Chengzhang Yu

Chengzhang Yu

he/him

Chengzhang Yu is a Ph.D. candidate at South China University of Technology, specializing in Large Language Models (LLMs), Multimodal Medical Artificial Intelligence, and Agent & Multi-Agent Systems (MAS). His research focuses on advancing AI-driven clinical decision-making and intelligent autonomous systems through multimodal learning and collaborative agents.

Publications

Journal Articles

DenseGCN: A multi-level and multi-temporal graph convolutional network for action recognition

Published in IET Image Processing, 2023

This paper proposes DenseGCN, a novel multi-level and multi-temporal graph convolutional network for skeleton-based action recognition that effectively captures both spatial and temporal dependencies in human motion sequences.

Recommended citation: Yu, C., et al. (2023). "DenseGCN: A multi-level and multi-temporal graph convolutional network for action recognition." IET Image Processing. 17(11), 3299-3312.
Download Paper

Conference Papers

A Self-Supervised Pressure Map Human Keypoint Detection Approach: Optimizing Generalization and Computational Efficiency Across Datasets

Published in IEEE Conference, 2024

This paper presents a novel self-supervised approach for human keypoint detection using pressure maps, achieving improved generalization and computational efficiency across different datasets without requiring manual annotations.

Recommended citation: Yu, C., et al. (2024). "A Self-Supervised Pressure Map Human Keypoint Detection Approach: Optimizing Generalization and Computational Efficiency Across Datasets." IEEE Conference Proceedings. DOI: 10.1109/10447055.
Download Paper