We propose a technique called voice augmented manipulation (VAM) for augmenting user operations in a mobile environment. This technique augments user interactions on mobile devices, such as finger gestures and button pressing, with voice. For example, when a user makes a finger gesture on a mobile phone and voices a sound into it, the operation will continue until the sound stops or another finger gesture is started. The VAM interface also provides a button-based interface, and the function connected to the button is augmented by voiced sounds. Two experiments verified the effectiveness of the VAM technique and showed that repeated finger gestures significantly decreased compared to current touch-input techniques suggesting that it is useful in supporting user control in a mobile environment.

In this paper, we propose an arranged text communication with the user's text messages and onomatopoeic texts of both the user's bodily motion and ambient sounds. The bodily motion is the user-subjective information, and the ambient sound is one of environmental information. The text communications have limitations on their expressions compared to other media such as sound, images or video. Instead of other modalities, the proposed method adopts onomatopoeic texts converted from the user's bodily motion and the environmental situation so that nonverbal expressions including the user's underconscious or uncontrolled information but without privacy information can be expressed in the text communication. We evaluated the different types of the expressive modality for pala-linguistic information in parallel with text communications. From the results, it is confirmed that onomatopoeic texts attached to the text communication can enrich communication information.

We propose a novel scheme of gesture interface that guides the user toward obtaining better performance and usability. The accuracy of gesture recognition is heavily affected by how the user makes postures and moves, as well as environmental conditions such as lighting. The usability of the gesture interface can potentially be improved by notifying the user of when and how better accuracy is obtained. For this purpose, we propose a method for estimating the performance of gesture recognition in its current condition, and a method for suggesting possible ways to improve performance to the user. In performance estimation, accuracy in the current condition is estimated based on supervised learning with a large number of samples and corresponding ground truths. If the estimated accuracy is insufficient, the module searches for better conditions that can be reached with the user's cooperation. If a good improvement is possible, the way to improve is communicated to the user in terms of visual feedback, which shows how to avoid or how to recover from the undesirable condition. In this way, users benefit, i.e., better accuracy and usability, by cooperating with the gesture interface.

A degenerated character input method used in the 1990s enabled users to enter a hiragana character by single-tap to the key where multiple Japanese hiragana were allocated. This method might remain ambiguities in hiragana strings since it actually inputs a numeral code, though this was a very convenient way because of its easy-to-learn and easy-to-use manipulation. Taking this advantage, we tried to develop a smartphone application to encourage the information retrieval illiterates, especially for elderly. By restricting the retrieval field, we tried to reduce the canditates derived from ambiguous input, for uses' easy selection. Key tiles were also designed larger than other existing digit designs on a touch screen. A Who's who database is formed from the Wikipedia open source by collecting personal names containing "yomigana" and their URLs. Converted numeral codes were added. An application "Shirabetai" was developed and installed on the Android smartphones. This application was demonstrated to the test subjects and opinion were collected. Operation times and error rates were also collected. About 70% of subjects recognized the usefulness of the proposed single-tap input method and the application. This method was also presented positive from performance evaluations for the elderly to assist input work in information retrieval.

For students who want to voluntarily choose a certain field, in the career education to increase their motivation, it is desirable that teacher considers the following two points: (1) In pre-education, assessing their psychological state's desirable and undesirable fluctuation, (2) In post-education, assessing its effect to their psychological state. Therefore, the purpose of our research is to develop a supporting method for (1) and (2) by using visualization's information between students' psychological states and their motivation of intention to choose a certain field. In this paper, our target is high school students who are applicants of science fields. From high school students (n=126), we collected the psychological test constructed by questionnaire about the psychological states of science fields and the degree of aspiration to choose science fields. We divided the obtained data of questionnaire into two data: training data (n=90) and test data (n=30). Using training data and Support Vector Machine, we visualized relationship between students' psychological state and motivation of intention to choose science fields. We also verified the classification accuracy using test data. As the results, we obtained 83% classification accuracy. Moreover, we proposed a supporting method for (1) and (2) by using the obtained visualization information.

In this research, we propose a communication support system that provides moderate relationship among third persons without a sense of duty of reply. Users communicate anonymously through a photograph. A user uploads a photograph with a short text. The photograph is sent to more than one user chosen at random. Reply of comments is optional so a receiver of a photograph can let it slide. In reply, users can use not only short text also reaction button such as ‘cute!’ or ‘don't mind!’ to reduce the effort of reply. From the result of an experiment, we cleared that the proposed system support the communication while encouraging a reply to not feel a sense of duty.

We propose a new method to analyze human relations and mental process through touch interface dynamically. There is a sociometry analysis that applies the fuzzy graph an approach to clearly express the measurement and evaluation of the structure of human relationships. The analysis conditions are not constant in the analysis of human relationships, and not in addition to performing the analysis by trial and error for that. Previously in the study, in the relationship such as between the groups, it has not been studied the individual relationships analysis in the groups. Similarly, it has never been also studied to analyze the human relationships dynamically. We propose a method to dynamically the human relationships by intuitive operation through the touch interface. In this paper, we describe our principle of the sociometry analysis that using the fuzzy graph, the our method through the touch interface, and their application to sociometry analysis in case of 30 people of human relationships.

