外文翻譯--配有鼻子追蹤器的氣味生成器（英文）

1、Projection-Based Olfactory Display with Nose Tracking Yasuyuki Yanagida1, Shinjiro Kawato1, Haruo Noma1, Akira Tomono2,1, and Nobuji Tetsutani11 ATR Media Information Science Laboratories {yanagida | ska

2、wato | noma | tetsutani}@atr.jp2 Dept. of Information Media Technology Tokai University tomono@keyaki.cc.u- tokai.ac.jpAbstractMost attempts to realize an olfactory display have involved capturing an

3、d synthesizing the odor, processes that still pose many challenging problems. These difficul- ties are mainly due to the mechanism of human olfaction, in which a set of so- called “ primary o

4、dors”has not been found. Instead, we focus on spatio- temporal control of odor rather than synthesizing odor itself. Many existing interactive olfactory displays simply diffuse the scent into t

5、he air, which does not provide the ability of spatio- temporal control of olfaction. Recently, however, several researchers have developed olfactory displays that inject scented air under the n

6、ose through tubes. On the analogy of visual displays, these systems correspond to head- mounted displays (HMD). These yield a solid way to achieve spatio- temporal control of olfactory space,

7、but they require the user to wear something on his or her face. Here, we propose an unencumbering olfactory display that does not require the user to attach anything on the face. It works

8、 by projecting a clump of scented air from a location near the user’ s nose through free space. We also aim to display a scent to the restricted space around a specific user’ s nose, rat

9、her than scattering scented air by simply diffusing it into the atmosphere. To implement this concept, we used an “ air cannon”that generates toroidal vortices of the scented air. We conducte

10、d a preliminary experiment to examine this method’ s ability to display scent to a restricted space. The results show that we could successfully display incense to the target user. Next, we

11、 constructed prototype systems. We could successfully bring the scented air to a specific user by tracking the nose position of the user and controlling the orientation of the air cannon to the

12、 user’ s nose. 1. Introduction We perceive the surrounding environment by using our five senses. VR systems so far have been developed to cover visual, auditory, and haptic sensations, so it is

13、 a natural progression to incorporate olfaction into VR systems. However, there are various reasons that olfac- tion has been left in the backcountry of the VR research field. For example: (1

14、) Olfaction is activated by chemical stimuli. This is very different from visual, auditory, and haptic sen- sations, which are activated by physical stimuli. (2) A set of “ primary odors,”i.e.,

15、a small number of bases to represent arbitrary smells, has not been found. Nevertheless, incorporating olfactory interfaces in VR systems could be effective for achieving a high level of prese

16、nce [1]. Dinh et al. [2] found that an olfactory cue, along with auditory and tactile cues, could increase the sense of presence in virtual environments. Olfactory sensation itself has a lo

17、ng history of research [3], including attempts to find a set of “ primary odors.”Amoore [4] categorized 7 primary odors, but later he extended this number to 20– 30. Buck and Axel [5] reported at

18、 least 100 kinds of receptive proteins, based on the theory of odorant receptor proteins [6]. The number of receptor proteins is theoretically estimated to exceed 1000. Therefore, we can har

19、dly use the same strategy adopted for visual display, by which we synthesize any color by mixing the 3 primary colors (red, green and blue). For tactile displays, recent progress has been

20、based on the idea of selectively stimulating several kinds of mechano- receptors [7– 9]. In these visual or tactile displays, virtual- ity is achieved by the characteristics of receptors that c

21、an detect only a part of the target physical phenomena. If we wanted to make an olfactory display that could provide arbitrary scent based on a similar strategy, we would have to somehow

22、control thousands of odorants. We are not attempting to solve such an immense prob- lem; our current focus is not on smell recording or syn- thesis. Instead, we focus on the spatio- temporal

23、control of odor, assuming that the odorant itself is ready to use. There are several subtasks of recording and displaying environments by a certain sensory modality: sensing, coding, storage

24、or transfer, decoding or rendering, and display (Figure 1). Our interest is on the final stage of display: the olfactory counterpart of HMD, CAVE [10], autostereoscopic displays [11], and so

