Kihoon Lee

Chung-Ang University, Korea Advanced Institute of Science and Technology

Papers

Total Citations

H-Index

About

Kihoon Lee is a researcher whose work spans intelligent fault diagnosis and bipedal robotics, demonstrating a unique ability to bridge data-driven machine learning with mechanical design. His most impactful contribution is the development of a **Multi-Objective Instance Weighting-Based Deep Transfer Learning Network** for intelligent fault diagnosis (2021, 33 citations). This work addresses a critical industrial challenge: the scarcity of large-scale labeled data in manufacturing. By innovating a transfer learning framework that weights training instances to overcome domain shifts, Lee has provided a practical, high-accuracy solution for predictive maintenance, directly improving the health management of industrial processes. Earlier in his career, Lee explored the fundamentals of legged locomotion with the design of a **4-joint, 3-link biped robot** (2002, 9 citations). In this work, he demonstrated that stable walking could be achieved with a minimal mechanical structure, introducing novel gaits such as four-crossing and crawling. This foundational research in minimalist, dynamically stable walking robots showcases his versatility and early contribution to the field of robotics. Lee’s work is notable for its clear focus on solving real-world problems—from ensuring machinery reliability to simplifying robotic locomotion.