Chongyang Zhong

University of Chinese Academy of Sciences, Institute of Computing Technology

Papers

Total Citations

140

H-Index

About

Chongyang Zhong is a computer vision researcher whose work centers on human motion prediction, a critical challenge at the intersection of deep learning, spatial-temporal modeling, and graph-based neural networks. His research addresses the fundamental problem of forecasting future human movements from historical motion sequences — work with far-reaching implications for autonomous driving, robotics, and human-computer interaction. Zhong's most prominent contribution is his development of the Spatio-Temporal Gating-Adjacency Graph Convolutional Network (GCN), which advances the modeling of complex relationships between body joints and motion dynamics across both space and time. This work has garnered over 122 citations, signaling strong recognition within the computer vision community. His approach leverages adaptive graph structures to more accurately capture the nuanced interdependencies in human skeletal motion, pushing beyond the limitations of earlier, more rigid GCN formulations. Extending this line of inquiry, Zhong has also investigated stylized human motion prediction, exploring how individual movement styles can be incorporated into spatial-temporal models — an important step toward more personalized and realistic motion synthesis. Collectively, his research represents meaningful progress in enabling machines to understand and anticipate human behavior with greater precision and contextual awareness.