論文

Human–agent interaction is a means of extending our society through social agents, artificial beings that work through social channels to humans. This paper describes the trends of human–agent interaction and the directions of HAI research that will occur in the future.

One of the possible benefits of robot-mediated education is the effect of the robot becoming a catalyst between people and facilitating learning. In this study, the authors focused on an asynchronous active learning method mediated by robots. Active learning is believed to help students continue learning and develop the ability to think independently. Therefore, the authors improved the UGA (User Generated Agent) system that we have created for long-term active learning in COVID-19 to create an environment where children introduce books to each other via robots. The authors installed the robot in an elementary school and conducted an experiment lasting more than a year. As a result, it was confirmed that the robot could continue to be used without getting bored even over a long period of time. They also analyzed how the children created the contents by analyzing the contents that had a particularly high number of views. In particular, the authors observed changes in children’s behavior, such as spontaneous advertising activities, guidance from upperclassmen to lowerclassmen, collaboration with multiple people, and increased interest in technology, even under conditions where the new coronavirus was spreading and children’s social interaction was inhibited.

Driven by the rapid development of artificial intelligence (AI) and anthropomorphic robotic systems, the various possibilities and risks of such technologies have become a topic of urgent discussion. Although science fiction (SF) works are often cited as references for visions of future developments, this framework of discourse may not be appropriate for serious discussions owing to technical inaccuracies resulting from its reliance on entertainment media. However, these science fiction works could help researchers understand how people might react to new AI and robotic systems. Hence, classifying depictions of artificial intelligence in science fiction may be expected to help researchers to communicate more clearly by identifying science fiction elements to which their works may be similar or dissimilar. In this study, we analyzed depictions of artificial intelligence in SF together with expert critics and writers. First, 115 AI systems described in SF were selected based on three criteria, including diversity of intelligence, social aspects, and extension of human intelligence. Nine elements representing their characteristics were analyzed using clustering and principal component analysis. The results suggest the prevalence of four distinctive categories, including human-like characters, intelligent machines, helpers such as vehicles and equipment, and infrastructure, which may be mapped to a two-dimensional space with axes representing intelligence and humanity. This research contributes to the public relations of AI and robotic technologies by analyzing shared imaginative visions of AI in society based on SF works.

We are now faced with the wall of the diversity of values. We are often required to consider or respect the values of other people though, it is not easy to sense them since we tend to think within our scope of knowledge, experience, and imagination. We have worked on how to exchange values and achieve a synergetic effect among people. In this paper, we are presenting our prototype system, called AIR-VAS, aiming to support becoming aware of values in group discussion. AIR-VAS has been developed as the system which recognizes characteristic opinions of a group and shares them among all engaging groups on the discussion. The recognized and shared opinion is based on the values of the people of the group. Through the sharing of opinion, people can know the different viewpoints on the issue of the current discussion, so that AIR-VAS can provide stimulation to people for idea generation. We have developed AIR-VAS on the approach that visualizing statements, which are presented during a discussion, as the word co-occurrence network. We implemented opinion sharing as the process of the network re-construction including presented sub-network as an opinion of a certain group. According to the experimental usage of the developed system, we analyzed the relationship between discussion and visualized information, and discuss what information gives awareness.

本研究では，裸眼立体視ディスプレイを使用したバーチャルロボットヘッドと人間の対人距離の有効性と，視線推定について検討する．人間と同じような身体を持つ物理的なロボットが人間とコミュニケーションを行う際の問題点として，人を傷つける恐れがあるため接近動作ができず，常に適切な対人距離を維持できるとは限らないことが挙げられる．身体動作の欠点を克服するため，平面ディスプレイに表示されるオンスクリーンエージェントを使用することが検討されてきたが，これはモナリザ効果を発生させ，観察者が視点を変えてもエージェントから見つめられているように感じてしまう問題がある．本研究では，これらの問題を解決するため，裸眼立体視可能なディスプレイ上に，人型の3Dモデルを表示することによりバーチャルロボットヘッドを開発した．美術館見学のシナリオを用意し，25名の参加者に美術館案内エージェントからどの程度移動したいか尋ねた．また美術館案内エージェントが見つめる先を推定することを求めた．実験の結果から，バーチャルロボットヘッドと人の間に対人距離が適用されることがわかった．またバーチャルロボットヘッドに箱を被せ，観察者に身体を想起してもらうことにより，モナリザ効果が減少することがわかった．

