Kei Katsumata

About

I am a Master's student in Information and Computer Science at Keio University, advised by Prof. Komei Sugiura. My research lies at the intersection of vision and language, computer vision, and embodied AI, with a focus on building intelligent systems that understand and act upon multimodal inputs — agents that interpret visual scenes and generate natural language in real-world settings.

Alongside research, I work as a full-stack engineer with experience building scalable applications in TypeScript and Go on modern cloud infrastructure, combining research insight with practical software engineering.

Research Interests

• Vision and Language — Multimodal Retrieval, Referring Expression, Instruction Generation
• Computer Vision — 3D/4D Reconstruction
• Embodied AI — Autonomous Driving, Mobile Manipulation

News

Education

• 2025–Present M.S. in Information and Computer Science, Keio University
• 2021–2025 B.E. in Information and Computer Science, Keio University
• 2018–2021 Seikei High School

Experience

• 2025–Present Research Assistant, Keio University
• 2023 Backend Engineer Intern, CyberAgent, Inc.
• 2023 Backend Engineer Intern, DMM.com LLC
• 2023–2024 Full Stack Engineer Intern, looking up Inc.
• 2022–2024 Education Internship, Keio University AIC
• 2021–2023 Data Science Intern, Coconala Inc.