Max Robotics — Global Service Robot Directory + US Certification

About

Hyung-gun Chi is a rising researcher at the forefront of embodied AI and human-robot interaction, whose work bridges computer vision, action recognition, and tactile perception. His primary research areas include skeleton-based action recognition, spatial action localization, human pose estimation under occlusion, and multi-modal representation learning for robotics. Chi’s most influential contribution is InfoGCN++, a pioneering model that enables online skeleton-based action recognition by predicting future frames, overcoming the critical limitation of requiring complete action observation—a breakthrough for real-time applications. This work has garnered 18 citations since 2024. He also introduced the Interacting Objects dataset, a novel resource capturing object-object interactions in dynamic environments like factories and surgical robotics, which has already received 10 citations. Chi’s AdamsFormer tackles the challenging task of predicting future action locations, essential for human-robot collaboration, while his Pose Relation Transformer addresses occlusion in human pose estimation by leveraging advances in NLP. Notably, his recent work on multi-modal representation learning integrates tactile data with vision and language, pushing beyond conventional approaches to enable richer, more robust robotic manipulation. With a growing citation record and a clear trajectory toward solving real-world interaction challenges, Chi is shaping the future of intelligent, context-aware robotic systems.