The authors evaluate the extent to which a user’s impression of an AI agent can be improved by giving the agent the ability of self-estimation, thinking time, and coordination of risk tendency. The authors modified the algorithm of an AI agent in the cooperative game Hanabi to have all of these traits, and investigated the change in the user’s impression by playing with the user. The authors used a self-estimation task to evaluate the effect that the ability to read the intention of a user had on an impression. The authors also show thinking time of an agent influences impression for an agent. The authors also investigated the relationship between the concordance of the risk-taking tendencies of players and agents, the player’s impression of agents, and the game experience. The results of the self-estimation task experiment showed that the more accurate the estimation of the agent’s self, the more likely it is that the partner will perceive humanity, affinity, intelligence, and communication skills in the agent. The authors also found that an agent that changes the length of thinking time according to the priority of action gives the impression that it is smarter than an agent with a normal thinking time when the player notices the difference in thinking time or an agent that randomly changes the thinking time. The result of the experiment regarding concordance of the risk-taking tendency shows that influence player’s impression toward agents. These results suggest that game agent designers can improve the player’s disposition toward an agent and the game experience by adjusting the agent’s self-estimation level, thinking time, and risk-taking tendency according to the player’s personality and inner state during the game.

Eye contact is crucial in social interactions, linking with sincerity and friendliness. However, blind people cannot see and make eye contact when they communicate with sighted people. It influences the involvement of blind people in blind-sighted conversations. Based on this context, we implemented Social glasses with an eye-tracking system, aiming to improve the communication quality between blind and sighted people in face-to-face conversations. Social glasses attempts to simulate the appropriate gaze for blind people, especially establishing the “eye contact” in blind-sighted conversations. To evaluate the impact of the interactive gaze displayed on the Social glasses, we performed dyadic-conversation tests under four experimental conditions (No Gaze, Constant Gaze, Random Gaze, and Interactive Gaze) for 40 participants. Quantitative results showed that the Interactive gaze has a positive impact on improving the communication quality between blind and sighted people, which were consistent with a qualitative analysis of the participants’ comments.

Gaze behaviors contain rich information regarding a person's emotions and engagements. Reciprocal eye contact can invoke feelings of liking between two strangers. But blind people cannot perceive and establish the eye contact with sighted counterparts, causing their feelings of social isolation and low confidence in conversations. Thus, our research purpose is to let blind people perceive and react gaze behaviors in social interactions. A Social Glasses system has been implemented iteratively to deliver the multisensory feedback channels of the "eye contact", integrating both visual and tactile feedback. Specifically, the system consists of a Social Glasses device and a tactile wristband, which are worn by a blind person. The Social Glasses simulates the natural gaze for the blind person, aiming at establishing the "eye contact" between blind and sighted people. The tactile wristband enables the blind person to perceive the corresponding tactile feedback when an "eye contact" happens. To test the system, we conducted a user experiment with 40 participants, including 10 blind-sighted pairs (N = 20) and 10 blindfolded-sighted pairs (N = 20), to see how it could help increase the communication quality between blind and sighted people, as well as to suggest implications for its design. Our main findings demonstrated that both the simulated gaze and the tactile feedback were significantly effective to enhance the communication quality in blind-sighted conversations. Overall, we contribute (1) empirical research findings on how a Social Glasses system enhances the communication quality in blind-sighted conversations; and (2) design principles to inform future assistive wearable device for augmenting social interactions.

The authors investigated the effect of gaze tracking of humans by an agent. Psychological studies have shown that the presence of others facilitates task with implying partnership. This effect also occurs if the presence of others is replaced with a robot. Furthermore, the previous research showed that being touched by a robot motivates a person. However, direct contact such as the touch of a robot has a risk of disturbing human work. The authors used gaze tracking for social facilitation because gaze tracking can also facilitate human efforts and implemented a social facilitation robot agent by using gaze tracking. The agent projects its face on a surface to show quickly changing expressions, which is difficult to achieve with ordinary robots. The agent can also express gaze tracking by implementing eye trackers. The authors prepared three conditions with monotonous tasks to compare the effects of gaze tracking: The agent gazes at the instructed point moves its gaze randomly and traces the gaze point of a participant. The results showed that the participants' motivation increased if the agent is tracking gazed point by a human, even though there are no differences between the achievement scores of the task and the continuation time.

This study analyzes the value conflict of a paper on fan fiction writing that used online fan fiction novels as a source to extract and filter sexual expressions from text. The boundaries of public and private information are ambiguous because users are not always aware of or have agreed to the fact that their content is to be used openly. The case was complicated by the fact that the use of these data by researchers violated an unconsciously infringed upon right of a vulnerable community with a weak legal position. This paper describes the debate on this topic among researchers from engineering and humanities fields on whether the purpose of the research was ethically acceptable; how the systems can be embedded in ethical values; and what ethical, legal, social, and educational lessons are appropriate for governance of artificial intelligence (AI). Our analysis aimed not only to clarify the abstract concept of privacy but also to make changes to the submission guidelines for authors. We hope that our analysis contributes to the governance of ethical AIs and AI ethics on handling sensitive aspects of online activities.

