Feng Pan

Marquette University, Jiangnan University

Papers

Total Citations

H-Index

About

Feng Pan is a researcher whose work bridges foundational robotics theory with modern deep learning applications. His early, highly cited research on "Efficient contact state graph generation for assembly applications" (2004, 13 citations) made a significant contribution to automated assembly strategies by developing a method to automatically generate all geometrically feasible contact states—a critical step for designing robust robotic manipulation and assembly processes. This work remains a key reference in the field of assembly planning and contact modeling. More recently, Pan has advanced the state of the art in visual SLAM (Simultaneous Localization and Mapping) for mobile robotics. His 2021 paper on "Loop Closure Detection in RGB-D SLAM by Utilizing Siamese ConvNet Features" (5 citations) addresses a core challenge in autonomous navigation: accurately recognizing previously visited locations to correct mapping drift. By leveraging Siamese convolutional neural networks, he introduced a data-driven approach to improve the robustness and accuracy of loop closure detection in RGB-D systems. This work demonstrates his ability to integrate classical robotics problems with contemporary computer vision techniques. Pan’s research trajectory, from contact state graphs to deep learning for SLAM, highlights a sustained focus on enabling more intelligent and autonomous robotic systems.