Jinhui Yi

Papers

Total Citations

H-Index

About

Jinhui Yi is a researcher at the forefront of computer vision and human motion analysis, with a primary focus on skeleton-based action recognition and trajectory forecasting. His most significant contribution is the development of the Pose Refinement Graph Convolutional Network (PR-GCN), a novel architecture that addresses fundamental limitations in how graph convolutional networks process skeleton data for action recognition. This work, published in 2021 and accumulating 58 citations, introduces a pose refinement mechanism that enhances the spatial representation of skeletal joints, leading to more accurate and robust recognition of human actions. Yi’s research extends to autonomous systems through his work on the Spatial-Temporal Consistency Network for low-latency trajectory forecasting, which tackles the critical challenge of modeling spatial interactions between moving objects in real-time scenarios. By moving beyond frame-to-frame motion modeling, this approach enables safer navigation for autonomous vehicles and mobile robots. With a total of over 80 citations across his key publications, Yi’s work bridges the gap between graph-based deep learning and practical applications in human-computer interaction and autonomous navigation, establishing him as an emerging voice in the field of spatiotemporal modeling.