The phenomenon of perceiving lifelikeness in the movements of non-living objects is referred to as animacy perception. Clarifying its mechanisms may provide a foundation for designing agents capable of more natural and intuitive interaction with humans. This study hypothesized that when humans infer intentionality from motion information, they initially estimate the direction of the goal of the movement in a Bayesian manner. The magnitude of the change in this estimated direction reflects the strength of intentionality and self-propelledness, which are correlated with the perceived strength of animacy. We tested this hypothesis through an experiment in which participants evaluated animacy, intentionality, and self-propelledness for dots moving on a screen. Path analysis and model comparison revealed that changes in motion direction had limited direct impact on perceived intentionality and self-propelledness, which are considered key components of animacy. In contrast, both the magnitude of goal direction changes and goal uncertainty significantly influenced these perceptions. These findings suggest that animacy perception may be realized through hierarchical Bayesian estimation, where the inference of goal direction plays a crucial mediating role.

As robots evolve, relying solely on “cuteness” based on baby-like appearance is no longer sufficient to elicit positive human responses. This research highlights “Azatoi,” an intentional and self-aware form of cuteness, as a key factor in promoting interpersonal interactions. Experiments were conducted with an avatar exhibiting” Azatoi” expressions, using the Ultimatum Game and impression assessment. The results reveal that the avatar with the “Azatoi“ smile elicited a stronger impression of intentionality, and its attractiveness increased only for female avatars. However, no significant differences in proposal acceptance rates were found in the Ultimatum Game based on the type of smile. This suggests that the situations in which the “Azatoi“ smile can be effectively used may be limited. Future research will focus on measuring the effects of“ Azatoi“ avatars in more socially relevant contexts.

In recent years, the concept of “cuteness” in appearance has been used to elicit positive reactions in human-robot interactions. However, as robots evolve and diversify, it becomes crucial to incorporate cuteness that extends beyond mere appearance to encompass other factors. Therefore, the focus of this study is on “Azatoi,” a self-conscious, intentional, and calculated form of “cuteness” as a new element aimed to evoke interpersonal reactions. To investigate the perception of “Azatoi,” we conducted an experiment in which participants rated facial expressions based on smiles. Our objective was to identify the elements that contribute to the perception of “Azatoi” and determine if it could be solely perceived through facial expressions. The results of the study indicated that certain smiles were generally perceived as “Azatoi,” differing from other facial expressions in terms of intentionality, emotionality, and perceived second-order intention. “Azatoi” smiles were associated with specific movements around the mouth (action units ((AU) 1, 4, 20, and 23) and eyebrows (AU 1, 4, and 7), potentially serving as triggers for perceiving “Azatoi.” The results demonstrate the significance of integrating “Azatoi” into robotic designs.

It is acknowledged that FaceNet, a deep learning-based face-recognition algorithm, failed to replicate certain features of the uncanny valley owing to the large difference between the evaluations by people and FaceNet for specific face images. For the images that FaceNet rated as highly human-like, localized attention to the lower part of the face (e.g., the mouth and chin) functioned as the basis for assessment. Such localized attention is considered to be part of the characteristics of individuals with autism spectrum disorder (ASD). This study investigated the similarity in face recognition between the ratings by FaceNet and those by individuals with ASD. Regression analyses were conducted with the ratings by FaceNet as the dependent variable and those by typically developing (TD) individuals or individuals with ASD as the independent one. The ASD group provided a better explanation of the FaceNet evaluation. These results indicate similarities between FaceNet and individuals with ASD.

