Suyi Liu

Papers

Total Citations

H-Index

About

Suyi Liu is a pioneering researcher in intelligent robotic perception and 3D vision, with foundational contributions to automated welding systems and advanced point cloud analysis. His early work on robot welding seam tracking using structured light vision (2010, 24 citations) established a robust framework for real-time, vision-guided industrial automation. Liu further advanced this field by developing subpixel accuracy methods for V-groove center extraction (2006), employing Laplacian of Gaussian edge detection and corner detection via extremum curvature—a technique that significantly enhanced the precision of laser-based seam tracking in robotic welding. More recently, Liu has pushed the boundaries of 3D scene understanding with his work on recurrent slice networks (2023), which addresses critical challenges in simultaneous semantic and instance segmentation of point clouds. By mapping unordered point clouds onto ordered sequences and leveraging bidirectional LSTM networks, his method achieves superior feature extraction and multi-scale fusion, effectively mitigating issues like semantic misclassification and edge blurring in dense or small-object scenarios. With over 28 total citations spanning two decades of innovation, Liu’s research bridges classical industrial robotics with cutting-edge deep learning, demonstrating lasting impact on both manufacturing precision and autonomous 3D perception.