Fenglu Ge

Papers

Total Citations

H-Index

About

Fenglu Ge is a researcher whose work lies at the intersection of robotics, machine learning, and industrial automation, with a particular focus on the challenging domain of mining tunnel inspection. Ge’s primary contribution is in the field of Learning from Demonstration (LfD), where robots are taught complex tasks through human-provided examples rather than explicit programming. In their most cited work, "Robot learning by a mining tunnel inspection robot" (2012), Ge pioneered the application of Discrete Hidden Markov Models (DHMM) to train a robot for autonomous inspection tasks, achieving 6 citations. This was complemented by a comparative study (2015, 5 citations) that systematically evaluated Gaussian Mixture Models (GMM), Continuous Hidden Markov Models (CHMM), and DHMM, providing a valuable roadmap for practitioners. Ge also introduced an innovative "Information Extraction" method for training dataset selection (2011, 3 citations), addressing a critical bottleneck in LfD by ensuring more relevant and efficient learning. With a total of 14 citations across these three foundational papers, Ge’s work has laid essential groundwork for deploying intelligent, adaptable robots in hazardous underground environments, demonstrating how LfD can bridge the gap between human expertise and autonomous operation.