Zheng Ge

Bournemouth University, ShanghaiTech University

Papers

Total Citations

H-Index

About

Zheng Ge’s research bridges the gap between human-centered sensing and autonomous perception, with a focus on human activity recognition (HAR) and 3D object detection for robotics and autonomous driving. His most cited work, a 2021 paper on a novel attention-based convolutional neural network for HAR, has garnered 30 citations and addresses critical challenges in smart homes, healthcare, and human-robot interaction by fusing CNN architectures with attention mechanisms to improve recognition accuracy. In 3D perception, Ge introduced **PersDet** (2022), a monocular detection method that operates directly in perspective bird’s-eye-view, eliminating the need for specialized feature-sampling operators and enabling deployment on edge devices—a practical breakthrough for real-world autonomous systems. His more recent work, **WildRefer** (2024), tackles the complex task of 3D object localization in large-scale dynamic scenes using multi-modal visual data and natural language, pushing the boundaries of human-robot communication. With a growing citation footprint and contributions that prioritize both algorithmic innovation and real-world deployability, Zheng Ge is establishing himself as a versatile researcher advancing the frontiers of embodied AI and intelligent perception.