FACULTY
TATSUYA YATAGAWA
Graduate School of Social Data Science/Associate Professor
Field of Study
Computer Graphics
Geometry Processing
Image and Video Processing
INTERVIEW
Unique Appeal of the Social Data Science Program
First of all, I think our department, the School of Social Data Science, is distinctive in terms that the target application of data science is clear, namely the social sciences. If you are a student of an ordinary department of science and engineering, you will study common subjects in such departments, and you may not think about how you will use the techniques you learn in these subjects in your future research. In contrast, our department claims that its mission is to broaden the capability of data science techniques in applying to the problems in social studies. Therefore, once you join our department, it is imperative that you learn and consider what you want to apply the data science techniques in terms of your research and business. In addition to the above characteristics, there are various professors in our department, and their expertise ranges from social sciences to traditional science and engineering. Therefore, once you have decided what you want to use the data science techniques for, we assure you that you can study with a professor who has expertise in your purpose. Thus, in our department, you can deeply think about your goals from different perspectives throughout your student life, and this is an attractive part of our department.
Innovative education and research encouraged in the Social Data Science Program
Since my own background is in pure science and engineering, I hope that Hitotsubashi University will become a university and research institution where students and researchers can study data science and computer science from the perspective that they are interested in without being conscious of the fact that Hitotsubashi University has been traditionally known as a famous university of social studies. However, to achieve this goal, the students who will join our department may have to engage in both traditional computer science and social studies, i.e., they each must be a two-way player, which must be harder than simply studying only regular computer science or data science. For example, in a regular computer science department, you may be able to take a lecture on computer architecture, operating systems, compilers, and so forth. However, in our department, you are expected to study these subjects on your own depending on what you want to achieve in your research and business. I hope that the students are willing to such difficulty and, after graduation, they can be an expert in social data science, and furthermore, they will even step out of the area of social data science and open up a new field of research and business.
CLASS
- (U)Programming 2
- (U)Image Processing
- (U)Project-Based Learning Reserch A
- (G)Advanced Machine Learning (practice)
RESEARCH
Research Website
My research field is in between computer graphics and computer vision and aims to use computers to reproduce the physical phenomena happening in the real world. Unfortunately, in Japan, computer graphics is considered an insignificant technology that can only be applied to a narrow range of entertainment applications, such as video games and movies, and they are considered useless to solve more serious problems. However, considering one of the ultimate goals of computer graphics, i.e., to accurately reproduce all real-world phenomena within a computer, its range of applications is everything in the real world. Therefore, computer graphics technologies are essential to conducting experiments that are hard to do in the real world due to the problems of time, money, and human resources. In such a research field, I am particularly interested in the techniques to more accurately measure the appearances, textures, and shapes of real-world objects, and computational models to show the acquired data of real-world objects in a computer. These fields are rapidly developing day by day and becoming increasingly complicated. However, I myself would like to develop a new approach with a wide range of applications based on simple theories and techniques that everyone is familiar with.
Keywords
- Image and video processing
- 3D geometry processing
- Physically based rendering
- Computer-aided design
- Geometric deep learning
- Deep generative model
- Self-supervised learning