-left mapping has an advantage over the reverse mapping (orthogonal SRC effect). In this paper, three experiments were conducted with four stimuli (two above and two below) and two responses (one in each side). All of the stimuli and responses were set within the same task space. When the distinction between each stimulus side was enhanced by serving a fixation or by grouping the stimuli in each side, orthogonal SRC effect occurred. However, when the response position was used as a reference point, orthogonal SRC effect disappeared. This suggests that the stimulus space and the response space are not integrated even if the both are set within the same coordinate.

抄録全体を表示

A man applies load on a soft object with his fingers to evaluate its hardness and detects a force, deformation and touch sensation. As soft objects in our environment are finite, the boundary conditions have a significant effect on our evaluation. In this paper, hardness evaluation of soft objects with different kinds of thicknesses was examined and its accuracy was confirmed. In the experiment isolating the thickness information from subjects, the accuracy of evaluation decreased. This result suggests that the accuracy of the thickness and the size of the object severely affect the hardness evaluation in a virtual environment. In a further experiment restricting touch sensation, subjects evaluated the hardness inaccurately according to the force and deformation which was affected by the boundary conditions.

抄録全体を表示

Visually induced self-motion perception (vection) arises with a latency period of several seconds. It was tested whether the sense of self-motion induced by an optical flow remained when the direction of the optical flow changed to a new one. Measurements of the latency period for vection in the new direction showed that the latency period was only 15% shorter than that for the first vection. The shortening can be explained partly by the fact that the visual patterns preceding the second flow were dynamic. It was concluded that vection was lost when the direction.of the flow changed, and that vection in a new direction requires a new latency period.

抄録全体を表示

As the human characteristics of space perception, the phenomena of subjectively straight lines such as Helmholtz's horopter, the parallel alley, and the equi-distance alley are well-known in binocular visual space. In this paper, these phenomena and other similar phenomena in tactile space and auditory space are introduced. Mathematical models to explain these phenomena are also introduced, and an example of the application to the virtual system is explained.

抄録全体を表示

This paper was aimed to estimate the perceptual strength of visually induced self-motion perception (vection) by focusing on the beating field (BF) ofoptokinetic nystagmus (OKN). For this purpose we used an immersive type VR system which provided 3D visual stimuli for observers at the most of the entirely visual field. The BF during vection was found to displace to the perceived heading direction in proportion to the strength of vection. This displacement preceded the difference in the subjective estimation of the strength of vection between viewing conditions. Thus the strength of vection was estimated by measuring BF of OKN even before the actual observation of the subjective estimation of vection.

抄録全体を表示

This paper describes a new concept system for physical and mental activation of the elderly which is designed to emphasize the factors of interaction and amusement in virtual reality(VR). It presents 3D visuals, sounds, and a vibration feedback and detects the user's movement A 60-inch display, two liquid crystal projectors with polarizing filters and polarizing glasses are used to produce 3D images. A pair of virtual maracas with a speaker, a vibrator and a LED marker for detecting 3D position was developed for interaction. After some improvement and assessment the developed system was introduced to the group recreation program for the elderly in day-nursing services. Two kinds of VR games, that is a balloon smashing and a mole hitting were applied for a week to 38 and 49 persons respectively. Operation of the adjustment in difficulty was successful and almost all participants, both with and without dementia symptoms were able to enjoy VR games. The effects of mental and physical activation were observed in movements, facial expressions, and verbal responses.

