UCLA朱松纯：A Cognitive Architecture for Human-Machine Teaming

人工智能前沿讲习班（AIDL）是中国人工智能学会主办并由学会副理事长谭铁牛院士牵头发起，旨在帮助学员们在短时间内集中学习这一领域的基础理论、最新进展和落地方向，并促进产·学·研相关从业人员相互交流。对于硕士、博士、青年教师、企事业单位相关从业者、预期转行AI领域的爱好者均具有重要的意义。AIDL已经成功举办三次，前三次主题分别为《深度学习如何促进AI发展》（学术主任：陶建华）、《机器学习前沿》（学术主任：周志华）和《大数据：理论与应用》（学术主任：程学旗）。

全球著名计算机视觉专家、统计与应用数学家、人工智能专家朱松纯教授受邀参加AIDL4，将于2017年9月24日在中科院自动化所作《A Cognitive Architecture for Human-Machine Teaming》，与大家分享他在人机协作方面的思考与成果！

朱松纯：《A Cognitive Architecture for Human-Machine Teaming》

摘要：The recent advances in fields, such as vision, language and learning have inspired renewed interest in academics and the public for developing general AI machines that are capable of communicating and collaborating with humans. In the first half of this talk, I will discuss my personal assessment of the current state of AI. In short, AI is entering an era of big integration, embracing six areas: vision, NLP, cognition, learning, robotics and social ethics (game theories). It calls for a unified representation and inference framework for a wide range of tasks. I propose my own solution --- the spatial, temporal and causal and-or graph (STC-AOG) as a unifying representation, and review some ongoing work in these areas. In the second part, I will discuss the layered infrastructures underneath human communication and collaborations, and propose a paradigm called communicative learning to formulate the complexity of human learning and communication, including the theory of mind, i.e. the beliefs and intents of others. Then I will discuss the challenges for constructing a cognitive architecture for human-machine communication and teaming, and the fundamental limit of learning. I will show a few examples of human robot collaborations.

简介：全球著名计算机视觉专家，统计与应用数学家、人工智能专家。美国洛杉矶加州大学(UCLA)统计系兼计算机系教授、计算机视觉与图像科学中心主任、长江学者、千人计划专家、IEEE Fellow。在计算机视觉、统计建模与推理方面发表论文140余篇。在美国获得多项国家级与世界级奖励，包括三次获得计算机视觉大会（ICCV）颁发的Marr prize和2008年国际模式识别学会（IAPR）颁发的Aggarwal prize，目前主持多项美国重大研究计划。