Welcome to my portfolio website.
Here is Naoya’s portfolio website.
日本語版はこちら
Education
- Apr. 2019 - Mar. 2023: The University of Tokyo, Japan, Bachelor of Computer Science
- Mar. 2023 - current: The University of Tokyo, Japan, Master of Computer Science
Skill levels
| Level | Description |
|---|---|
| 4 | I have a lot of experience with it and I can handle it easily. |
| 3 | Although I have used it several times, it takes me a while to get used to it again because of the long blank period. |
| 2 | I have a little experience of using it, but I can use it. |
| 1 | I have little experience of using it and do not have the confidence to work primarily with it. |
Programming languages, frameworks and libraries
| Languages, frameworks, etc. | Levels |
|---|---|
| C, C++ | 4 |
| Java | 4 |
| — Swing | 4 |
| Python | 4 |
| — Pytorch | 4 |
| — Keras, Tensorflow | 2 |
| OCaml | 3 |
| C# | 2 |
| JavaScript | 2 |
| - Vue | 2 |
| TypeScript | 2 |
| Kotlin | 2 |
| SystemVerilog | 1 |
| x64 Assembly Language | 1 |
Applications
| Applications | Levels |
|---|---|
| Blender | 4 |
| Docker | 3 |
| AndroidStudio | 3 |
| Unity | 3 |
| Vivado | 1 |
Publication
Exploring Weight Balancing on Long-Tailed Recognition Problem
- Authors： Naoya Hasegawa, Issei Sato
- Conference： ICLR2024
- Links： ArXiv, OpenReview, Github
Test-time Adaptation Meets Image Enhancement: Improving Accuracy via Uncertainty-aware Logit Switching
- Authors： Shohei Enomoto, Naoya Hasegawa, Kazuki Adachi, Taku Sasaki, Shin’ya Yamaguchi, Satoshi Suzuki, Takeharu Eda
- Conference： IJCNN2024
- Links： ArXiv
Certification
- TOEFL iBT … 92 / 120 (April 2022)
Internship experience
- Aug. 2022 - Sept. 2022: I did a research internship on domain adaptation in deep learning models at NTT Laboratories (NTT研究所). For more information, please click here (sorry, this article is in Japanese).
- Aug. 2023 - Sept. 2023: I did a research internship with the entertainment team on the theme of “Application of Deep Learning to Creative Work” at Preferred Networks, Inc.
Field of interest
Machine learning，Image processing.
I particularly prefer to solve more general, practical but difficult problem setups, e.g. when the distribution of the training data is skewed (imbalanced learning).
Slides
Slides used in my circle workshops, lightning talks, etc. are available on this page. Sorry, all slides are in Japanese.
Works
My developments are summarized here.