抄録全体を表示

実体験のように感じる実画像表示の主な目的は、写真に写された実際の場所を画面上で見渡すことです。そのような画像表示を実現するためのハードウェア部品は低価格で購入できますが、一般用のソフトウェアシステムは、デバイスとコンピュータ基底対話型コントロールの基で統合してすぐに利用することはできません。ステレオグラフィック表示システムで頻繁に使われる特別な部品の需要に応えるために、私たちはアップル社のQuickTime VR技術とJava言語を用い、JavaのためのQuickTimeアプリケーションプログラミングインターフェース(API)の支援を受け、パノラマ&オブジェクトムービーブラウザを開発しました。この斬新なQTVRブラウザにより、仮想環境で視野を調整することかできるようにうなりました。ただし、撮影位置、ユーザーの周りで使用できるモニターとプロジェクターの数とサイズにより、表示できる視野は限られてしまいます。VR_4U_2Cというブラウザは会津大学のスぺーシャルメディアグループマルチモデルグループウェアにおけるクライアントのひとつですVR_4U_2Cは、コンピュータと伝統的な入力装置(マウス、トラックボ-ル、回転式コントローラ、トラックパッド、キーボード)、その他のインターフェース(スピーカー配列空間オーディオ装置、携帯電話、サーボモーターコントロール付回転イス)を使用することにより、対話形式で多次元の仮想空間を表現することかできます。VR_4U_2Cは、動的な旋回技術であり、QTVR画像を立体的に表示する際に生じ得る問題を解決することかできます。

抄録全体を表示

This paper presents a mixed reality (MR) system with tangible interface as well as visual fusion in an MR space. The sense of touch is given by physical objects on which computer generated imagery is accurately registered and superimposed. This approach is different from most of tangible or Haptic displays, which utilize special display hardware to give the feeling of touch, and "physical icons" or "phicons," which mediate the user and controlled objects. The proposed approach is especially useful in industrial design and production process where digital mockups and physical mockups are thoroughly utilized. Concept and technical details are described on a practical system specially designed and implemented for an evaluation of automobile interior design.

抄録全体を表示

In this paper we present experimental results that show how people perceive other's pointing gestures. We measured human perception of natural finger pointing under a face-to-face condition to find out how people understand other's pointing gestures. The experiment revealed that people try to locate pointed position by extrapolating an orientation of an index finger of the pointer without using gaze information of the pointer even if the finger provides misleading information. Also, when a target is pointed using both eyes and a finger, people concentrate on the finger information even if people can locate the pointed position by watching eyes.

抄録全体を表示

This paper presents a new mixed reality technique for controlling the real object's color appearance with light projection. Our system illuminates the real object with at least 4 mono-RGB color lights in advance, and then it estimates what color should be illuminated for a target color. This approach can be applied to planning of the color restoration of color-degraded cultural properties, and a presentation in a museum. In this paper, we explain the experimental result for changing the color appearance of some gradation patterns and a painting and the effectiveness of our proposed method.

抄録全体を表示

We have been studied on an information display based on saccadic eye movements. The display can present 2D images using only a single line of flickering light sources. When people make a saccade across the light sources, they can perceive 2D images thanks to the retinal afterimage drawn by the saccade. Though, so far, saccades of observers are induced by other stimuli such as two sequentially flashing light sources, if the saccades are remotely detected, practical applications can be realized. In this study, we present a new technique for detecting saccades, based on the fast, robust pupil detection method. The method uses two infrared light sources synchronized with the camera's frame rate. The two light sources generate bright and dark pupil images, and we can carve out the position of the pupil by subtracting them. From the series of the captured image, change of the eye gaze is measured as the displacement of the pupil's position. Measuring the displacement with a line scanning camera, which can take one line image with high frequency, the saccade can be detected. We investigated the feasibility of this technique by clarifying the retroreflective features of the retina and required spatial resolution for detecting saccadic eye movements.

抄録全体を表示

This paper proposes an "asobi-litation" system that utilizes a haptic sensation in order to enhance non-varbal communication such as body contact between handicapped child and his/her teacher. The haptic sensation generated by a 3DOF manipulator enables the child to handle virtual objects computed with PC. The manipulator also helps both child and teacher to touch each other while they share the haptic feedback. We developped a prototype system and then 23 children (20 of them are handicapped) were experienced it. As a result body contact was observed through their performances. Except for some children who did not have any conversations with us, the subjects answered that they are familiar with the system.

抄録全体を表示