This paper describes a development of a presentation robot system to interact with a small group and a demonstration result of a presentation by a robot. Our research target is that a robot interaction encourages participation of participants for a robot presentation, and interests for a long time. In this paper, we propose a robot interaction using cards to reflect the opinion of a small group of participants. We developed an integrated presentation robot system using a humanoid robot, PC and camera. By using the presentation robot system, we demonstrate a robot presentation for high-school students and freshman of university students. We verified that the proposed method encourages participation of participants for a robot presentation, and interests participants in the robot presentation.

In recognizing an object, humans are thought to rely on how that object is used; that is, on the interaction between a person and that object. This study develops a method that extracts interrelationships between people and objects (actions humans take on an object) as the object’s attribute characteristic and use that attribute information as a characteristic of object recognition. This paper proposes a method that represents the shape of an object and human movements – extracted with an RGB – D sensor–as attributes of the object.

In recent years the personal robots for the purpose of contacting with human such as the entertainment robot and nursing care robot have been increasing. By this reason, the ability to transfer some informations in the communication to human has become necessary for these robots. As the medium to communicate smoothly, it is used a lot of non-verbal (non-language) communication like facial expressions and gestures.

In this research, by focusing on the body motion, we have proposed an interactive emotion communication (IEC) as a method which robot and human attempt to transfer information by the body motion with emotion. We aim to perform interactive communication by the body motion with emotion to increase the interpersonal affinity to the robot by amplifying and suppressing the human emotion. In the previous research, it has been found it is necessary to have not only sympathetic but also not bored emotional behavior response model to enhance interpersonal affinity.

A method for the emotion generation using Markovian emotional model is proposed in this paper. Using the Markovian emotional model is capable of modeling the state transition that takes into account the emotional preference reactions and the internal state of recognized emotion. In this research, an emotion generation model with the interaction sustained human interest is constructed. Performing the interaction experiment by the robot equipped with this model, we confirmed that the interpersonal affinity is improved in a humanoid robot with Markovian emotional model.

With the development of the Internet, broadcasting of sports watching has also increased, and annotation display on this is an important factor for watching support. This paper proposes to use the psychological warfare of players as watching support information in watching sports. The psychological warfare between players is inferred based on the expectation of viewers and given. The viewer enjoys games by seeing the compensation of the strategies of the players, but at this time he felt the subjective interestingness by anticipating the psychology of the player such as “I worked hard” and “I am bearish.” We constructed a system SBF which calculates the psychological state of the player from the subjective situation of the viewer and the objective information of the game by fuzzy reasoning and presents the illustration icon via the Russell space. We conducted a subject experiment on baseball game watching and clarified that it is effective for improving empathy degree, coincidence, and fun with the system.

Recently, more educational-support robots, which support learning, are paid attention to. However, the problem of these robots is that a user loses his/her interest in them. To solve this problem, the previous study proposed a sympathy expression method for educational-support robots to enable them to express sympathetic emotions when interacting with learners. The previous study suggested the robots which uses the sympathy expression method prompt learners to reduce losing his/her interest with robots in collaborative learning. Although the sympathy expression method expresses the emotions by changing robot’s facial expression, the existing study of Human-Robot-Interaction suggested robot’s body movement was effective for interaction with humans. We think that the sympathy expression method that expresses the emotions by robot’s body movement and it’s facial expression can give more good impression to learners than the sympathy expression method that expresses the emotions by only robot’s facial expression. Thus, this paper investigates the impression effect of the robot using the sympathy expression method that expresses the emotions by robot’s body movement and it’s facial expression.

The purpose of this research is to realize smooth and natural mutual avoidance of human and robot by communicating the intention of avoiding robot to people. Assuming a situation where a person and a robot encounters at a doorway or corridor of a building or a situation, where the robot cannot perform a lower body movement. We aim to construct a system that can communicate and actively guide people in a direction not to collide. By this proposed method, a risk of collision between human and robot decreases, it is thought that we can avoid avoiding behaviors of each other and minimize mutual avoidance distance. In order to verify whether upper body movement is effective as a transmission means of the robot’s avoidance direction intention, we posted it to the subject by three presentation methods with the position of both arms and the orientation of the chest as motion elements. As a result of the experiments, the result that upper body motion affected a determination of avoidance direction was obtained.

