We conducted an experiment, in which we found that training a drone increased its animacy, attachment to it, and affinity with it. Additionally, we found that the drone impressed trainers as a puppy while it impressed non-trainers as a fly.
{"title":"Trainability Leads to Animacy: A Case of a Toy Drone","authors":"Yutai Watanabe, Yuya Onishi, Kazuaki Tanaka, Hideyuki Nakanishi","doi":"10.1145/3349537.3352776","DOIUrl":"https://doi.org/10.1145/3349537.3352776","url":null,"abstract":"We conducted an experiment, in which we found that training a drone increased its animacy, attachment to it, and affinity with it. Additionally, we found that the drone impressed trainers as a puppy while it impressed non-trainers as a fly.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124263363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With an ever increasing demand for personal service robots and artificial assistants, companies, start-ups and researchers aim to better understand what makes robot platforms more likable. Some argue that increasing a robot's humanlikeness leads to a higher acceptability. Others, however, find that extremely humanlike robots are perceived as uncanny and are consequently often rejected by users. When investigating people's perception of robots, the focus of the related work lies almost solely on the first impression of these robots, often measured based on images or video clips of the robots alone. Little is known about whether these initial positive or negative feelings persist when giving people the chance to interact with the robot. In this paper, 48 participants were gradually exposed to the capabilities of a robot and their perception of it was tracked from their first impression to after playing a short interactive game with it. We found that initial uncanny feelings towards the robot were significantly decreased after getting to know it better, which further highlights the importance of using real interactive scenarios when studying people's perception of robots. In order to elicit uncanny feelings, we used the 3D blended embodiment Furhat and designed four different facial textures for it. Our work shows that a blended platform can cause different levels of discomfort towards it depending on the facial texture and may thus be an interesting tool for further research on the uncanny valley.
{"title":"Let Me Get To Know You Better: Can Interactions Help to Overcome Uncanny Feelings?","authors":"Maike Paetzel, Ginevra Castellano","doi":"10.1145/3349537.3351894","DOIUrl":"https://doi.org/10.1145/3349537.3351894","url":null,"abstract":"With an ever increasing demand for personal service robots and artificial assistants, companies, start-ups and researchers aim to better understand what makes robot platforms more likable. Some argue that increasing a robot's humanlikeness leads to a higher acceptability. Others, however, find that extremely humanlike robots are perceived as uncanny and are consequently often rejected by users. When investigating people's perception of robots, the focus of the related work lies almost solely on the first impression of these robots, often measured based on images or video clips of the robots alone. Little is known about whether these initial positive or negative feelings persist when giving people the chance to interact with the robot. In this paper, 48 participants were gradually exposed to the capabilities of a robot and their perception of it was tracked from their first impression to after playing a short interactive game with it. We found that initial uncanny feelings towards the robot were significantly decreased after getting to know it better, which further highlights the importance of using real interactive scenarios when studying people's perception of robots. In order to elicit uncanny feelings, we used the 3D blended embodiment Furhat and designed four different facial textures for it. Our work shows that a blended platform can cause different levels of discomfort towards it depending on the facial texture and may thus be an interesting tool for further research on the uncanny valley.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114453388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper introduces the concept of team design patterns and proposes an intuitive graphical language for describing the design choices that influence how intelligent systems (e.g. artificial intelligence, robotics, etc.) collaborate with humans. We build on the notion of design patterns and characterize important dimensions within human-agent teamwork. These dimensions are represented using a simple, intuitive graphical iconic language. The simplicity of the language allows easier expression, sharing and comparison of human-agent teaming concepts. Having such a language has the potential to improve the collaborative interaction among a variety of stakeholders such as end users, project managers, policy makers and programmers that may not be human-agent teamwork experts themselves. We also introduce an ontology and specification formalization that will allow translation of the simple iconic language into more precise definitions. By expressing the essential elements of teaming patterns in precisely defined abstract team design patterns, we provide a foundation that will enable working towards a library of reusable, proven solutions for human-agent teamwork.
