Yuichi Bannai

Presentation Technique of Scent to Avoid Olfactory Adaptation

Several tracking techniques for augmented reality have been proposed. In feature point tracking, a pose is computed by minimizing the error between the observed 2D feature points and the back-projected feature points from the 3D scene model. This minimization problem is usually solved by nonlinear optimization. The main advantage of this approach is its accuracy. However, it is difficult to compute the correct pose unless an appropriate initial value is used. In addition, when an observation contains some errors, this approach does not guarantee a correct pose even if it converges to the global minimum. Therefore, once an incorrect pose is computed in a frame, either the tracking in the next frame may fail or the result will deviate from the correct pose. In this paper, we propose a new tracking framework for augmented reality. The proposed method tracks features as multiple local hypotheses based on not just one pose but multiple poses that are computed from pose estimation in the previous frame. Since multiple poses are maintained as global hypotheses, as long as the correct pose is contained in the hypotheses, tracking can be continued even in difficult situations such as a simple iterative scene with high-speed movements.Trials on the transmission of olfactory information together with audio/visual information are currently being conducted in the field of multimedia. However, continuous emission of a scent creates problems of human adaptation to the lingering olfactory stimuli. During long movie scenes, viewers can not detect an emitted scent continuously. To overcome this problem we applied pulse ejection to repeatedly emit scent for short periods of time to ensure the olfactory stimuli do not remain in the air to cause adaptation. This study presents the decision procedure for the ejection interval Deltat required while considering the olfactory characteristics of subjects. The developed method provided the user with an olfactory experience over a long duration, avoiding adaptation.

Effective Presentation Technique of Scent Using Small Ejection Quantities of Odor

Trials on the transmission of olfactory information together with audio/visual information are currently underway. However, a problem exists in that continuous emission of scent leaves scent in the air causing human olfactory adaptation. To resolve this problem, we aimed at minimizing the quantity of scent ejected using an ink-jet olfactory display developed. Following the development of a breath sensor for breath synchronization, we next developed an olfactory ejection system to present scent on each inspiration. We then measured human olfactory characteristics in order to determine the most suitable method for presenting scent on an inspiration. Experiments revealed that the intensity of scent perceived by the user was altered by differences in the presentation method even when the quantity of scent was unchanged. We present here a method of odor presentation that most effectively minimizes the ejection quantities.

Time characteristics of olfaction in a single breath

The transmission of olfactory information together with audiovisual information is now attracting much attention. However, the information is difficult to synchronize because of problems of scent lingering in the air and olfactory adaptation. We aimed at minimizing the amount of odorant presented to users in order to mitigate these problems, and developed an olfactory display that is able to present scents precisely. The display uses pulse ejection, whereby scents are emitted for only short periods of time. In this study, we aimed to mitigate the above-mentioned problems and to measure the time characteristics of olfaction in a single breath, which are difficult to measure by conventional methods. As a result, the most effective conditions for using a small amount of odorant in a single breath were revealed. These results are expected to ease the synchronization of olfactory and audiovisual information.

Development of a High-Performance Olfactory Display and Measurement of Olfactory Characteristics for Pulse Ejections

For transmitting olfactory information together with audio/visual information in the field of multimedia, it is necessary to control the scent presentation in accordance with the changing images and sounds presented to receivers. However, continuous emission of scents creates problems of lingering scent in air and human olfactory adaptation. To overcome such problems we applied repeated pulse ejections to emit a scent for short periods of time. We remodeled an olfactory display having high emission control ability. This display can control emit for a unit of 667 sec. It has been found out that the shortest time of pulse ejection which everyone can sense is 33 msec in this preliminary experiment. So we measured olfactory characteristics for 33 msec of pulse ejection. “Response time” and “Duration of scent perception” were measured for pulse ejection. Further, we defined and measured the minimum ejection interval that subject could discriminate the two individually emitted pulses of scent, which we call “Separable detection threshold”. It is expected that scent presentation based on these results can match visual and audio information and better enhance the sense of reality.