In recent years, many support systems intended for elderly and disabled people have been developed. Among them, the study on the interface utilizing human gaze has been attracting attention. Human vision contains a lot of important information. Some studies that estimate a next action of operator intended to perform by using the visual information have been reported. As an example, the visual saliency map model is one of the means using to estimate the intention of the human gaze. This model enables to estimate the place directed human attention from the image. Also, this model has an advantage that uses only the image information. As the research dealing with the visual saliency map model, the study suppressing the changing value of saliency map by the movement of operator’s head and the study controlling the weight of saliency map used human knowledge have been reported. In this research, we focus on the combination of visual saliency map and gaze information, and extract a behavioral intention estimation map from the landscape and the movement of human eyes. A gaze instruction system using panoramic expansion image for controlling the locomotion of omni-directional wheelchair is proposed in this paper. In this system, the human intention by various input information is estimated by using fuzzy reasoning. To evaluate the effectiveness of this method, we carried out the comparison experiments for gaze instruction system by using only movement of human eyes.

Autonomous mobile robots that guide people in dynamic environments need to carry out the following two tasks; to move safely avoiding obstacles, and to keep the relative positional relation to people without losing sight of them. In this study, we propose a robot that avoids both static and dynamic obstacles (such as oncoming pedestrians) safely while guiding multiple people. The robot tracks the target people robustly by using both laser range finders and an omnidirectional stereo camera. Furthermore, the robot distinguishes between static and dynamic obstacles and avoids them safely by giving them different repulsive potential accordingly. By combining these two functions, the robot can achieve appropriate guidance.

In recent years, the research and development of human symbiotic robots have advanced remarkably. Many studies of human symbiotic robots often adopt interaction designs in which robots understand speech and feelings of people. In this research, we focus on the interaction design where robots transmit information to people, and people understand and respond it. For constructing the interaction, we utilize toddler-like utterance by young children. There have been studies to investigate the acceptable range of toddler-like utterance using a baby-type robot, but general knowledge has not been obtained at present. In this paper, we investigate which kind of robots can utilize the toddler-like utterance of a two year old child without discomfort by using 24 types of communication robots. Furthermore, we show the general knowledge using the obtained results.

In recent years, non-task-oriented dialogue system which aims at natural chatting of computer with human has received much attention. In order to increase the satisfaction of users, it is effective to generate utterances which can attract users’ interests in addition to performing natural conversation. Therefore, in this paper, we propose a method to acquire tweets that represents “ARUARU NETA” (typical story for many people) from Twitter and use them as system’s utterances. “ARUARU NETA” is one of the techniques in entertainment inviting laugh by getting empathy of the audience. As a result of the evaluation experiment of one round trip conversation, it was confirmed that the utterances of the proposed method were the most preferable compared with the chat dialogue API and normal tweets.

In recent years, for the development of agents capable of communicating as partner of humans, research of to realize a more natural conversation between humans and agents is drawing attention. On the other hand, few researches attempt to systematize the effect of agent’s linguistic behavior on the impression and relationship building. Humans are selects linguistic consideration in conversation taking psychological distance and social position into consideration in order to build good relationships with opponents. Based on the idea of Reeves’s media equation and Dennett’s intentional stance, it seems that there is effective linguistic consideration for relationship building also in human-agent conversation. There, in this research, we aim to verify the effect of agent’s linguistic behavior on the impression and relationship building. As an approach, based on the politeness theory, we subjectively evaluate the impression when the agent used jokes that can be expected to have an effect to attribute intention to users among the consideration to actively reduce the distance to the opponent (Positive politeness strategy), and consideration for keeping it (Negative politeness strategy), by participant experiments.

Recent interactive design tools can provide many kinds of parameters for users. The user repeats to select and settle the parameters again and again according to his/her image which he/she wants to create. The process of the design isn’t an easy task for beginners who aren’t familiar to the design tool. We can deal the process of the iterative design work as well as interactive optimization. Our study aims to reduce the load of the iterative optimization. We introduce onomatopoeias to set proper initial values of the optimization. This paper reports the experimental result of our proposed method which adopts initial values setting with onomatopoeias. The paper discusses the experimental result and concludes that the proposed method might be useful for the beginners to reduce the load of the iterative design process.

In this paper, we propose a method for morphing input characters by using onomatopoeias. This method morphs input characters by using impression value of onomatopoeias proposed by Akiyama et al., and expresses wet and dry effects by using pen pressure and speed. Using this method, users can express characters intuitively, because they just input onomatopoeias as their imagination.