産業向け画像認識を中心に、異常検知や生成モデルを用いた少量データ環境での実用的な手法と、実運用を見据えたシステム化を研究しています。
Selected projects demonstrating research-driven implementation with measurable business impact.
外観検査AI 学習プラットフォーム
A cloud-based, data-centric training platform with human-in-the-loop workflows for reproducible and continuously improvable visual inspection AI
System architecture design, full-stack development, ML training pipeline integration
自動外観検査設備の実導入
Deployment of multiple automated visual inspection systems combining AI-based visual judgment and rule-based image processing under real-world constraints
System design, on-site deployment, inspection logic integration
図面解析による見積積算RPAシステム
RPA system using drawing analysis AI for automated cost estimation, from internal tool to external product
Full-stack development, AI inference engine implementation, product expansion
