Fanyang Meng
Papers
2
Total Citations
12
H-Index
2
About
Fanyang Meng is a leading researcher in computer vision and human motion analysis, with a focus on advancing human-robot interaction and intelligent surveillance systems. His work centers on developing sophisticated graph-based deep learning models to understand and predict complex human behaviors. Meng’s most impactful contribution is the Dynamic Compositional Graph Convolutional Network (DC-GCN), introduced in his highly cited 2023 paper (7 citations), which revolutionized composite human motion prediction by efficiently modeling the intricate spatial-temporal dynamics of body movements. This work addresses a critical challenge in the field: enabling machines to anticipate human actions in real-time for applications ranging from autonomous driving to collaborative robotics. In his more recent 2025 publication (5 citations), Meng extended his expertise to interactive action recognition, proposing a novel "Mutual Excitation" framework that leverages graph convolution to capture the subtle, reciprocal dynamics between interacting individuals—whether hand-to-hand or human-to-human. This breakthrough is particularly significant for video surveillance and human-robot collaboration scenarios. Meng’s research consistently pushes the boundaries of how machines perceive and predict human motion, earning him recognition as a rising authority in topology-aware feature learning for interactive behavior analysis.
Research Focus
Key Achievements
Top Papers
- 1
- 2