In this study, we have developed a social robot that considers the concepts of" addressivity" and " dialogicality" and" hearership" which are considering hearer's attitude and responses; (1) recognize the eye gaze of the hearer by the eye mark recorder, and (2) state of the hearer to adjust the utterance strategy. As a result, the utterance organized with the hearer is original utterances. In the experiment, we conducted a survey to explore effectiveness of the utterance generation system (social robot). The utterance generated by the proposed approach was able to get a high commendation for most of the questions. It is suggested that, the utterance of the robot is considered to be equipped with" address". Because the robot was uttered while consciously aware of the eye gaze of the hearer and hearer thinks," I have been conscious as the hearer". We got the knowledge that utterance with " address" inspire " intentional stance" from hearer.

We developed a robot user interface where the robot controller's motions were reflected in the motions of certain characters appearing on a computer display. We then conducted a simple experiment to investigate the effectiveness of this interface system to see whether users felt that this controller was enjoyable. We also conducted several user experiments that investigated the effectiveness of the appearance and material of this interface.

Recently, the various systems that present directional information are studied and put into practical use. Many of these systems are intended to add sensory stimulus to users to present the directional information. For example, these systems present directional information, such as map information, voice guidance and others. In general, such methods are useful to navigate the user to certain destinations. However, by using such explicit presentation method, users' behavior is restricted by the presented information or the system. For example, in case of that user pays attention to their surround environment or is lost in meditation, the explicit presentation is not required. On the other hand, we propose a novel method to present directional information by decreasing sensory stimulus. Our method uses only environmental sound to present directional information. With the proposed method, we implement a prototype system. The device decreases the sensory stimulus by controlling the sound field of environmental sound. The result of user experiments shows that the information presented by the system is recognized only when the user explores it consciously.

When people are walking together with taking the hand of each other, they can sense the partner's thought and intention by sensing the patterns of grasping. Platform of Mako-no-Te is sign with one-handed robot with small-scale head-appearance as its body. Through this platform, our motivation to explore how those grasping patterns are effectively communicate and building relationships between human and robot. Through the experiment, we explored that subjective evaluation when robot and human takes the hand of each other. Through the questionnaire, we examine subjective impression about the rapport, effect and influences of the interaction, intention, closeness and etc. Results showed that through the variety of grasping patterns it's potential to establish the perceivable communication between the robot and human. In this paper, we describe our concept, design, mechanism of Mako-no-Te and results of the experiment.

Today, reducing CO2 emissions is the most important issue in the world. Then, smart homes have attracted attention as an attempt at household level. Equipping various appliances and a home energy management system to optimize energy consumption, smart home should have a unique livability. Therefore, we performed a research to clarify the structure of livability evaluation of smart home from the residents' viewpoints by such quantitative approach as interviews. That is, we performed interviews to a family living in a smart home more than a year, analyzed the results by applying the Modified Grounded Theory Approach, and clarified the structure.

Mobile phones and video game controllers using gesture recognition technologies enable easy and intuitive operations, such as those in drawing objects. Gesture recognition systems generally require several samples of training data before recognition takes place. However, recognition accuracy deteriorates as time passes since the trajectory of the gestures changes due to fatigue or forgetfulness. We investigated the change in gestures and found that the first several samples of gestures were not suitable for training data.Therefore, we propose two methods of finding appropriate data for training for long-term use. We confirmed that the proposed methods found better training data than the conventional method from the viewpoints of the number of data collected and recognition accuracy.

In order to improve road safety, automobile manufacturers have been developing several types of Advanced Driver Assistance Systems (ADAS) including collision warning, arouse attention and so on. This paper describes effects of collision warning and arousing attention on driver behaviour in different driving situations by using a high fidelity driving simulators. First it was investigated how car-following behaviour was influenced in response to different types of driver support systems. Next it was investigated effect of collision warning and arousing attention on car-following behaviour and its impact on driving safety when critical situations occur. The results showed that arousing attention may contribute to longer time headway during car-following depending on driving situations. Moreover there was a possibility that combining the use of collision warning and arousing attention may induce decreased collision risk compared to driving condition in which only collision warning is available. Potential implication of these results includes how to design driver support systems with arousing attention.

There has been a significant progress in wearable computing technology. However, the existing character input methods for wearable computers have some defects and are not practical. This paper proposes and evaluates two character input methods for wearable computers using a distance sensor at the wrist and five mechanical switches at the fingertips. The results of our evaluation show that the proposed methods are sufficiently fast and easy to learn for practical use. This paper also discusses reasons for user errors and limitations of the methods.