This paper presents NAMIDA, social interface consisting of three robots that help the driver to navigate by considering him as a bystander in a multi-party conversation between the agents. Through this model the cognitive workload of the driver can be reduced compared to the conventional one-to-one communication based approach that directly addresses the driver. We set up an experiment to compare the multiparty conversation based NAMIDA (MPCN) and the one-to-one conversation based NAMIDA (OOCN) to explore the effects of both methods on the cognitive workload factors and attention behaviours of the participants. The results showed that the MPCN system yields reduction in certain workload factors which can help the driver in increasing his attention on the road.

It is still an enormous challenge within the HRI community to make a significant contribution to the development of a robot's utterance generation mechanism. How does one actually go about contributing and predicting the future of robot utterance generation? Since, our motivation to propose a robot's utterance generation approach by utilizing both addressivity and hearership. Novel platform of Talking-Ally is capable of producing an utterance (toward addressivity) by utilizing the state of the hearer's behaviors (eye-gaze information) to persuade the user (states of hearership) through dynamic interaction. Moreover, the robot has the potential to manipulate modality, turn-initial, and entrust behaviors to increase the liveliness of conversations, which are facilitated by shifting the direction of the conversation and maintaining the hearer's engagement in the conversation. Our experiment focuses on evaluating how interactive users engage with an utterance generation approach (performance) and the persuasive power of robot's communication within dynamic interactions.

We developed soccer ball-shaped interactive artifacts (COLUMN) consisting of eight modules that are connected to twelve servomotors. In the interaction, COLUMN becomes a social mediator to prompt the connectivity of the users. We claim that the above process generates a variety of user's interpersonal coordinates and these can be mapped to the shape of COLUMN's. Therefore, the motivation of our study is to extract the patterns of user's interpersonal coordination. It is also interesting to explore how these patterns lead to imagine behavior for an artifact which has different degree of freedom of its body than the existing.

The advantage of the proposed approach is that users can invent their own communication protocol (based on the knock patterns) to communicate with the creature. This approach is a novel concept to establish future human-robot communication protocols within many contexts or human centric applications. However, through symbolic communication (knock-based communication) a human is able to convey to the robot to adapt and communicate with their personalized communication protocol.

We investigate the potentiality of sociable spotlights (sociable artifacts) to change the participants' roles by providing turn-yielding signals at the Transition Relevance Place (TRP) by changing the artifacts' behaviors (the effect of colors combined with the artifact behaviors as a bunch) that moot the state of the conversation.

In this study we develop Core Less Unformed Machine (COLUMN) as novel & transformable robotic platform to explore how visually mediated information is useful to coordinate (interpersonal coordination) to establish connectedness of three participants to obtain COLUMN's behaviors (transformable rolling motions).

In this paper, we propose a MAWARI-based social interface as an interactive social medium to broadcast the information to users. The interface consists of three creatures (MAWARIs) and is designed with minimalism designing concepts. MAWARI is a small scale robot that has only body gestures to express (or interact) its attractive social cues. The participant's role is reduced to that of a bystander when the interface is of a passive social state. In this context, the user does not need to participate in the conversation but still gains information without exerting effort (i.e., less conversational workload).

We developed a creatures-based social dining table that can communicate through a knocking sound, which in Japanese is pronounced as "KonKon." Our main focus was to create a minimal number of cues for a proto-communication by establishing social interactions between a creature and a human. In particular, humans used the "KonKon" interface to communicate with a creature to demonstrate the social behaviors necessary to adapt to a person's intentions. The creature used a mutual adaptation model for achieving a more ideal adaptation during the interactions. In the experimental results, we discuss the concept of the creature and indicate the effectiveness of the communication-protocol on the "KonKon" interface for mutual adaptation.

We developed a Sociable Trash Box (STB) as a children-assisted robot able to collect the trash in order to convey its intentional stance to children. The STB is capable of engaging manifold affiliation behaviors to build a social rapport with children by collecting the trash around their environment. In particular, the STB is a child-dependent robot that walks alone in a public space for tracing humans and trash for the purpose of collecting the trash. The robot is incapable of collecting the trash by itself, and it engages by using interactive behaviors and vocalizations to make a social coupling with children based on the robot's anticipation to accomplish its goal. The present experiment investigates how STB behaviors are effective in conveying intentions to evoke children's social interactions and to assist in collecting the trash in their environment.

A sociable PC (SPC) is capable of engaging and interacting with social cues while users use it. SPC is a kind of artifact which is capable of coordinating and synchronizing its behaviors based on user attention and information on a display. In particular, SPC can exhibit behaviors to induce a trust through social rapport with the user while responding to the user's behaviors and activities on a PC. We used the concept of a minimalism designing mechanism to invent the SPC. The SPC appearance is much like soft "Tofu," so the user can touch and sense it. The SPS can also provide feedback to the user using attractive social cues such as shaking its body, displaying an attractive motion and joint attention with the user, etc.

The STB is capable of engaging manifold affiliation behaviors to build a social rapport toward the goal of collecting trash around an environment. In particular, STB is a child-dependent robot that walks alone in a public space for the purpose of tracing humans and trash and to collect the trash. In a crowded space, STBs move toward the trash by engaging with an attractive twisting motion (behaviors) and vocal interaction to convey STB's intention to children. Our STB robot is incapable of collecting the trash by itself. In this sense, children have to infer a robot's intentional stance or expectation for interaction with the STB. To collect trash while creating social rapport with children is a novel concept. The STB engages with twisting and bowing motions when children put trash into an STB container.

