China
Zhejiang University of Technology
Human object interaction, Action recognition, Feature refinement and learning, Multi-stream neural network, Human pose, Object semantic information