Scent Presentation Technique of Pulse Ejection Synchronized with Breathing

Trials on the transmission of olfactory information together with audio/visual information are currently being conducted in the field of multimedia. However, continuous emission of odor creates problems of human adaptation to the lingering olfactory stimuli. To overcome such a problem we developed an olfactory display. This display has high emission control so that it can provide stable pulse emission of scents. We previously proposed two scent presentation methods synchronized with the user’s breathing pattern.The first method presents smell using a breath sensor. The second method presents smell using the ejection interval Δt,determined by the olfactory characteristics. The results of evaluation experiments and a questionnaire survey of users revealed the proposed methods provide a continuous sense of smell to the user, avoiding adaptation.

Scent presentation expressing two smells of different intensity simultaneously

Studies aiming at increasing the sense of reality by using scents with the picture medium, such as movies, have lately attracted much attention. In the picture medium, multiple objects with different scents are seen at the same time. Therefore, it is thought that by controlling the presentation of scent in accordance with the various scented objects seen, the viewer will experience a greater sense of realism. This realism will be further heightened by expressing through scent the distance relation between the multiple objects, that is, stronger scent for objects in the foreground and weaker scent for those in the background. However, this requires a technique that can not only present multiple scents which must be perceived simultaneously, but also express a distance relation between them. In this study, by using pulse ejection which emits scents for only very short periods of time and by measuring the olfactory characteristics for various scent presentation patterns by changing the relative amount and presentation frequency of the two scents, we succeeded in enabling around 90 percent of users to sense the distance relation between two scents. The proposed method for presenting scent depicting multiple objects shown simultaneously in the picture medium is expected to further enhance the sense of reality.

Real 3 communication and aromatic group computing: HCI and CSCW research at Canon Media Technology Lab.

INTRODUCTION AND BACKGROUND The Media Technology Laboratory is one of Canon Inc.ts corporate research labs. Originally called the Information Systems Research Center, the laboratory changed its name when it started research and development in information media, Now, approximately 70 research scientists and engineers, some of them managers, are in charge oi mearch and development mainly in HCI, CSCW, and other fields, Their ‘range of expertise covers AI, natural language understanding, computer vision and graphics, computer architecture, and system software (OS, database, etc.).

An Instruction Method for Displaying Trajectory of an Object in Remote Collaborative MR on the Basis of Changes in Relative Coordinates

We aim at developing a remote collaborative Mixed Reality(MR) system targeting the interaction between instructors and workers. In this system, both instructors and workers have their own real objects, and the instructor shows the worker how to operate the object by moving his own real object. Since the movement of the instructors object is displayed as a CG image of the trajectory overlapping the workers object in each operation, the worker can easily see the changes in the position of the instructors object and trace the movement along the trajectory. We created this model using the object coordinates instead of the world coordinates in order to be able to ignore the changes in the position of the users and other objects in the world coordinates. We implemented a prototype and used it to conduct validation experiments. The results indicated that worker operation time is shorter in the proposed model in comparison to conventional models, where each users orientation towards the object is different.

A communication model of scents mediated by sense-descriptive adjectives

Studies on olfactory communication using scents in addition to audio/visual information have recently become active. In a scent communication system, no representation scheme of scent information has been established due to lack of a basis for the smell. We introduced background scents that provide the ambience of image contents, and proposed a communication model for the background scents using sense-descriptive adjectives for association. The results of our experiments showed that the communication model is effective for enhancing the image contents.

ARGUS (video program) (abstract only): an active awareness system using computer-controlled multiple cameras

ARGUS is a prototype system which achieves awareness before starting face-to-face communication by utilizing multiple far-end controllable cameras on a broadband network. A user desiring awareness information can change camera directions and angular fields of view at will in order to get the desired image. We call this active control of cameras to aquire awareness information “active awareness.” Camera locations and viewing fields can be quickly grasped by camera icons on the Map Viewer, which shows the actual office layout. ARGUS has a merit to provide both wide area views and close, detailed views via the control of remote cameras. However, the privacy of people may be violated, for example, through extreme magnificaiton. Therefore, it is very important to carefully consider the camera placement and usage. We have modeled office enviroments and have introduced two camera types: private cameras and public cameras. The aim of the former type is to catch a personal view whereas the aim of the latter type is to monitor common spaces. To protect privacy, ARGUS has several levels of access restrictions for each of these camera types. In this video, we describe the policy and implementation of ARGUS from the viewpoints of user interface and privacy.