沼尾 正行 教授Numao, Masayuki

Graduate School of Information Science and Technology
Department of Information and Physical Sciences, Architecture for Intelligence Lab (Institute of Scientific and Industrial Research)

<!--:ja-->沼尾 正行 教授<!--:--><!--:en-->Numao, Masayuki<!--:-->

Professor Numao is working on developing artificial intelligence for computers that can empathize and learn. His goal is to create computers that can co-exist alongside humans. This is not to say computers should be entrusted with any and all tasks. The defining characteristic of computers is that they serve to provide ease and convenience for humans.

A computer that composes and arranges music especially for a listener

Our “computer that can empathize” is a system that can compose and arrange music specifically for a person. It measures that person’s brainwaves and pulse and composes and arranges music tailored to those measurements. While it is difficult to evaluate which songs a person may like, human emotion can be measured in several objective and subjective ways, such as with psychological behavior markers. Our examination and evaluation of these factors is ongoing. Subjective elements beyond brainwaves and a pulse are also important, and sometimes we have our research subjects fill out surveys. I am also furthering my research with researchers in other fields as I work on systems that create sensory stimulus such as odors and lighting matched to individuals.


Mutual understanding is an important theme in artificial intelligence research

I think the appeal of interdisciplinary research will colleagues from other fields is the chance to talk about things that are new to one another. However, the difficulty then lies in being able to understand each other! Technical vocabulary is different, as are the worlds we each experience every day. What is objective to one person may be a subjective matter to another. It is always interesting to see how researchers from different fields come to understand one another.
In fact, my own research is closely connected to these ideas of mutual understanding. In the field of artificial intelligence, there is much discussion of how to connect machines with the processes whereby people come to understand one another. What this means is, to me, the process of interdisciplinary research itself becomes material for my own research.

Interdisciplinary research is a mutually stimulating opportunity for growth

numao3As an example of activity in the Humanware Innovation Program, some life sciences research students come to my lab as part of their research lab rotation. Their strength lies in experimentation using live organisms. My experiments sometimes involve the measurement of brain waves, and so for the information science students in my lab, I expect that the insights of the life sciences students could prove to be very valuable, and their interaction would be a mutually stimulating experience.
I advise students that undertake the challenge of integrated research with those in other fields to increase the time they work together, and to make an effort to fill in any gaps in communication that may exist. I also actively support their participation in joint research projects with researchers in other universities, regardless of their field of study.

Researchers must broaden their interests while maintaining a firm grasp of their own discipline

I have an important request for students interested in interdisciplinary research: while it is good to broaden your interests and try new things, it is also very important to have a firm grasp of the fundamentals in your field of expertise, which you then need to develop into a mastery that cannot be matched by others. Both of these aspects are important. When focusing on your own field of study, your perspective tends to become narrower and narrower. However, if you follow your interests into interdisciplinary research projects, you will find it difficult to produce good results if you do not have a firm grasp of the fundamentals of your own field. As a researcher, first and foremost you must have a solid grasp of where your own expertise lies.




沼尾 正行 教授Numao, Masayuki

Numao, Masayuki

情報科学研究科 情報数理学専攻 知能アーキテクチャ講座(産業科学研究所)

<!--:ja-->沼尾 正行 教授<!--:--><!--:en-->Numao, Masayuki<!--:-->