Spatial computing 空間
AR/VR from optics to runtime. Calibration and registration — the quiet math that makes a virtual object stay put in a real room.
KEIKEN/ ANUBHAV,ENGINEERED
[06_DOF]
Keiken (経験) - embodies the idea software comes from experience, curiosity, and continuous exploration. I build intelligent systems that perceive the world, preserve experience, and make complex ideas tangible through interaction.
scroll to register layers ↓
01 — thesis
I build instruments for seeing — hardware that overlays worlds, models that recognize motion, and memory systems that give machines a past.
One practice, three disciplines: spatial computing, machine learning, and local-first software. Whether I'm designing pipelines, training models, or building knowledge systems, the goal is the same: helping computers perceive, remember, and interact with the world more naturally.
02 — laboratory
03 — practice
Machine learning 学習
Event detection over time-series and telemetry. Small annotated corpora, honest baselines, reproducible feature pipelines.
Systems 構築
Local-first by conviction. Rust, TypeScript, and databases that live on your machine, not someone else's.
Craft 描く
Drawing, Blender, and mathematics — the hand skills that keep the engineering pointed at something worth seeing.
04 — transmission