{"title":"Team Design Patterns","authors":"J. Diggelen, Matthew Johnson","doi":"10.1145/3349537.3351892","DOIUrl":"https://doi.org/10.1145/3349537.3351892","url":null,"abstract":"This paper introduces the concept of team design patterns and proposes an intuitive graphical language for describing the design choices that influence how intelligent systems (e.g. artificial intelligence, robotics, etc.) collaborate with humans. We build on the notion of design patterns and characterize important dimensions within human-agent teamwork. These dimensions are represented using a simple, intuitive graphical iconic language. The simplicity of the language allows easier expression, sharing and comparison of human-agent teaming concepts. Having such a language has the potential to improve the collaborative interaction among a variety of stakeholders such as end users, project managers, policy makers and programmers that may not be human-agent teamwork experts themselves. We also introduce an ontology and specification formalization that will allow translation of the simple iconic language into more precise definitions. By expressing the essential elements of teaming patterns in precisely defined abstract team design patterns, we provide a foundation that will enable working towards a library of reusable, proven solutions for human-agent teamwork.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129001373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Currently, it is difficult for humans and AI agents to cooperate because the agent has incomplete intention understanding. In this paper, we propose the design of cooperative interaction between humans and AI creatures. As an experiment, two creatures learned to lift a heavy box in a virtual environment simultaneously. As a result, one of the creatures was able acquire the behavior of following the other creature automatically. We need to verify whether cooperation between humans and AI can be established through ongoing investigations.
{"title":"Design of Cooperative Interaction between Humans and AI Creatures through Reinforcement Learning","authors":"Ryosuke Takata, Yugo Takeuchi","doi":"10.1145/3349537.3352771","DOIUrl":"https://doi.org/10.1145/3349537.3352771","url":null,"abstract":"Currently, it is difficult for humans and AI agents to cooperate because the agent has incomplete intention understanding. In this paper, we propose the design of cooperative interaction between humans and AI creatures. As an experiment, two creatures learned to lift a heavy box in a virtual environment simultaneously. As a result, one of the creatures was able acquire the behavior of following the other creature automatically. We need to verify whether cooperation between humans and AI can be established through ongoing investigations.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"20 5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125648486","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Like a human listener, a listener agent reacts to its communicational partners' non-verbal behaviors such as head nods, facial expressions, and voice tone. When adopting these modalities as inputs and develop the generative model of reactive and spontaneous behaviors using machine learning techniques, the issues of multimodal fusion emerge. That is, the effectiveness of different modalities, frame-wise interaction of multiple modalities, and temporal feature extraction of individual modalities. This paper describes our investigation on these issues of the task in generating of virtual listeners' reactive and spontaneous idling behaviors. The work is based on the comparison of corresponding recurrent neural network (RNN) configurations in the performance of generating listener's (the agent) head movements, gaze directions, facial expressions, and postures from the speaker's head movements, gaze directions, facial expressions, and voice tone. A data corpus recorded in a subject experiment of active listening is used as the ground truth. The results showed that video information is more effective than audio information, and frame-wise interaction of modalities is more effective than temporal characteristics of individual modalities.
{"title":"An Investigation on the Effectiveness of Multimodal Fusion and Temporal Feature Extraction in Reactive and Spontaneous Behavior Generative RNN Models for Listener Agents","authors":"Hung-Hsuan Huang, Masato Fukuda, T. Nishida","doi":"10.1145/3349537.3351908","DOIUrl":"https://doi.org/10.1145/3349537.3351908","url":null,"abstract":"Like a human listener, a listener agent reacts to its communicational partners' non-verbal behaviors such as head nods, facial expressions, and voice tone. When adopting these modalities as inputs and develop the generative model of reactive and spontaneous behaviors using machine learning techniques, the issues of multimodal fusion emerge. That is, the effectiveness of different modalities, frame-wise interaction of multiple modalities, and temporal feature extraction of individual modalities. This paper describes our investigation on these issues of the task in generating of virtual listeners' reactive and spontaneous idling behaviors. The work is based on the comparison of corresponding recurrent neural network (RNN) configurations in the performance of generating listener's (the agent) head movements, gaze directions, facial expressions, and postures from the speaker's head movements, gaze directions, facial expressions, and voice tone. A data corpus recorded in a subject experiment of active listening is used as the ground truth. The results showed that video information is more effective than audio information, and frame-wise interaction of modalities is more effective than temporal characteristics of individual modalities.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130123792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Akane Matsushima, N. Oka, Chie Fukada, Kazuaki Tanaka
evel (Austin 1962). DAs constitute the most fundamental part of communication, and the comprehension of DAs is essential to human-agent interaction. The purpose of this study is to enable an agent to behave properly in response to DAs without their explicit representation on one hand and to estimate the DAs explicitly on the other hand. The former is realized by reinforcement learning and the latter by Bayesian inference. The simulation results demonstrated that the agent not only responded to DAs successfully but also inferred the DAs correctly.
