In this paper, we introduce various devices and toolkits for developing ubiquitous interfaces effectively. Moreover, we propose novel middleware, the “MobiServer”, which helps users treat multiple devices and toolkits easily. Finally, we show some research projects developed with the MobiServer.
To supplement existing forms of communication such as telephone and e-mail, this research proposes a new method of communicating “awareness” between people who are separated by long distances. In this paper, we investigate cases where coincidences in daily behaviors lead to casual conversation and thus intimacy and togetherness. We propose a new method of communicating these “happy coincidences” between a pair of remotely located houses. By equipping furniture and appliances such as doors, sofas, refrigerators and televisions with sensors, we developed a system where these items are connected to remote equivalents and their near simultaneous use is communicated.
We have developed a program called “Sequential Graphics.” This program lets us simulate the experience of drawing. It enables us to draw pictures using broad strokes. The program's canvas has a looped time line, and therefore, images can be played back repeatedly. Users can add strokes at any point during the sequential playback. Users can also add strokes in real time while viewing the playback. This program records the motion of a painter's hand and can present the sensation of drawing to the user. Using this program, a user can create a “moving-still picture” that simultaneously exhibits the characteristics of both a still picture as well as an animation.
PhotoelasticTouch is a novel tabletop system designed to intuitively facilitate touch-based interaction via real objects made from transparent elastic material. The system utilizes vision-based recognition techniques and the photo-elastic properties of the transparent rubber to recognize deformed regions of the elastic material. Our system works with elastic materials over a wide variety of shapes and does not require any explicit visual markers. Compared to traditional Interactive Surfaces, our 2.5 dimensional interface system enables direct touch interaction and soft tactile feedback. In this paper we present our force sensing technique using photoelasticity and describe the implementation of our prototype system. We also present two practical applications of PhotoelasticTouch, a force-sensitive touchpanel and a paint application.
In this paper we propose an interaction technique using onomatopoeia. Although onomatopoeia, imitative words such as ‘Kira Kira’ and ‘Moko Moko’, is frequently used in ordinary conversation, many articles and comics, it is rarely used for human-computer interaction. We consider that onomatopoeia helps users to tell emotions and sensuous information to a computer more effectively. We developed a prototype of painting system called Onomatopen, which enables a user to change brushes by saying onomatopoeia. For example, when the user draws a line while saying ‘Giza Giza’, a jaggy line is drawn. As a result of our user test, we found that users can easily understand the usage and enjoy drawing more than conventional painting software.
Fingering on playing contrabass has significant effects on its musicality. Therefore, it is very important for contrabass players to learn correct fingerings. Especially, it is effective to share real-time fingering information of professional performances. In response to these requirements, the goal of our study is to construct a real-time fingering detection system for contrabass players. Our system detects the fingering in real-time by integrating a simple camera-based image processing and musical rules on contrabass fingerings. The evaluation results confirmed 87% accuracy in fingering detection. Additionally, we implemented two applications; a self-study application and a remote lesson application. They show the effectiveness of detecting fingerings in real-time.
Many amateur photographers feel difficulty to shoot natural faces of people. While many subjects become tense when caught on camera, other camera-ready subjects -familiar with be taken photos- often create stage face. This paper proposes a novel interaction technique, the EyeCatcher, which helps photographers take natural and various faces, by keeping eyes of subjects to the camera without be nervous with be taken photos. We develop a prototype system of the EyeCatcher and verify the effectiveness through user tests and discussion.
We developed a solver of the satisfiability problems of quantifier free first order formulas containing stringrelated functions and predicates. In this paper, we describe the design and the implementation of the solver. Our method is based on Makanin/Jaffar's word unification algorithm, but is extended to be able to handle constraints of assignable lengths of variables and possible occurrence counts of letters in the variables. To implement this, we propose a hooking calculation method (HC method) to combine a word unification solver and a QFP solver. The soundness of the combination is kept by ‘basic relation’ of strings and their attributes. Jaffar's algorithm is known to be complicated, but we show our solver can solve relatively small-sized problems, which usually appear in software verification situations, in practical time.