Kaoru Arakawa

Home  >>  刊行物について  >>  ロールモデル集  >>  Kaoru Arakawa


I have Lived my life without ever compromising my goals.

Kaoru Arakawa
Professor of Interdisciplinary Mathematical Science
Department of Frontier Media Science

“Early Childhood to Student Days”

 My interest in mathematics and physics goes back to my high school days. After entering the University of Tokyo, I was torn between science and engineering. Eventually I decided to pursue my studies in the Faculty of Engineering’s Department of Electronic Engineering, drawn by the notion of “turning ideas into things.” I was only the sixth female student ever to study in that department. I went on to pursue a doctoral degree at the Graduate School of Engineering of the University of Tokyo, and while there I did a stint at the California Institute of Technology (Caltech) in the United States as a Fulbright Scholar. After returning to Japan, I completed my doctoral studies and earned a Ph.D. in engineering in 1986.
 My graduate studies focused on information and communication engineering, digital signal processing in particular. At Caltech, I was involved in researching the application of digital signal processing to EEG analysis.
 At most academic conferences I participated in, there were very few women aside from myself. This was not a problem in itself, but I was often the object of curiosity.
 I married during the first year of my doctoral program. My area of study did not involve experiments and I did not have to stay at the university until late at night, so I was able to squeeze in household chores in between my research work.

“Working Life to Present”

 After completing the doctoral program I became a researcher at the Japan Society for the Promotion of Science under the Research Fellowship for the Young Scientist Program. Later, a former graduate school supervisor offered me a job as an assistant in the Faculty of Engineering of the University of Tokyo, and I accepted. But being a temporary position, I soon began looking for another job. I wanted to take a job in academia that allowed the freedom, relatively speaking, to pursue my own research interests, but it was rare for a woman to teach at a university back then. To make matters more difficult, I had a small child at that time, and finding a university that would hire me was a challenge.
 Fortunately, this was around the time when the age of information technology was just beginning and a growing number of universities were creating departments on computer science and engineering. This meant they needed to hire a large number of researchers with a Ph.D. in the field of computer and information studies as faculty members. Meiji University was one such university. Its School of Technology, renamed the School of Science and Technology in 1989, offered me a faculty position in its newly-established Department of Computer Science.
 In the following years I became a successful researcher, being promoted to associate professor in 1992 and professor six years later. In 2013, I was transferred to the School of Interdisciplinary Mathematical Science, where I now teach as a professor in the Department of Frontier Media Science. Over the years I relied on daycare, babysitters, and later on, after-school programs to care for my children. In 1994, I left Japan for the United States with my young son to participate in a foreign residency research program at the University of California, Santa Barbara. I placed my son in a local elementary school and he spent his afternoons in an after-school program offered by the university. After returning to Japan, I gave birth to another son and again I had to rely on outside childcare support, such as daycare and babysitters. But by then personal computers were smaller and more capable, and broadband networks widely available, so that working at home had become more feasible. I marveled at how easier things had become compared with earlier years when my elder son was a child.

“Research Descriptions and the Thrills of Work”

 My specialty is digital signal processing. More specifically, I analyze signals perceived by humans, such as images or sounds, and extract data from these signals or process them using a computer. In information and communications engineering, much has been achieved in terms of technologies to represent signals in mathematical models and design mathematically optimal processing systems. Mathematical processing alone, however, cannot effectively reflect human perceptions or subjective attitudes towards these signals. I am now studying a new signal processing technique that takes into account human senses and subjective attitudes.
 Since around the year 2000, my focus has been on digitally enhancing photographic facial images by removing skin imperfections, such as blemishes and wrinkles, and improving complexions. This technology involves classifying and retouching skin elements to remove blemishes and wrinkles from the image. I am also working on an image processing technique that tightens facial contours or sharpens facial features, which makes the skin texture appear more refined and the face lifted and firmer, just as an aesthetic treatment does. The base of this technique is a mathematical model of the face and the skin, but we also take into account human senses and subjective perceptions in processing the facial or skin elements. Developed in partnership with a commercial company, this technology has been released as a smartphone app; it is also used in “puri-kura” photo booth machines.
 Digital processing of music is another area of research interest. I am currently investigating ways to embed extra data, such as intellectual property rights data, into musical signals in a way that is undetectable to humans. This technology takes into account human hearing properties. Embedding data into musical signals can alter the sound, making it unpleasant at times. By manipulating sound signals to create sounds pleasant to the human ear, data embedding can be accomplished without degrading the quality of the music.
 Most recently, I have been studying ways for a machine to estimate the most desirable designs for an individual. Again, this technology is based on a mathematical technique called optimum estimation theory. By integrating this theory with technology for humans to interact with computers, we are able to display designs that best meet an individual’s subjective preferences.
 The thrill of research lies in creating something completely new and innovative and having a great number of people use it to make their lives more comfortable and emotionally fulfilling. My ultimate goal is to develop an artificial intelligence that understands the human mind and sensitivities, one that is capable of presenting the most desirable designs or suitable features for any individual.

Messages to Students and Young Researchers

 Looking back on my life thus far, I can say one thing for certain: If there is something you wish to accomplish with your life, you should not give it up easily, no matter what obstacles you face. Your aspirations may seem unusual or unrealistic at that particular moment, but times change. There are plenty of instances where the unconventional becomes the norm in only a few decades’ time. When I first decided to study electronic engineering, people often asked me why a woman like me would be attracted to such a field. These days, no one thinks twice about women choosing to study engineering. In fact, women in this area are valued exactly because there are so few of us. I had aspired to teach at a university at a time when female university professors were a rarity; years passed, and without my being aware of it, times had shifted and the prospects for women to become university professors had improved significantly. My experience tells me that if you hold on tight to your dreams, there is every possibility that the world around you will change and that the problems you may have faced will no longer exist, all in good time.