{"title":"Understanding Dialogue Acts by Bayesian Inference and Reinforcement Learning","authors":"Akane Matsushima, N. Oka, Chie Fukada, Kazuaki Tanaka","doi":"10.1145/3349537.3352786","DOIUrl":"https://doi.org/10.1145/3349537.3352786","url":null,"abstract":"evel (Austin 1962). DAs constitute the most fundamental part of communication, and the comprehension of DAs is essential to human-agent interaction. The purpose of this study is to enable an agent to behave properly in response to DAs without their explicit representation on one hand and to estimate the DAs explicitly on the other hand. The former is realized by reinforcement learning and the latter by Bayesian inference. The simulation results demonstrated that the agent not only responded to DAs successfully but also inferred the DAs correctly.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133420081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Suguru Honda, Taishi Sawabe, Shogo Nishimura, Wataru Sato, Yuichiro Fujimoto, Alexander Plopski, M. Kanbara, H. Kato
Humanitude is a multimodal communication care method that utilizes seeing, touching, and speaking. Moreover, a touch-care method is well known as an effective care method mainly focus on touch motion. These kinds of care techniques are effective in practical situations, however, it is difficult to provide such care therapy to all patients due to the lack of human resources. To address this problem, researchers try to develop a touch-care robot that can provide touch-care automatically. Conventional research of touch-care robot mainly focuses on the movement of stroke or the speech that only considers the impression of the contents of speech but not prosodic information. Therefore, in this research, we focus on the speech rate in the prosodic information with stroke motion. In this work, we investigate the effects of speech rate on the prosodic information and evaluate the relationship between stroke pace and speech rate to improve human comfort. We conducted a user study with 6 participants around 20 years old males. As a result of the list of the questionnaire suggests a correlation between stroke pace and speech rate that provides comfort.
{"title":"Evaluation of Relationship between Stroke Pace and Speech Rate for Touch-Care Robot","authors":"Suguru Honda, Taishi Sawabe, Shogo Nishimura, Wataru Sato, Yuichiro Fujimoto, Alexander Plopski, M. Kanbara, H. Kato","doi":"10.1145/3349537.3352793","DOIUrl":"https://doi.org/10.1145/3349537.3352793","url":null,"abstract":"Humanitude is a multimodal communication care method that utilizes seeing, touching, and speaking. Moreover, a touch-care method is well known as an effective care method mainly focus on touch motion. These kinds of care techniques are effective in practical situations, however, it is difficult to provide such care therapy to all patients due to the lack of human resources. To address this problem, researchers try to develop a touch-care robot that can provide touch-care automatically. Conventional research of touch-care robot mainly focuses on the movement of stroke or the speech that only considers the impression of the contents of speech but not prosodic information. Therefore, in this research, we focus on the speech rate in the prosodic information with stroke motion. In this work, we investigate the effects of speech rate on the prosodic information and evaluate the relationship between stroke pace and speech rate to improve human comfort. We conducted a user study with 6 participants around 20 years old males. As a result of the list of the questionnaire suggests a correlation between stroke pace and speech rate that provides comfort.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115739297","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Emotions play an important role in human-agent interaction. To realise natural interaction it is essential for an agent to be able to analyse the sentiment in users' utterances. Modern agents use a distributed service model in which their functions can be located on any number of computers including cloud-based servers. Outsourcing the speech recognition and sentiment analysis to a cloud service enables even simple agents to adapt their behaviour to the emotional state of their users. In this study we test whether sentiment analysis tools can accurately gauge sentiment in human-chatbot interaction. To that effect, we compare the quality of sentiment analysis obtained from three major suppliers of cloud-based sentiment analysis services (Microsoft, Amazon and Google). In addition, we compare their results with the leading lexicon-based software, as well as with human ratings. The results show that although the sentiment analysis tools agree moderately with each other, they do not correlate well with human ratings. While the cloud-based services would be an extremely useful tool for human-agent interaction, their current quality does not justify their usage in human-agent conversations.
