Kyungdo Kim

Kyungdo Kim is a fourth-year Ph.D. candidate in Biomedical Engineering at Duke University, where he leads the TULIP project, a multimodal AI system for quantitative assessment of neurological disorders, including Parkinson's disease. His research integrates markerless multi-camera 3D pose estimation, EHR modeling with heterogeneous disease progression, and explainable AI to bridge the gap between clinical scoring and continuous behavior measurement. He has authored multiple publications spanning diverse healthcare data modalities including vision, audio, biosignals, clinical NLP, and EHR, with a first-author CVPR 2024 paper among them. Beyond academia, Kyungdo interned at Samsung Research America during his Ph.D. and previously co-founded a hardware startup as CTO, building IoT and AI systems from the ground up. His long-term vision is a "physical human to digital human" framework that captures and quantifies the full spectrum of human data across movement, physiology, and behavior to enable personalized healthcare.