This paper proposes a new type of display system by which images come out in an open space according to the motion of users or devices. This is an application of the specialized projector with a function of pixel-by-pixel visible light communication, which the authors have previously developed. This projector permits image data as invisible high-speed flickering signals into the open space. In the space filled with the invisible image data, by scanning a specialized device, we can make the image show up. Based on this concept, this paper introduces two implementations: "SteganoScan" and "SteganoScan Orbs." SteganoScan is a kind of persistence of vision display. When a performer waves an array of LED module with photo-detectors quickly in the projection light, viewers can perceive floating 2D images caused by the effect of retinal afterimage. SteganoScan Orbs consists of tens of balls which glow like moving pixels and form an image.

Kinetic typography, which temporally modifies display attributes of text, is able to express implicative meaning, including emotion.In contrast to animation of geometrical figures and body motion of CG avatars, kinetic typography can convey specific messages by text itself. The effect of motion pattern on emotive expression, especially interaction effect between text and motion, has not been evaluated quantitatively. We thus conducted a basic investigation using elemental words of emotions and frequently-used motion patterns. The results showed that words and motion patterns can be each categorized into three groups and effects were varied according to a combination of a word group and a motion group.

This paper discusses a new concept named"device art."The recent decades has shown significant advance in the interactive artwork created in Japan.Those artworks employ innovative interface devices. Device art has different characteristics from western art; bottom-up creation process, playfulness, and commercialization. Traditional Japanese culture, such as tea ceremony or flower arrangement, uses sophisticated devices, which are roots of device art. This paper describes characteristics of device art as well as its methodologies for creation of works.

We have developed a mixed reality (MR) painting system called the MR-based Artistic Interactive (MAI) Painting Expert, which simulates painting of physical objects in the real world. This system includes an interaction device called MAI Painting Brush, which has the characteristics of an actual paintbrush, and a painting model developed using this device. However, this system could not give the feeling of painting on virtual objects in an MR space. Therefore, we proposed and developed mechanisms that simulate the effect of touch and move when a brush device is used with a virtual painting target. In this paper, four mechanisms were proposed; the bend, slider, belt, and flap mechanisms, to provide the feeling of touching and painting a virtual object with the brush device, using visual and tactile feedback. These mechanisms were implemented and tested as prototypes, and a new brush device named MAI Painting Brush++ was designed and developed based on the result.

In recent years, researches on a very human-like android have became popular. The main purposes of past android researches were investigating: (1) how people treat very human-like android and (2) whether it is possible to replace existing communication media such as telephone or TV conference system, by androids as a communication medium. We found that androids have advantages compared to humans in a specific context.For example, the android drama that is one of a collaboration project with the artist, visitors reported that the android impress them especially when the android was reading a poem in the drama. We, therefore, did the experiment to investigate the advantages of the android compared to humans in the context of poem reading. An experiment was conducted to illustrate influences of the android poem reading.Participants were listened to the poem that was read by three kinds of poem reading agents: the android, the model of the android and the box. Experiment results showed that an entrainment for the poem gained the most highly score under the android condition. It indicated that the android have an advantage for communicating the meaning of the poem.

Recently, equipment for home use and online services requiring specific technical knowledge to operate have increased. However, an average home user cannot be expected to have in advance all the required knowledge for proper operation. Therefore, we propose an active help system which provides only required information to users according to a user ' s acquisition knowledge level, which is inferred using a separate proposed method. More concretely, we propose a method for expressing the acquisition level of each knowledge item of a user with a quantitative index, referred to as the "knowledge fixing degree " , and inferring this index formally using the number of times a knowledge item is referenced and the number of times an appropriate answer is given requiring that knowledge. We show experimentally the effectiveness of the proposed method in network operation in a proto-type implementation. To generalize this method, we need to verify the effectiveness of the system under more realistic conditions in the future.

The typical input for the mirror interface is the user's life-sized video stream. Because of this, the reachable area depends on the user's arm length and users sometimes have to intentionally move forward to select objects. In order to avoid such movement, we propose two new methods for selecting the target object with the mirror interface. The first is a pointing stick method which uses a virtual pointing stick that functions in a similar manner to a physical pointing stick.The second is a dragging method which allows users to drag targets towards the user's palm. The experimental results indicate that both of new methods are able to expand the reachable area compared to the traditional method, so that users do not have to move to select objects. Furthermore, it is also found that the combinatorial input of the traditional method and the pointing stick method is suitable for the mirror interface.

Awareness plays an important role in facilitating effective interaction in distributed environments. In text-chat interaction, it is difficult for participants to read all utterance texts and understand who is talking to whom about what. For progressing the effective interaction, it is required that participants can be aware of the information about utterers and the utterance targets. Additionally, they need to focus on the utterances which are useful for understanding the topic. This paper proposes an interactive interface for managing awareness of utterances to support successful interaction among participants. Utterance texts displayed in the interface move from utterers to their targets intuitively. To represent focusing utterances emphatically in the interface, the utterances are timely detected according to utterance types and understanding degree of each keyword. Experimental result showed that our detection method could distinguish more than 60% of focusing utterances correctly. In addition, participants could intuitively grasp the conversation flow and detected focusing utterances by observing the interface.

This paper describes a series of experiments to evaluate the readability of short stories on electronic devices focusing on the operability of page-turning. We compared reading speed from an iPad and a Kindle against that from a paper book and a notebook PC. As a result, although no significant differences were found among the four media without accompanying page-turning, reading from a paper book was significantly faster than a Kindle and the notebook PC when they turned pages. We also investigated usability of different user interactions for turning pages on iPad. Speed and cognitive load of reading was compared between three conditions: an iPad in which page-turning interaction was restricted to swiping (swiping condition), an iPad in which page-turning interaction was restricted to tapping (tapping condition), and a paper book. As a result, we could not find any significant differences in speed and cognitive load in the case of reading without turning pages. However, in the case of reading with page-turns, reading from the paper book induced less cognitive load than reading from both conditions of the iPad, and swiping induced less cognitive load than tapping. These results suggest that page-turning on paper books is performed almost unselfconsciously, and swiping of iPad is an effective interaction technique, as well as it provides an attractive user interface. Moreover, we discuss methodologies to improve electronic reading devices.

A research goal on usability evaluation methods is to achieve reliable usability "scoring". Toward the goal, the authors have developed quantitative usability evaluation method that uses a usability checklist and a checking support tool. Our checklist consists of 114 checking items with more clearly and precisely defined checking steps and judgment criteria than existing checklists. Such clear and precise definitions enable checklist-based evaluations more reliable and less independent of differences among evaluators. Effectiveness of our checklist has been evaluated by experiments with novice evaluators. Besides, the authors have evaluated effectiveness of our checking support tool by an experiment. This article reports our evaluation method and support tool, and the method and results of the experiments.