{"title":"Cloud-Based Sentiment Analysis for Interactive Agents","authors":"M. Keijsers, C. Bartneck, H. Kazmi","doi":"10.1145/3349537.3351883","DOIUrl":"https://doi.org/10.1145/3349537.3351883","url":null,"abstract":"Emotions play an important role in human-agent interaction. To realise natural interaction it is essential for an agent to be able to analyse the sentiment in users' utterances. Modern agents use a distributed service model in which their functions can be located on any number of computers including cloud-based servers. Outsourcing the speech recognition and sentiment analysis to a cloud service enables even simple agents to adapt their behaviour to the emotional state of their users. In this study we test whether sentiment analysis tools can accurately gauge sentiment in human-chatbot interaction. To that effect, we compare the quality of sentiment analysis obtained from three major suppliers of cloud-based sentiment analysis services (Microsoft, Amazon and Google). In addition, we compare their results with the leading lexicon-based software, as well as with human ratings. The results show that although the sentiment analysis tools agree moderately with each other, they do not correlate well with human ratings. While the cloud-based services would be an extremely useful tool for human-agent interaction, their current quality does not justify their usage in human-agent conversations.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122150279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Autonomous machines are in the near future likely to increasingly interact with humans, and carry out their functions outside controlled settings. Both of these developments increase the requirements of machines to be trustworthy to humans. In this work, we argue that machines may also benefit from being able to explicitly build or withdraw trust with specific humans. The latter is relevant in situations where the integrity of an autonomous system is compromised, or if humans display untrustworthy behaviour towards the system. Examples of systems that could benefit might be delivery robots, maintenance robots, or autonomous taxis. This work contributes by presenting a biologically plausible model of unconditional trust dynamics, which simulates trust building based on familiarity, but which can be modulated by painful and gentle touch. The model displays interactive behaviour by being able to realistically control pupil dynamics, as well as determine approach and avoidance motivation.
{"title":"A Computational Model of Trust-, Pupil-, and Motivation Dynamics","authors":"Trond A. Tjøstheim, B. Johansson, C. Balkenius","doi":"10.1145/3349537.3351896","DOIUrl":"https://doi.org/10.1145/3349537.3351896","url":null,"abstract":"Autonomous machines are in the near future likely to increasingly interact with humans, and carry out their functions outside controlled settings. Both of these developments increase the requirements of machines to be trustworthy to humans. In this work, we argue that machines may also benefit from being able to explicitly build or withdraw trust with specific humans. The latter is relevant in situations where the integrity of an autonomous system is compromised, or if humans display untrustworthy behaviour towards the system. Examples of systems that could benefit might be delivery robots, maintenance robots, or autonomous taxis. This work contributes by presenting a biologically plausible model of unconditional trust dynamics, which simulates trust building based on familiarity, but which can be modulated by painful and gentle touch. The model displays interactive behaviour by being able to realistically control pupil dynamics, as well as determine approach and avoidance motivation.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115246728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ayane Hisatomi, Yutaka Ishii, T. Mochizuki, Hironori Egi, Yoshihiko Kubota, H. Kato
This paper describes a conversational holographic agent to help learners assess and manage their participation in order to encourage co-regulation in a face-to-face discussion. The agent works with a voice aggregation system which calculates each participant's utterances, silence ratio, ratio of participation, and turn-taking during the discussion in real-time, and produces prompting utterances and non-verbal actions to encourage learners' participation, summarization, and clarification of what they say. During discussions with each other, learner follow the prompts, and might model how the agent regulates the participation.
{"title":"Development of a Prototype of Face-to-Face Conversational Holographic Agent for Encouraging Co-regulation of Learning","authors":"Ayane Hisatomi, Yutaka Ishii, T. Mochizuki, Hironori Egi, Yoshihiko Kubota, H. Kato","doi":"10.1145/3349537.3352802","DOIUrl":"https://doi.org/10.1145/3349537.3352802","url":null,"abstract":"This paper describes a conversational holographic agent to help learners assess and manage their participation in order to encourage co-regulation in a face-to-face discussion. The agent works with a voice aggregation system which calculates each participant's utterances, silence ratio, ratio of participation, and turn-taking during the discussion in real-time, and produces prompting utterances and non-verbal actions to encourage learners' participation, summarization, and clarification of what they say. During discussions with each other, learner follow the prompts, and might model how the agent regulates the participation.","PeriodicalId":188834,"journal":{"name":"Proceedings of the 7th International Conference on Human-Agent Interaction","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122547062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：