There have been several attempts in recent years to develop a remote communication device using sensory modalities other than speech that would induce a user's positive experience with his/her conversation partner. Specifically, Hugvie is a human-shaped pillow as well as a remote communication device enabling users to combine a hugging experience with telecommunication to improve the quality of remote communication. The present research is based on the hypothesis that using Hugvie maintains users' level of trust toward their conversation partners in situations prone to suspicion. The level of trust felt toward other remote game players was compared between participants using Hugvie and those using a basic communication device while playing a modified version of Werewolf, a conversation-based game, designed to evaluate trust. Although there are always winners and losers in the regular version of Werewolf, the rules were modified to generate a possible scenario in which no enemy was present among the players and all players would win if they trusted each other. We examined the effect of using Hugvie while playing Werewolf on players' level of trust toward each other and our results demonstrated that in those using Hugvie, the level of trust toward other players was maintained.

Social brain theory hypothesizes that the human brain becomes larger through evolution mainly because of reading others’ intentions in society. Reading opponents’ intentions and cooperating with them or outsmarting them results in an intelligence arms race. The authors discuss the evolution of such an arms race, represented as finite state automatons, under three distinct payoff schemes and the implications of these results, which suggest that agents increase complexity of their strategies. The analyses of the high-ranking agents’ automata suggests the process to acquire complex strategy in delayed reward condition.

技術の発達によって機器の機能が増加している．増加した機能をユーザに分かりやすく伝える手法として，擬人化エージェントを用いた機能説明が提案されてきた．本研究では擬人化エージェントを実世界に拡張させる演出を用いて，複合機（Multi-Function Printer，MFP）の操作説明を行い，人間のインタラクションがどのように変化するか検証した．検証結果として，スクリーン上のエージェントの説明を実世界に拡張することにより，その後のユーザの操作時間が有意に短くなることを確認した．また，ユーザの発話量が有意に増加すること，説明後に学習への興味が向上することが評価結果から示唆された．

「人狼知能」プロジェクトでは、コミュニケーションゲームである「汝は人狼なりや(Are You a Werewolf?)」をプレイ可能なエージェントの作成問題を、チェスや将棋に続く人工知能の標準問題として解くことを目指している。我々は、CEDEC2015において『人工知能が人ともにゲームを繰り返し楽 しめる』ことを最終目標とした、知的エージェントによる人狼ゲーム大会を開催した。大会には38体のエージェントが参加し、15体のエージェントによる決勝戦が行われた。優勝エージェントは、他のエー ジェントと比べ有意に勝率が高く、人狼知能エージェントの強さに差があることが示された。また、役職と勝率の関係を分析した結果、勝率に特徴のある役職が存在することが明らかとなった。

人工知能技術の「責任ある研究とイノベーション（Responsible Research and Innovation: RRI）」のためには多様なステークホルダー間による対話・協働が不可欠である。本研究では，情報学系研究者のほか情報学の倫理的・法的・社会的問題を考える人文・社会科学研究者，政策系の専門家やSFなど創作/編集活動関係者，メディア，一般市民など多様なステークホルダーにアンケート調査を行った。10年後の運転，育児，介護，人生選択，健康管理，創作活動，防災，軍事の8分野では，全体として運転・防災・軍事分野など「知的な機械・システム」の導入に社会的合意が必要とされる分野の機械化には積極的な意見が多い。一方，ライフイベントにおける意思決定や健康管理など個人選択に委ねられる分野は「人間が主体で機械を活用する」傾向にあった。回答者の専門や価値観，経験によって意見は多様であり，個別の技術導入場面において機械と人間の関係をいかに創造的に組み替えていくかの議論が重要である。

Helping hand robots have been the focus of a number of studies and have high potential in modern manufacturing processes and for use in daily living. As helping hand robots interact closely with users, it is important to find natural and intuitive user interfaces for interacting with the robots in various situations. This study describes a set of gestures for interacting with and controlling helping hand robots in situations in which users need to manually control the robot but both hands are not available, for example, when users are holding tools or objects in their hands. The gestures are derived from an experimental study that asked participants for gestures suitable for controlling primitive robot motions. The selected gestures can be used to control translation and orientation of an end effector of a helping hand robot when one or both hands are engaged with tasks. As an example for validating the proposed gestures, we implemented a helping hand robot system to perform a soldering task.

In designing and developing a gesture recognition system, it is crucial to know the characteristics of a gesture selected to control, for example, an end effector of a robot arm. We conducted an experiment to collect a set of user-defined gestures and investigate characteristics of the gestures for controlling primitive motions of an end effector in human-robot collaboration. We recorded 152 gestures from 19 volunteers by presenting virtual robotic arm movements to the participants, and then asked the participants to think about and perform gestures that would cause the motions. It was found that the hands were the parts of the body used most often for gesture articulation even when the participants were holding tools and objects with both hands: a number of participants used one- and two-handed gestures interchangeably, gestures were consistently performed by the participants across all pairs of reversible gestures, and the participants expected better recognition performance for gestures that were easy to think of and perform. These findings are expected to be useful as guidelines in creating a gesture set for controlling robotic arms according to natural user behaviors.