Shuangfeng Wei

Ministry of Natural Resources, Beijing University of Civil Engineering and Architecture

Papers

Total Citations

H-Index

About

Shuangfeng Wei is advancing the frontier of robotic perception by tackling one of visual SLAM’s most stubborn challenges: dynamic, semantically ambiguous environments. His research centers on integrating semantic information with simultaneous localization and mapping (SLAM) to enable mobile robots to operate reliably in the messy, unpredictable indoor spaces where traditional systems fail. Wei’s most influential work, “A Semantic Information-Based Optimized vSLAM in Indoor Dynamic Environments” (2023, 10 citations), directly addresses the static-scene assumption that cripples conventional vSLAM, offering a framework that allows robots to build accurate maps and localize themselves even when objects move around them. Building on this, his 2025 paper “DeepLabV3+-Based Semantic Annotation Refinement for SLAM in Indoor Environments” tackles the critical bottleneck of poor semantic data quality, using advanced deep learning to refine annotations and dramatically improve 3D reconstruction from monocular cameras. Though early in his career, Wei’s focused contributions are already gaining traction among researchers working on robust, real-world robotic navigation. His work represents a vital step toward autonomous systems that can truly understand and adapt to the dynamic, semantically rich environments we inhabit.