Search

Understanding the emergence of symbol systems, especially language, requires the construction of a computational model that reproduces both the developmental learning process in everyday life and the evolutionary dynamics of symbol emergence throughout history. This study introduces the collective predictive coding (CPC) hypothesis, which emphasizes and models the interdependence between forming internal representations through physical interactions with the environment and sharing and utilizing meanings through social semiotic interactions within a symbol emergence system. The total system dynamics is theorized from the perspective of predictive coding. The hypothesis draws inspiration from computational studies grounded in probabilistic generative models and language games, including the Metropolis-Hastings naming game. Thus, playing such games among agents in a distributed manner can be interpreted as a decentralized Bayesian inference of representations shared by a multi-agent system. Moreover, this study explores the potential link between the CPC hypothesis and the free-energy principle, positing that symbol emergence adheres to the society-wide free-energy principle. Furthermore, this paper provides a new explanation for why large language models appear to possess knowledge about the world based on experience, even though they have neither sensory organs nor bodies. This paper reviews past approaches to symbol emergence systems, offers a comprehensive survey of related prior studies, and presents a discussion on CPC-based generalizations. Future challenges and potential cross-disciplinary research avenues are highlighted. [ABSTRACT FROM AUTHOR]

It is well known that, in human communication, physical contact, such as holding hands, has the effect of relieving stress and providing a sense of intimacy. In this study, we verified whether walking hand-in-hand has a positive effect on relationship building between children and robots. Specifically, an interaction experiment was performed in which a child and a robot play one-on-one for approximately 30 min with 37 children aged 5–6 years. The robot is teleoperated by a nursery teacher in this experiment. The children are divided into two groups: the experimental group, in which children walk hand-in-hand during their first encounter with the robot, and the control group, in which children do not have any physical contact with the robot. The change in the interaction is analyzed while taking into consideration the distance between the child and the robot, eye contact rate, and a questionnaire completed by the parents and children. The results reveal that the children in the experimental group interacted significantly with the robot. Moreover, the parents of the children in the experimental group tended to feel that their children appeared to experience intimacy with the robot. These results suggest that walking hand-in-hand has a positive effect on child-robot relationship building. [ABSTRACT FROM AUTHOR]

Based on the little big-five inventory, we developed a technique to estimate children's personalities through their interaction with a tele-operated childcare robot. For personality estimation, our approach observed not only distance-based but also face-image-based features when a robot interacted with a child at a close distance. We used only the robot's sensors to track the child's positions, detect its eye contact, and estimate how much it smiled. We collected data from a kindergarten, where each child individually interacted for 30 min with a robot that was controlled by the teachers. We used 29 datasets of the interaction between a child and the robot to investigate whether face-image-based features improved the performance of personality estimation. The evaluation results demonstrated that the face-image-based features significantly improved the performance of personality estimation, and the accuracy of the personality estimation of our system was 70% on average for the personality scales. [ABSTRACT FROM AUTHOR]