Z
Zeyuan Hu
Publications - 6
Citations - 21
Zeyuan Hu is an academic researcher. The author has contributed to research in topics: Convolutional neural network & Deep learning. The author has an hindex of 2, co-authored 6 publications receiving 11 citations.
Papers
More filters
Journal ArticleDOI
Dual Attention-Guided Multiscale Dynamic Aggregate Graph Convolutional Networks for Skeleton-Based Human Action Recognition
Zeyuan Hu,Eung-Joo Lee +1 more
TL;DR: This paper proposes a dual attention-guided multiscale dynamic aggregate graph convolution neural network (DAG-GCN) for skeleton-based human action recognition and exploits the relationship of joints hierarchically and the spatial temporal correlations through two modules.
Proceedings ArticleDOI
Human Motion Recognition Based on Improved 3-Dimensional Convolutional Neural Network
Zeyuan Hu,Eung-Joo Lee +1 more
TL;DR: A new time transition layer that models variable temporal convolution kernel depths is put forward that embeds this new Hybrid Model in the proposed 3D CNN, and the DenseNet architecture is extended with 3D filters and pooling kernels.
Journal ArticleDOI
Human Motion Recognition Based on Spatio-temporal Convolutional Neural Network
TL;DR: The results show that the improved convolutional neural network has a significant improvement in improving the recognition rate and has obvious advantages in action recognition.
Journal ArticleDOI
Improved DT Algorithm Based Human Action Features Detection
TL;DR: The improved Dense Trajectories(iDT) algorithm is used to optimize and extract the optical flow features in the movement of human action, then it will be combined with Support Vector Machine methods to identify human behavior, and the image will be used in the KTH database for training and testing.