In this paper, a creature-based futuristic sociable dining table is proposed that communicates through a knock on a table. The main objective was to create a minimal number of cues for protocol communication for the purpose of establishing social interactions between a creature and a human. The creature uses a knock (within a specific period) with a Temporal Difference (TD) learning model to learn and to adapt to a user's intentions, and the creature moves a different direction (right, left, straight, and back) on a table based on the user's knock. By ascertaining the creatures' behaviors, the user can convey his/her intentions by a number of knocking sounds while looking at the creature adaption. The creature is capable of doing mutually adaptive learning through a TD model in real time to satisfy the user's goal or intentions. In the context of the experimental results, we discuss the concept of creatures and indicate the effectiveness of the communication-protocol with the “knocking” interface for reciprocal adaptations.

In this paper, we propose a therapeutic-assisted robot for children with autism to ameliorate their skill of joint attention. The robot conducts a goal-directed based interaction to establish engagement between the child and robot in order to establish a beneficial learning environment for autistic children. An unsupervised Mixture Gaussian-based cluster method is proposed to detect the child's intention in real time to process the goal-directed task smoothly. The novelty of this approach is that does not require the use of any training data or a trained model to detect the child's intention. Our autonomous robotic system is tested with several autistic children at a School for the Disabled in Nagoya, Japan. The results of the initial interaction showed that the children enjoyed interaction with and feedback from the robot, which confirmed that the robot can be used as mediator or an object of joint attention. The unsupervised approach was able to detect the children's intention at every time segment to process the goal-directed task with a higher accuracy rate. The results of the goal-directed task showed that the proposed interaction was highly effective in enhancing their joint attention. Since most of the children attempted to imitate the robot's gestural behaviors and used a variety of learning patterns to attend to the robot's fingered object in the environment to obtain joint attention with robot.

The paradigm of animated pedagogical agent has been developed in recent years. These lifelike autonomous characters inhabit leaning environments and make rich, face-to-face learning interactions possible. Children seem to develop an immediate and deep affinity to such characters: the agents abilities open up new possibilities, e.g. agents can demonstrate complex tasks, employ locomotion and gestures, focus children's attention on the most important aspects of a task at hand, and convey emotional responses to the game. This create a comfortable environment with the ability to both engage and motivate children. The pedagogical agent uses emotion, intensity of emotion, and game states in order to create rules for adapting the system according to child behavior. Results showed that an affective gesture recognition model was able to recognize a child's emotion with a considerably high rate of over 82.5%, and the social agent (estimate intensity of emotion) had a strong relationship with observers' feedback, except in low intensity levels.

Conveyance and recognition of human emotion and affective expression is influenced by many factors, including culture. Within the user modeling field, it has become increasingly necessary to understand the role affect can play in personalizing interactive interfaces using embodied animated agents. However, little research within the computer science field aims at understanding cultural differences within this vein. Therefore, we conducted a study to evaluate if differences exist in the way various cultures perceive emotion from body posture. We used static posture images of affectively expressive avatars to conduct recognition experiments with subjects from three cultures. After analyzing the subjects' judgments using multivariate analysis, we grounded the identified differences into a set of low-level posture features. We then used Mixture Discriminant Analysis (MDA) and an unsupervised expectation maximization (EM) model to build separate cultural models for affective posture recognition. Our results could prove useful to aide designers in creating more effective affective avatars.

Today, creating systems that are capable of interacting naturally and efficiently with humans on many levels is essential. One step toward achieving this is the recognition of emotion from whole body postures of human partners. Currently, little research in this area exists in computer science. Therefore, our aim is to identify and measure the saliency of posture features that play a role in affective expression. As a case-study, we collected affective gestures from human subjects using a motion capture system. We first described these gestures with spatial features. Through standard statistical techniques, we verified that there was a statistically significant correlation between the emotion intended by the acting subjects, and the emotion perceived by the observers. We examined the use of discriminant analysis to measure the saliency of the proposed set of posture features in discriminating between 4 basic emotions: angry, fear, happy, and sad. Our results show that the set of features discriminates well between emotions, and also provides evidence about the strong overlap between descriptors in both acting and observing activities.

Abstract One of the challenging issues in affective computing is to give a machine the ability to recognize the mood of a person. Efforts in that direction have mainly focused on facial and oral cues. Gestures have been recently considered as well, but with less success. Our aim is to fill this gap by identifying and measuring the saliency of posture features that play a role in affective expression. As a case study, we collected affective gestures from human subjects using a motion capture system. We first described these gestures with spatial features, as suggested in studies on dance. Through standard statistical techniques, we verified that there was a statistically significant correlation between the emotion intended by the acting subjects, and the emotion perceived by the observers. We used Discriminant Analysis to build affective posture predictive models and to measure the saliency of the proposed set of posture features in discriminating between 4 basic emotional states: angry, fear, happy, and sad. An information theoretic characterization of the models shows that the set of features discriminates well between emotions, and also that the models built over‐perform the human observers. Copyright © 2004 John Wiley & Sons, Ltd.