Ruixiang Zhang

Stanford University, National University of Singapore, Harbin Institute of Technology

Papers

Total Citations

105

H-Index

About

Ruixiang Zhang is a robotics and perception researcher whose work spans autonomous navigation, climbing robotics, and intelligent sensor design. His most influential contribution is **Meta-RangeSeg** (2022, 50 citations), a deep learning framework for real-time LiDAR sequence semantic segmentation. By introducing multiple feature aggregation mechanisms, this work directly addresses the computational bottleneck of processing 3D point clouds for autonomous vehicles and robots, enabling efficient, real-world deployment without sacrificing accuracy. Earlier, Zhang demonstrated his versatility in robotics with **Capuchin** (2013, 27 citations), a quasi-autonomous four-limbed robot capable of free-climbing vertical terrain—a significant step beyond aid climbing that requires engineered features. This work integrated sensing, planning, and control for unstructured environments. He also pioneered evolutionary gait generation for bipedal robots (2003, 21 citations), using trajectory-based parameters to satisfy the ZMP criterion for stable walking on flat ground and stairs. Most recently, Zhang has ventured into flexible electronics, developing a laser-induced graphene sensor with macroscopic crack arrays for tunable gauge factor modulation (2024). His career reflects a rare breadth—from perception algorithms to mechanical design to novel materials—making him a researcher who bridges theory and hardware in intelligent systems.