In this study, we developed an interview training system in which a virtual agent provides feedback on improvements in interviewees’ nonverbal behaviors. We verified the effectiveness of providing assertive feedback from the agent. Assertive communication involves expressing opinions and feelings while respecting others perspectives and opinions. Whether feedback is helpful or threatening depends on the individual characteristics of those receiving it. We classified participants according to their tendency to learn from failures and compared the effectiveness of assertive feedback. The results indicated that assertive feedback effectively reduced the agent's aggressiveness for those with a low tendency to learn from failures and improved the agent's perceived friendliness for those with a high tendency to learn from failures.

This study aimed to investigate whether Japanese participants have subconscious gender biases against a male and a female virtual agents and quantify those biases if any. The participants were given a negative feedback after taking a logical thinking test regardless of their performance by the virtual agents. Then we measured their subconscious gender biases with IAT. The preliminary results indicated the participants showed general gender biases against the male agent. On the other hand, stereotypical gender biases strongly appeared against the female agent. These results provide implications for more gender-neutral agent design in applications where they might give negative feedbacks such as education and training.

The COVID-19 pandemic has had a significant socio-economic impact on the world. Specifically, social distancing has impacted many activities that were previously conducted face-to-face. One of these is the training that students receive for job interviews. Thus, we introduce a job interview training system that will give students the ability to continue receiving this type of training. Our system recognizes the nonverbal behaviors of an interviewee, namely gaze, facial expression, and posture using a Tobii eye tracker and cameras. The system compares the recognition results with those of models of exemplary nonverbal behaviors of an interviewee and highlights the behaviors that need improvement while playing back the interview recording. Most current interview training systems require high-end Hardware and Software and are not designed for general users, and there are few systems using CG agents to give feedback. The purpose of this study was to conduct an initial evaluation of the usefulness of this system and to identify areas for improvement. The results of the initial evaluation of the system indicate that improvements in the recognition accuracy of nonverbal behaviors and the quality of the interaction with the CG agent are needed.

The COVID-19 pandemic has had a significant socio-economic impact on the world. Specifically, social distancing has impacted many activities that were previously conducted face-to-face. One of these was the training that students receive for job interviews. Thus, we developed a job interview training system that will give students the ability to continue receiving this type of training. Our system recognized the nonverbal behaviors of an interviewee, namely gaze, facial expression, and posture and compares the recognition results with those of models of exemplary nonverbal behaviors of an interviewee. A virtual agent acted as an advisor gives feedback on the interviewee's behaviors that need improvement. In order to verify the effectiveness of the two kinds of feedback, namely, rationalized feedback (with quantitative recognition results) vs. non-rationalized one, we compared interviewees’ impression. The results of the evaluation experiment indicated that the virtual agent with rationalized feedback was rated as more reliable but less friendly than the non-rationalized feedback.

The purpose of this study is to develop gaze behaviors of a virtual agent that adapt to the gaze behavior of a human conversant so that both introverts and extroverts feel comfortable with. The proposed adaptive gaze model changes the agent's gaze percentage to the user according to the user's gaze percentage to the agent. The results of the evaluation experiment suggest the adaptive gaze model enhances perceived friendliness of the agent and lowers stress toward conversation for introverts.

Gaze plays an important role in human-human communication. Adequate gaze control of a virtual agent is also essential for successful and believable human-agent interaction. Researchers on HAI have developed gaze control models by taking account of gaze duration, frequency, and timing of gaze aversion. However, cultural differences in gaze behaviors have not been focused enough. We aimed to investigate cultural differences in gaze behaviors and their perception by developing virtual agents with Japanese gaze behaviors, American gaze behaviors, and fixed gaze behaviors. We then compared their effects on the impressions of the agents and interactions. Our first experimental results with Japanese participants suggested that the impression of the agent is affected by participants' shyness and familiarity of the gaze patterns performed by the agent. This paper reports the results of our second experiment that investigated the effects of the gender of agents and the gender of participants on perception of culture-specific gaze behaviors. The preliminary results suggests there are differences between the gender of participants and agents on the perception of favorable gaze behaviors. The female participants had a better impression on the agent of the same gender that performed the American gaze behavior, while the male participants preferred the agents with the Japanese gaze behavior. These results indicate favorable gaze behaviors vary according to users' gender and agents' gender.

This paper reports our preliminary results of a cross-cultural study of perception and acceptance of cultural specific self-adaptors performed by a virtual agent. There are culturally-defined preferences in self-adaptors and other bodily expressions, and allowance level of expressing such non-verbal behavior are culture-dependent. We conducted a web experiment to evaluate the impression and acceptance of Japanese culture specific self-adaptors and gathered participants from 8 countries. The results indicated non-Japanese participants' insensitivity to the different types of self-adaptors and over sensitivity to Japanese participants' to stressful self-adaptors.

This paper focuses on “animacy” of a robot in order for the robot to be accepted to our daily lives. Related research shows as the robot's perceived animacy increases, perceived friendliness toward the robot also increases. We hypothesize as the animacy of the robot's behavior increases, our tolerance level for the robot's malfunction also increases. Lego Mindstorms is used to evaluate the robot's animacy, friendliness, and tolerance level while performing experimental tasks. The results showed the contrary evaluation of the tolerance level, as the higher animacy of the robot resulted in lower tolerance by the participants. The results also suggest that there is a dichotomy in the tolerance level of the robot's malfunctions between participants' mental stance to the robot.

Self-adaptors are bodily behaviours that often involve self-touch that is regarded as taboo in public. However, self-adaptors also occur during casual conversations between friends. We developed a virtual agent that exhibits self-adaptors during conversation with users. Our continuous evaluation of the interaction between the agents that exhibit self-adaptors and without indicated that there is a dichotomy on the impression on the agents between users with high social skills and those with low skills. People with high social skills feel more friendliness toward an agent that exhibits self-adaptors than those with low social skills. The result suggests the need to tailor non-verbal behaviour of virtual agents according to user’s social skills.

This paper reports a preliminary result of a cross-cultural study on the facial regions as cues to recognize virtual agents' facial expressions. We believe providing research results on the perception of cartoonish virtual agents' facial expressions to HRI research community is meaningful in order to minimize the effort to develop robot's facial expressions. The result implies Japanese weighed facial cues more heavily in the eye regions than Hungarians, who weighed facial cues more heavily in the mouth region than Japanese.

This paper introduces a preliminary result of a cross-cultural study on the facial regions as cues to recognize virtual agents' facial expressions. We believe providing research results on the perception of cartoonish virtual agents' facial expressions to HRI research community is meaningful in order to minimize the effort to develop robot's facial expressions. The result implies 1) the mouth region is more effective in conveying the emotions of the facial expressions than the eye region, 2) there are cultural differences in using facial regions as cues to recognize cartoonish facial expressions between Hungary and Japan.