Feeling that a non-human object has emotions (hereinafter referred to as “emotional state attribution”) is generally known as animacy perception. Previous studies have considered appearance and motion separately as factors that evoke emotional state attribution. Thus, if both the degree of human likeness in an object’s shape and the presence or absence of motion are considered simultaneously, we must consider whether strong emotional state attribution occurs even for objects that are not human-like in shape. In this study, we experimentally investigated the influence of human likeness in shape and movements evoking social relations on emotional state attribution, including their interaction, using three types of objects (human figure, shimeji mushroom, and match). In Experiments 1 and 2, although the human figure was rated as more human-like than the shimeji mushroom in terms of shape, emotions were attributed more strongly to the shimeji mushroom than to the human figure when accompanied by movements that evoked social relationships. The results suggest that people may attribute emotions more strongly to objects that resemble humans only to a certain extent in terms of shape when they show movements that evoke social relations.

Vicarious decisions are made on behalf of others that are not for the decision-makers themselves, but for the satisfaction of the others. They are often observed in interactive situations in the real-world, such as investment trusts in an outsourced agency (planners) and its clients (sponsors). We challenged the question of whether planners really could follow the intent of sponsors’ orders in making vicarious risk decisions. We designed and conducted an online experiment in which pairs of persons interacted with each other in the role of either sponsor or planner. Our results showed that planners adjusted the number of gambling or risky choices according to the sponsor’s orders, but did not take actions that reflected the sponsor’s risk preferences; nonetheless, sponsor’s satisfaction to the planner’s choice was substantially high. These findings shed light on the interaction design of how deeply vicarious agents (whether human or robot) should follow the client’s thoughts in collaborative tasks.

We examined the interaction in face-to-face selling situation. In particular, we examined how customer's nonverbal behaviors were related to their preference in a natural environment. We found that customers' body posture could be a good predictor about their preference. We also found that estimation on customer's preference could be achieved from two nonverbal behaviors better than single behavior. We discuss the present contributions toward constructing the efficient agent systems.

Interpersonal touch is said to have significant effects on social interaction. We used the ultimatum game to examine whether touch from a robot could inhibit a negative feeling to the robot. We set two experimental conditions: the one was "touch condition" in which unfair proposals were offered to a participant when a robot touched his/her arm and the other was "no touch condition" in which unfair proposals were offered when the same robot did not. We compared Medial Frontal Negativity (MFN) measured by EEG, whose amplitude is correlated with feeling of unfairness, between the two conditions. Result shows that MFN amplitude was larger in the no touch condition than in the touch condition. This indicates that touch from a robot may inhibit a sense of unfairness for the robot. Our finding suggests that touch from a robot could enhance positive feeling to the robot through human-robot interaction.

We attempted to determine whether the provision of diverse information could facilitate creative idea generation. For this purpose, we enlisted 35 students to generate two ideas individually by using the scanning material provided. The results revealed that the participants tended to select similar articles and to reduce their diversity even though diverse materials were provided. Further, the diversity of the materials they used actually was positively correlated with the quality of the generated ideas. We concluded that the diversity of materials that were not provided but nevertheless used had an effect on idea generation.

In the near future, it is expected for us to communicate with robots and computer agents in a natural way. In daily life, we usually speculate about partner's intentions from diverse nonverbal information expressed unconsciously. We thus need to investigate the method of speculating partner's intentions by using nonverbal information and to implement it with a robot or agent to realize the smooth communication. However, there is not a system satisfying necessary conditions which we considered for the measuring nonverbal information in natural communication. Therefore, we made a real-time system for readily measuring gaze directions and facial feature points at a time. And then, we established an experimental setting for measuring multimodal nonverbal information that participants expressed during communication. We used the system and setting to make an experiment for discriminating lie, as an example which intentions were unconsciously expressed by nonverbal information. As a result, we found that we could discriminate lies by using diverse nonverbal information in the same way people did

The aim of this paper is to reveal the relations between time scales and time series properties by concentrating on information requisite for speculators using a genetic learning model of investor sentiment. For this purpose, first the authors identified the conditions for describing investor sentiment by altering parameters of genetic algorithm. Then auto-correlations and conditional probabilities were calculated using the estimated models in the first step. The results show that both the amount and quality of information for the agents determine the time series properties. This implies that the preciseness of information which speculators permit depends on their time scales.

