See the world.
Create what’s next.
I work at the intersection of computer vision, generative AI, and human–computer interaction. My research spans image and video generation, large-scale model training, and human motion understanding.
I received my Ph.D. from Tokyo Institute of Technology in 2022. Previously, I worked at NAVER and NAVER Cloud, and was a visiting scholar at Carnegie Mellon University.
Explore my backgroundSelected work
IDEAS INTO SYSTEMSFOUNDATION MODELS · TECHNICAL BLOG
Sarashina Image 2.6B
A Japanese–English text-to-image foundation model with a diffusion transformer trained from scratch. Read the technical blog about the model and its development.
ICCV · HIGHLIGHT
ZIM: Zero-Shot Image Matting for Anything
Fine-grained image matting with visual prompts, connecting zero-shot segmentation to precise masks.
ECCV WORKSHOP · SYNTHETICDATA4CV
Synthetic data, closer to the real world
Diffusion-based synthetic dataset generation for egocentric 3D human pose estimation.
UIST 2020 · HUMAN MOTION
MonoEye
Multimodal human motion capture using a single ultra-wide fisheye camera.
WACV 2020 · EFFICIENT VISION
MoVNect
Lightweight 3D human pose estimation through teacher–student learning.
Experience
Full CV ↗JUL 2026 — PRESENT
Manager CURRENT
Noetra株式会社
AUG 2025 — PRESENT
Computer Vision Lead Research Engineer
SB Intuitions
Research and development of image-generation models.
JAN 2023 — JUN 2025
Research Engineer / Team Leader
NAVER Cloud
Generative AI and virtual human agents. Team Leader from May 2023.
FEB — DEC 2022
Research Engineer
NAVER
Human motion capture and generative motion research.
Research interests
Image & video generation / Large-scale model training / Human pose & motion / Human–computer interaction