报告题目：认知机器人学：从婴儿到机器人与人工智能

报告时间：2025年10月27日（周一）9：00

报告地点：文理楼290

内容摘要：

本次讲座将介绍认知机器人学这一跨学科领域，该领域通过整合人工智能、认知科学和生物学的知识来推动机器人智能的发展。这一研究范式促进了机器人专家、计算机科学家、心理学家和神经科学家之间的紧密合作。讲座以语言学习为核心案例，展示这一研究方法。发展心理学研究奠定了理论基础，揭示了具身性在认知与语言中的关键作用——这一理论被称为具身认知。这些发现直接指导了机器人语言能力的设计与人机协作的实现。研究采用发育机器人学方法，并通过iCub和Pepper机器人的实验展示成果。这些研究揭示了具身偏见如何影响早期词汇习得与语法学习。同一方法使机器人能够通过指称手势和手指计数来学习数字等抽象概念。讲座还将介绍一个研究心理理论及其与信任关系的新型发育机器人学模型。该工作同时考察人类对机器人的理解，以及机器人用于推断人类意图的人工心理理论。结果表明，意图透明且机器人决策可解释能显著增强信任与协作。讨论将延伸至该方法对具身人工智能和伴侣机器人的应用前景，同时涉及相关的哲学思考、可信人工智能的伦理问题，以及当前大语言模型的局限性。

报告人简介：

Angelo Cangelosi教授现任曼彻斯特大学（英国）机器学习与机器人学教授，并担任曼彻斯特机器人与人工智能中心的联合主任及创始人。他已成功获得欧洲研究理事会高级资助（由英国研究创新署资助）。其主要研究领域涵盖认知与发展机器人学、神经网络、语言具身化、人机交互与信任机制，以及健康与社会护理领域的伴侣机器人。作为项目协调员/首席研究员，他累计牵头获批科研经费超过4000万英镑，其中包括欧盟高级研究基金eTALK、英国研究创新署TAS信任节点与CRADLE繁荣计划、美国空军研究实验室CASPER++项目，以及众多地平线计划和玛丽·居里学者项目。Angelo Cangelosi教授已发表学术论著300余篇，现任《Interaction Studies》与《IET Cognitive Computation and Systems》期刊主编，并于2015年担任《IEEE Transactions on Autonomous Development》主编。他曾主持多项国际学术会议，包括2022年布里斯托尔ICANN会议和2021年北京ICDL会议。其著作《发展机器人学：从婴儿到机器人》（MIT Press）于2015年问世，已被译成中文和日文。与浅田稔合编的《认知机器人学》（MIT Press）于2022年最新出版。

Academic Report Notice of Angelo Cangelosi:Cognitive Robotics: From Babies to Robots and AI

Speaker: Angelo Cangelosi

Title:Cognitive Robotics: From Babies to Robots and AI

Time: 9:00 AM, October 27, 2025(Monday)

Location:290 Arts and Science Building

Abstract:  

This talk introduces Cognitive Robotics, an interdisciplinary field integrating AI, cognitive science, and biology to advance robotic intelligence. This paradigm fosters close collaboration between roboticists, computer scientists, psychologists, and neuroscientists. Using language learning as a central case study, the talk demonstrates this approach. Research in developmental psychology provides the foundation, revealing the critical role of embodiment in cognition and language—a theory known as embodied cognition. These findings directly inform the design of linguistic capabilities and human-robot collaboration in robotics. The research employs Developmental Robotics methods, showcased through experiments with the iCub and Pepper robots. These studies reveal how embodiment biases shape early word acquisition and grammar learning. The same methodology enables robots to learn abstract concepts, such as numbers, through pointing gestures and finger counting. A novel developmental robotics model investigating Theory of Mind and its link to trust is also presented. This work examines both human understanding of robots and a robot's Artificial Theory of Mind for inferring human intentions. Results demonstrate that transparent intentions and explainable robot decisions significantly enhance trust and collaboration. The discussion extends to the implications for embodied AI and companion robots, while also addressing associated philosophical questions, ethical issues in trustworthy AI, and the limitations of current large language models.

Personal Introduction:

Angelo Cangelosi is Professor of Machine Learning and Robotics at the University of Manchester (UK) and co-director and founder of the Manchester Centre for Robotics and AI. He was selected for the award of the European Research Council (ERC) Advanced grant (UKRI funded). His research interests are in cognitive and developmental robotics, neural networks, language grounding, human robot-interaction and trust, and robot companions for health and social care. Overall, he has secured over £40m of research grants as coordinator/PI, including the ERC Advanced eTALK, the UKRI TAS Trust Node and CRADLE Prosperity, the US AFRL project CASPER++, and numerous Horizon and MSCAs grants. Cangelosi has produced more than 300 scientific publications. He is Editor-in-Chief of the journals Interaction Studies and IET Cognitive Computation and Systems, and in 2015 was Editor-in-Chief of IEEE Transactions on Autonomous Development. He has chaired numerous international conferences, including ICANN2022 Bristol, and ICDL2021 Beijing. His book “Developmental Robotics: From Babies to Robots” (MIT Press) was published in January 2015, and translated in Chinese and Japanese. His latest book “Cognitive Robotics” (MIT Press), coedited with Minoru Asada, was recently published in 2022.