Current Research

This study begins with the observation that monocular 3D human pose estimation (3D HPE), widely used for assessing construction workers’ postures, inherits the joint definitions and average anthropometric statistics encoded in benchmark training datasets. As a result, when a worker’s body proportions deviate from the training distribution, the estimated skeleton may exhibit segment-length inconsistencies and depth-scale ambiguity—manifesting as anthropometric error. To address this issue, we develop a Personalized Skeleton Modeling post-processing framework that explicitly incorporates worker-specific body segment lengths. The proposed approach evaluates multiple candidate depth (Z-axis) scales and selects the one that simultaneously minimizes bone-length residuals and pose distortion, thereby transforming bone-length constraints from a simple output-stabilizing heuristic into a mechanism for correcting dataset-driven anthropometric bias.

• J. JEONG, S. SHIN, D. LEE, Worker Customized Stress Monitoring through Body Composition Analysis and Wearable Bio-Sensor, The 10th International Conference on Construction Engineering and Project Management (ICCEPM), Seogwipo, Jeju, South Korea, 2025. (Best Paper Runner-Up)

• J. JEONG, H. YOUNG, D. LEE, Thermal Imaging-Based Prediction of Facial Temperature Distribution for Fatigue Monitoring in Construction Workers, in: Proceedings of the 2025 Spring Conference of the Korea Institute of Building Construction (KIC), Gyeongju, South Korea, 2025.

• J. JEONG, D. KIM, D. LEE, A Study on Utilizing Carbon Dioxide Concentration During Respiration to Improve the Accuracy of Predicting Worker Stress, in: Proceedings of the 2024 Annual Conference of the Korean Institute of Construction Engineering and Management (KICEM), Seogwipo, Jeju, South Korea, 2024. 

• S. LEE, J. JEONG, D. LEE, Automated Quality Control Using UGVs and LiDAR in Construction Site, in: Proceedings of the 2023 Annual Conference of the Korean Institute of Construction Engineering and Management (KICEM), Goseong, Kangwon, South Korea, 2023.