Hitotsubashi University
School of Social Data Science
Graphics and Vision Labratory

Join us

For international students

The following descriptions are for those who plan to join our lab as a post-graduate student. If you plan to enter the under-graduate school of social data science, Hitotsubashi University, you will be required the Japanese language skill because the official language in most classes is Japanese in most Japanese universities, unfortunately.

In contrast, if you plan to join the graduate school of social data science, it would be possible to get the degree. Though most classes will be in Japanese, you can communicate with professors (and also students) only in English.

In any case, it will be a good idea to contact the lab with [CONTACT] form at first.

School of Social Data Science

Our lab belongs to Shool of Social Data Science (to start next April), Hitotsubashi University. Since Hitotsubashi University was previously the Tokyo University of Commerce and has strong research background for economics and other social sciences. Accordingly, the new school aims to open up a new research avenue from the perspectives of computer science and social studies (e.g., political, economic, juristic sciences).

Therefore, different from other science schools of Japan’s national universities, you are required to have the knowledge of both computer science and social science in the entrance exam.

About our lab

Our lab will launch April 2023, and thus, only a faculty member, Tatsuya Yatagawa, belongs to the lab currently. This means that there are no “sempai” (Japanese word to mean senior colleagues). So, we will be happy if you are a self-motivated person and can work on your study with strong enthusiasm.

Basically, students in our lab aims to publish academic research papers in academic journals and international conferences. Getting such publications in your short university life will require hard-working because the papers need to provide more or less new idea to improve previous techniques and to achieve the world’s highest quality. However, despite the hard work, you can get one of the happiest experience where your research is admitted interesting by many professional people.

The previous studies of our lab are listed in the [[RESEARCH]]https://tatsylab.github.io/en/research/ page, so please look at the page first to understand what you will be able to study about in our lab.

Life in our lab

Basically, you are expected to come to the laboratory every day (excluding weekends) and work on your research except the time for lectures. Each student will be provided with a high-spec computer required for their research and can use it exclusively. The students can also work on some server machines if they need.

Once every 1-2 weeks, the students can discuss their studies in one-on-one meeting with faculty members. In addition, we will have a meeting with all the lab members once a week, where some of the students will introduce their research and discuss that with other members. We will also have a seminar to read English textbooks or papers to share the knowledge with lab members.

The students can engage in a part-time job, but once you are assigned to a laboratory, you will need to spend most of your time on research. Therefore, you need to manage your time by yourself to steadily move your research forward.

If you want to focus on research and want to reduce the amount of part-time work, we may be able to introduce some scholarships or ask you to do some lab work (related to research) to get salary from the lab, albeit a small amount. As such financial problems are a very important issue to continue your studies, please feel free to ask with faculty members if you have any troubles.

Research theme

Our lab plans to work on a range of studies related to computer graphics (CG) and computer vision (CV) from basic theories to applications. Therefore, each student can choose their research theme by yourself as far as the topic does not deviate greatly from the field of CG/CV.

The following keywords may be related to the research theme, but the students can proposed some other topics if they prefer. In that case, they can discuss with the faculty members how the topic can be grown up to research.

  • Image and video editing (target images are various, such as landscape photos, face images, paintings, etc.)
  • Physics-based rendering (technique to create photorealistic images by simulating the transport of light rays on a computer)
  • 3D shape processing (processing data such as point clouds and polygons, aiming for application to various industries)
  • Fundamental technologies of deep learning related to the above topics (it might be rather applications than the very basics of deep learning itself)

We think that many students who are interested in our faculty and graduate school are interested in machine learning, so some keywords related to it are also listed below.

  • Data generation models (GAN, normalizing flow, diffusion model, etc.)
  • Neural networks for processing 3D shapes (point cloud convolutional networks, graph convolutional networks, etc.)
  • Learning techniques for extracting useful knowledge from a small amount of data (few-shot learning, self-supervised learning, self-prior)


After reading the above, some of you may feel that research seems difficult, or you may be somewhat anxious about whether you can make it through life in the lab. However, I myself did not graduate an computer science department to get my under-graduate degree, and I was almost a novice when I started my research about image processing after I entered a graduate school.

So far, I’ve written a little harsh, but in the end, as long as you have the “motivation”, you don not have to worry about the lab life very seriously.

If you are interested in our laboratory, we would like you to visit our laboratory to solve your concerns.

Inquiries for the laboratory visit: Contact form