Title: Human-Agent Interaction: Problems and the solutions
Human-Agent Interaction (HAI) has widespread aspects in engineering and scientific research fields like artificial intelligence, user interface, machine learning, robotics, cognitive science and social psychology. There are some significant research objectives for HAI which have not been attacked in the traditional research area, and various solutions to them have been developed thus far. In this talk,
first of all, I try to figure out the original and important research perspective of HAI. It includes design of agent’s appearance and function, investigating and modeling how a human behaves against an artifact with agency, how expressions from an agent should be designed in engineering, how to determine physical properties of an agent. After discussions on the problems in the current HAI research, I introduce our various collaborations as examples of the concrete solutions to them. Our studies on adaptation gap, Artificial Subtle Expressions, Peripheral Cognition Technology, a cognitive model for human adaptation to an agent/non-agent, bias on human adaptation to agent’s behaviors are presented and I discuss the relationship between the studies and the problems in HAI. As concluding remarks, I summarize and position our work in an aspect of HAI research and to present insights on future research objectives of HAI not yet achieved.
Seiji Yamada is a Professor of the National Institute of Informatics, SOKENDAI, and also a visiting professor of Tokyo Institute of Technology. He received B.S. (1984), M.S. (1986) and the Ph.D. (1989) degrees in artificial intelligence, machine learning from Osaka University. He is one of pioneers of Human-Agent Interaction and has been developed various impressive studies which are related to
HAI. His current research interests are mainly in the Human-Agent Interaction and intelligent interactive systems.
Hiroyuki Itoh is the CEO of Crypton Future Media, Inc. He was born in 1965. After graduating Hokkai Gakuen University of Economics, he had worked for Hokkaido University. Then he launched the Crypton Future Media, Inc in Sapporo, 1995, which is best known worldwide for production and sales of speech synthesis software for computer music, e.g., this company has more than 50 business partners in US, Europe areas and around the world, and sells more than 10 million sound contents based on the licenses in Japan.
Especially, the release of the first character vocal series “Hatsune Miku” greatly raised this company’s profile. In recent years, this company is focussing on research and development of DTM software, music distribution service, applications for a mobile phone, 3D graphics and contents retrieval, under their slogan “Sound the future”.
Title: Human augmented cognition based on intention recognition
Human augmented cognition refers to the expansion of innately limited brain cognitive capabilities such as memory function, attention span and decision making. Human intention recognition is crucial for implementing an efficient human augmented cognition as well as human computer interaction. In this talk, I introduce a new framework for human augmented cognition based on understanding of human explicit and/or implicit intention. We are developing a glass type cognitive assistant device so called “visualizing earphone” and “hearing glass” which are used for not only recognizing human intention but also providing useful visual and audio information according to a situation. The visualizing earphone using two cameras and a headphone set in a glass interprets both human’s intention and outward visual surroundings based on incremental feature extraction, selective attention, gaze analysis, hierarchical object categorization and face recognition algorithms, and translates visual information into a audio signal so that the limited capacity of human cognition such as memory, inference and decision can be expanded. The hearing glass using two microphones and a tiny display monitor unit in a glass can transform audio signals into image and video information, which also helps human cognition in complex environment. Also, the humanoid robot NAO together with Microsoft KINECT is used to provide a suitable cognitive service according to human intention in an indoor environment. Experimental results with real data have demonstrated the validity of the proposed approach.
Minho Lee received the Ph.D. from Korea Advanced Institute of Science and Technology (KAIST) in 1995, and is currently a professor of School of Electronics Engineering, Kyungpook National University, Daegu, Korea. He established Mobile Technology Commercial Center at Daegu, and worked for Education & Training Department as a director from 2005 to 2006. Also, he was a visiting scholar for Dept. of Brain and Cognitive Science at MIT from 2006 to 2007. He received several best paper awards at international conferences including ICONIP (2007 and 2009), IDEAL(2008), ICAISC(2006) and ISABEL(2006). He has been served for several international journals (neural networks, nature intelligence, etc.) as an associate editor and also for international conference as program chairs, special session chairs, and now he is president for Asia-Pacific Neural Network Assembly (APNNA) and general chair for 20th international conference on neural information processing (ICONIP2013) at Daegu, Korea. His research interests include brain-neuroinformatics, biologically inspired vision systems, human augmented cognition, selective attention, autonomous developmental system based on action and perception cycle learning, brain-machine interaction and intelligent sensor systems.