Waseda WeeklyWaseda Weekly

Taking on the challenge of developing content production support technology that can be used in manufacturing

Graduate School Graduate School of Fundamental Science and Engineering Doctoral Program, March 2024 Graduate, Kouki Madono

In graduate school, I am conducting research into supporting content creation. In particular, during my doctoral studies, I have been pursuing technology to digitally reproduce the quality of ink paintings by artists. In recent years, various software programs have made it possible to draw beautiful pictures, but there was an issue of not being able to reproduce ink paintings, which are characterized by bleeding and smudges.

In my research, I proposed a method to build a database of images of drawings made by a robot and learn how to paint ink paintings from that database. When I actually used this method to paint a drawing, I found that it produced ink paintings that were more beautiful than those produced by software commonly used by artists today (Expresii, Adobe Fresco, Rebelle6 pro) or existing digital technology (DeepBrush).

A camera is used to capture the robot drawing.

This research was academically recognized, having been accepted for publication as a thesis and winning an award for excellence, but many people, including professors at art universities and doctoral students from other departments, were also interested in it, saying, "Being able to draw beautiful pictures seems really interesting." Graduate School of Fundamental Science and Engineering has many researchers who specialize in computer graphics (CG) and an environment where students can discuss ideas for coming up with technical solutions, which was useful for my research on "how to draw beautiful ink paintings."

The second from the right, "(e)Ours," is the method proposed by Madono. The bleeding and smudges are more clearly expressed than with other software.

My motivation for doing this research goes back to my first year as a master's student. During a joint research meeting on campus, Research to colorize black and white images I was impressed by what one of the group, Dr. Simosella Edgar (Faculty of Science and Engineering Associate Professor), said in her lecture. Here are some of his words.

1. Research is based on lots of hard work, not genius inspiration.
2. Success is built on mountains of failures.
3. Pick up hobbies while you can.
4. Take risk and aim high.
5. Know when to know and when to rest.
6. Be open to new experiences.

Until then, I had been studying various fields blindly, but after hearing this talk, I realized that in order to do interesting or good research, it is important to first set high goals and to value your own free sensibilities during the trial and error process. After listening to the lecture, I decided to find one really interesting problem and focus on solving it, and I think that has led me to where I am today.

After completing my doctorate, I will continue my research as a fixed-term lecturer in the School of School of Fundamental Science and Engineering. In my doctoral research, I used robots, but in the future, I would like to collaborate with talented artists and art university students to conduct research into mathematically elucidating the unique drawing methods and painting styles of talented artists. I would also like to take on the challenge of developing content production support technology that can be used in all kinds of manufacturing, including watercolor painting and sculpture, in addition to ink painting.

Finally, I would like to introduce "Early Bird," which has been a great support for my research life. Faculty of Science and Engineering has a one-year program called "Early Bird" that allows excellent doctoral students to interact with each other while receiving research funding. In the first half of the year, students introduce themselves and their research, and in the second half, they invite famous professors to give lectures and students find themes together and hold workshops.

The purpose of the event is to promote personnel exchanges between young researchers and research in different fields, so you can have useful experiences that you cannot get in regular laboratory activities, such as attending lectures and presentations by other members. I would be very happy if this article sparks an interest in Early Bird's activities.

Left photo: A lecture on how to write research funding applications. It was very helpful because I rarely have the chance to learn this. It was also a good opportunity to learn that writing styles differ depending on the research field.
Right photo: At the workshop. Listening to presentations from different research fields, I was able to encounter new ways of setting research objectives and interesting proposal methods that I don't usually see. I also realized that when explaining my research to people from different research fields, it is necessary to communicate more clearly than I thought.

Daily Schedule

• 09:30 Wake up, make coffee (late nights mean late mornings)
• 10:30 Research
• 11:30 Doing chores (I try to get chores done at this time so I can focus on my studies in the afternoon)
• 12:30 Lunch
• 13:00 Seminar (Since this is an information lab, many people participate online)
• 16:30 Study (I use a standing desk to help me concentrate and take breaks to get some fresh air)
• 21:00 Break (I sometimes have dinner, watch TV, or read, but I usually just sit around doing nothing)
• 23:00 Bathing and stretching
• 24:00 Research (I remotely access the lab's server and run programs to conduct experiments while I sleep. Since the number of computers in the lab is limited, I try to use them when I have free time.)
• 25:00 Bedtime