About Me
I am a second-year PhD student at Morishima Lab, Waseda University, broadly interested in computer vision, especially in visual content generation and representation learning. My research centers on diffusion models, particularly on rethinking the design choices that realize them, including spatial granularity, conditioning structure, and frequency representation, to enable more controllable image generation. A driving question is how the structural design of generative models affects controllability and consistency in visual content production, which connects naturally to current challenges in image and video generation and editing. Here is my CV.
Research Interests: Diffusion Models · Generative Models for Vision · Controllable Image/Video Generation · Synthetic Pre-training · Computer Vision · Computer Graphics
Education
Doctor of Engineering Apr 2025 – Present
Master of Engineering Apr 2023 – Mar 2025
Bachelor of Engineering Apr 2019 – Mar 2023
Exchange Experience
Visiting Student Feb 2025 – Jul 2025
Working Experience
Research Internship at CyberAgent AI Lab, Tokyo, Japan Aug 2026 – Oct 2026
- Developed a pre-visualization tool powered by video generation models to support film and content production
Technical Fellow at Insight Inc., Tokyo, Japan Feb 2025 – Present
- Optimizing the performance of the ADVENTURE project using CUDA
Software Engineer (part-time) at Insight Inc., Tokyo, Japan Mar 2021 – Jan 2025
- Developed high-performance computational mechanics software for physics simulation
- Designed Graphical User Interface for computer-aided engineering (CAE) using Java and Qt (C++)
- Developed and trained generative models for seismic wave simulation on the supercomputer Fugaku
Teaching Experience
Teaching Assistant, Research Seminar on Applied Physics 2023 – 2024
- Mentored two students in building an image generation model from scratch and conducting empirical evaluations
- Coached the students in presenting their work to faculty members in class