Interactive agents such as pet robots or adaptive speech interface systems that require forming a mutual adaptation process with users should have two competences. One of these is recognizing reward information from users' expressed paralanguage information, and the other is informing the learning system about the users by means of that reward information. The purpose of this study was to clarify the specific contents of reward information and the actual mechanism of a learning system by observing how 2 persons could create a smooth speech communication, such as that between owners and their pets. A communication experiment was conducted to observe how human participants create smooth communication through acquiring meaning from utterances in languages they did not understand. Then, based on experimental results, a meaning-acquisition model that considers the following 2 assumptions was constructed: (a) To achieve a mutual adaptive relationship with users, the model needs to induce users' adaptation and to exploit this induced adaptation to recognize the meanings of a user's speech sounds; and (b) to recognize users' utterances through trial-and-error interaction regardless of the language used, the model should focus on prosodic information in speech sounds, rather than on the phoneme information on which most past interface studies have focused. The results confirmed that the proposed model could recognize the meanings of users' verbal commands by using participants' adaptations to the model for its meaning-acquisition process. However, this phenomenon was observed only when an experimenter gave the participants appropriate instructions equivalent to catchphrases that helped users learn how to use and interact intuitively with the model. Thus, this suggested the need for a subsequent study to discover how to induce the participants' adaptations or natural behaviors without giving these kinds of instructions.

In this study, we propose an artificial market approach, which is a new agent-based approach to foreign exchange market studies. Using this approach, emergent phenomena of markets such as the peaked and fat-tailed distribution of rate changes were explained. First, we collected the field data through interviews and questionnaires with dealers and found that the features of dealer interaction in learning were similar to the features of genetic operations in biology. Second, we constructed an artificial market model using a genetic algorithm. Our model was a multiagent system with agents having internal representations about market situations. Finally, we carried out computer simulations with our model using the actual data series of economic fundamentals and political news. We then identified three emergent phenomena of the market. As a result, we concluded that these emergent phenomena could be explained by the phase transition of forecast variety, which is due to the interaction of agent forecasts and the demand-supply balance. In addition, the results of simulation were compared with the field data. The field data supported the simulation results. This approach therefore integrates fieldwork and a multiagent model, and provides a quantitative explanation of micro-macro relations in markets.

The classification of diffuse lung opacities in thin-section computed tomography (HRCT) images is an important step of developing a computer-aided diagnosis (CAD) system. We have evaluated the performance of a Gabor filter-based classification of typical diffuse lung opacities in HRCT images. The experimental results show that the Gabor filter-based approach is an effective means of classifying diffuse lung diseases.

In a hierarchical VLSI layout design, the block-level layout design is called a "chip floor plan." In this paper, a semi-automatic VLSI chip floor plan algorithm and its implementation are presented. The initial block placement is obtained by an attractive and repulsive force method (AR method), and the subsequent block packing process is performed by gradually moving and reshaping blocks with chip boundary shrinking. The chip area estimation is performed by using individual block area calculations from empirically obtained equations. A set of interactive commands is also provided to facilitate the manual optimization processes using a color graphic terminal. By processing several practical VLSI circuits, it is shown that the method is very effective for handling various kinds of blocks and is able to reduce the design effort required to achieve the chip floor plan.

A parallel processing algorithm for logic module placement is presented. Conventionally, such placement problems have been solved on a single processor in a sequential manner. In this paper, it is shown that a two-dimensional processor array structure can be applied to the placement problem, resulting in a substantial reduction of the processing time. This parallel processing algorithm is based on the concept that the adjacent pairwise exchange method could be expanded to the parallel processing case. By using simulation programs, it is shown that the placement results obtained by the parallel processing algorithm are a little better than those obtained by the sequential algorithm. In addition, the theoretical estimations in respect to the processing cycle iterations correspond well with the simulation results.