25、 on. In contrast to visual displays, not so many olfactory “ displays”have been developed so far. Many scent blenders simply diffuse the rendered scent into the air, and these are regarded as coun

26、terparts to (controllable) illumination. Some researchers have developed olfactory displays that transfer the scented air through tubes to the nose, to 43IEEE Virtual Reality 2004 March 27- 31,

27、Chicago, IL USA 0- 7803- 8415- 6/04/$20.00©2004 IEEE. Proceedings of the 2004 Virtual Reality (VR’04) 1087-8270/04 $ 20.00 IEEE ultimate autostereoscopic visual display. Our approach is to develop a n

28、ew kind of spatio- temporally controllable olfactory display with fewer encumbrances. Hence, our goal is to develop an olfactory display with the following characteristics: ? Unencumbering. Users

29、do not need to wear any de- vices or glasses. This is not only for users’conven- ience but also for application to bidirectional tele- communications. ? Localized. Scent should be perceived only

30、 within a limited range of space and at a certain time. By lo- calizing the scent, we can display different smells to multiple users. Also, we can significantly reduce the amount of (often

31、 expensive) odorant compared with simply diffusing the scent into an entire room, as the volume of the air sustaining the odorant is so small. The system can thus be relatively cost- effect

32、ive. In addition, this small clump of scented air can easily dissipate in a short time, which enables us to wield short- term control of the olfactory effect. To achieve an olfactory display

33、 with these features, we need to deliver a clump of scented air from a position near the nose through free space. We have explored the possibility of using an “ air cannon”for this purpos

34、e. Here, we overview the entire system design. How we use the air cannon is described later. The entire system (Figure 2) is composed of the fol- lowing components: ? Nose tracker ? Air cann

35、on platform ? Air cannon ? Scent generator First, we have to detect the position of the user’ s nose. For this purpose, popular head tracking technologies such as magnetic trackers or mechanic

36、al trackers can be used. Here, however, we again prefer methods that do not require the user to wear anything; therefore, computer- vision- based face tracking would be the most suitable appro

37、ach. Once the location of a part of the head/face is measured, it is easy to calculate the position of the nose. Next, the platform on which the air cannon is mounted is controlled so t

38、hat we can aim at the user’ s nose. Two degrees- of- freedom angles (azimuth and elevation) are controlled to determine the direction of the air cannon. It is not necessary for the scent ge

39、nerator to fill the chamber with scented air—it only has to spray the scent just before the air cannon launches a clump of air. Thus we can send a different scent with each launch and c

40、over multiple users with a single air cannon. An air cannon (also known as a vortex cannon) is a chamber with a circular aperture, and it is very popular in science demonstrations for chi

41、ldren. The simplest way to make an air cannon is to use a cardboard box, cutting out a hole and sealing the seams with packing tape. If we use a box whose dimensions are 30 cm by 20

42、cm by 20 cm with a hole of approximately 5 cm in diameter and hit it hard, a clump of air reaches 5- 10 m as if it were an invisible bullet. If we fill the box with smoke and push

43、it more gently, we observe a smoke ring moving smoothly forward. The speed of the smoke ring is approximately 50 cm to several meters per second. This ring demonstrates a toroidal vortex (

44、vortex ring) generated by the air cannon, and it shows that the vortex can carry particles that exist around the aperture when the air cannon launches the air. The schematic of an air can

45、non and a toroidal vortex is shown in Figure 3. It is said that this vortex ring occurs because of the difference in velocity at the edge (slow) and the center (fast) of the aperture. Th

46、e pressure at the center of the vortex (ring shape) is kept low so that the vortex keeps its shape for a while. The size of the vortex depends on the aperture size, while the speed and

47、 reaching distance of the vortex path depends on the velocity profile of the pushing TimeVortex ringPush Circular apertureSudden volume decreaseVelocity distributionAir extrusionFigure 3. An air cannon gene

48、rating a vortex ring.Tracking nose directionAiming at the noseAir cannonDelivering scented air by vortex rings CameraPlatformFigure 2. Concept of projection-based olfactory display. 45 Proceedings of the 2004 Virtual R