Mingyong Zeng

PLA Army Engineering University, Tianjin University

Papers

Total Citations

H-Index

About

Mingyong Zeng is a researcher whose work bridges computer vision and robotics, with a particular focus on person re-identification and multi-robot odor source localization. In his most cited work, "Efficient person re-identification by hybrid spatiogram and covariance descriptor" (2015, 38 citations), Zeng challenges the prevailing trend favoring metric learning by demonstrating that carefully designed features can still achieve competitive performance. He proposes a novel hybrid descriptor that combines spatiogram and covariance features, offering an efficient and effective alternative for person re-identification systems—a critical task in surveillance and security. Earlier, Zeng contributed to robotics with "Single Odor Source Declaration by Using Multiple Robots" (2009, 15 citations), where he introduces a three-step method for robots to collaboratively locate and declare odor sources in indoor environments. This work involves robot convergence, odor persistence judgment, and mass throughput calculation, showcasing his ability to tackle real-world environmental sensing challenges. With a career spanning feature engineering in computer vision to multi-robot coordination, Zeng’s research demonstrates a practical, systems-oriented approach that has influenced both fields.