I'm Kei Katsumata.
I am a Master's student advised by Prof. Komei Sugiura at Keio University in Japan.
Click to expand
My research interests lie at the intersection of vision and language, computer graphics, and autonomous driving, with a particular focus on building intelligent systems that understand and act upon multimodal inputs.
I am especially interested in enabling agents that can interpret visual scenes and generate natural language instructions in real-world settings.
As a full-stack engineer, I have extensive experience in building scalable applications using TypeScript, Go, and modern cloud technologies. I combine my research expertise with practical software engineering skills to create impactful solutions.
About me
I am a passionate researcher and full-stack engineer at Keio University, focusing on multimodal AI systems and their real-world applications. My work combines theoretical insights from machine learning research with practical software engineering to create impactful technology solutions. I have experience in both academic research, publishing at top-tier conferences/journals like IEEE RAL, and industry work, building scalable applications using modern technologies.
- Location:Tokyo, Japan
- Age:22
- Nationality:Japanese
- Interests:Vison and Language, Autonomous Driving, Computer Vision
- Study:Keio University
- Employment:Keio University
Publications
Journals
Mobile Manipulation Instruction Generation From Multiple Images With Automatic Metric Enhancement
Kei Katsumata, Motonari Kambara, Daichi Yashima, Ryosuke Korekata, Komei Sugiura
IEEE Robotics and Automation Letters, pp. 3022-3029, DOI: 10.1109/LRA.2025.3539086, 2025
Impact Factor (2023): 4.6, h5-index: 117
International Conferences
Mobile Manipulation Instruction Generation From Multiple Images With Automatic Metric Enhancement
Kei Katsumata, Motonari Kambara, Daichi Yashima, Ryosuke Korekata, Komei Sugiura
IEEE RA-L presented at International Conference on Robotics and Automation (ICRA), 2026
Viena, Austria
Domestic Conferences
NaiLIA: 多層的な依頼文に基づくネイルデザインのマルチモーダル検索
雨宮佳音, 小松拓実, 八島大地, 是方諒介, 勝又圭, 杉浦孔明
第28回画像の認識・理解シンポジウム, 2025
国立京都国際会館, 2025年7月
物体操作指示文生成モデルに基づくモバイルマニピュレーションのためのデータセット拡張
勝又圭, 神原元就, 八島大地, 是方諒介, 杉浦孔明
第39回人工知能学会全国大会, 2025
大阪国際会議場, 2025年5月
NaiLIA: 緩和損失に基づくネイルデザインのマルチモーダル検索
雨宮佳音, 小松拓実, 八島大地, 是方諒介, 勝又圭, 杉浦孔明
第39回人工知能学会全国大会, 2025
大阪国際会議場, 2025年5月
自動評価尺度を用いた強化学習およびマルチモーダル基盤モデルに基づく物体操作指示文生成
勝又圭, 神原元就, 杉浦孔明
第42回日本ロボット学会学術講演会, 2024
大阪工業大学梅田キャンパス, 2024年9月6日
Education
High School
B.E. in Information and Computer Science
M.S. in Information and Computer Science
Work
Keio University
CyberAgent, Inc.
DMM.com LLC
looking up Inc.
Keio University AI and Advanced Programming Consortium
Standard Inc.
Coconala Inc.
Skills
Get in touch.
I am always looking for new opportunities to collaborate and learn. Feel free to reach out to me via email or LinkedIn.
- ke59ka77[at]keio.jp
- Location
- Tokyo, Japan
- Github
- agate106k