Papers
12
Total Citations
398
H-Index
8
About
Jingwei Song is a robotics researcher whose work spans simultaneous localization and mapping (SLAM), surgical robotics, and robot perception, with a particular focus on enabling autonomous systems to operate reliably in complex, real-world environments. His most influential contribution, the OpenLORIS-Scene dataset (2020, 163 citations), addressed a critical gap in lifelong SLAM research by providing benchmark data capturing the dynamic, ever-changing conditions that service robots genuinely encounter — a resource that has become widely adopted by the robotics community. Equally notable is his pioneering MIS-SLAM system (2018, 127 citations), which achieved real-time, large-scale dense deformable mapping within minimally invasive surgical environments using heterogeneous computing, advancing the frontier of surgical augmented reality and robotic-assisted procedures. Song's more recent work pushes further into medical robotics, including vascular respiratory motion compensation, 3D-2D image registration for navigation, and lightweight CPU-based surgical SLAM systems. He has also contributed to geometric and learning-based approaches for robot perception and localization, fusing CNNs with geometric constraints for robust indoor positioning. Across his career, Song's research reflects a consistent drive to bridge theoretical advances in robot autonomy with demanding, safety-critical real-world applications.
Research Focus
Key Achievements
Top Papers
- 1Are We Ready for Service Robots? The OpenLORIS-Scene Datasets for Lifelong SLAM163 citations · 2020
- 2
- 3
- 4
- 5
- 6
- 7BDIS-SLAM: a lightweight CPU-based dense stereo SLAM for surgery8 citations · 2024
- 8
- 9Optical Flow-Based Vascular Respiratory Motion Compensation8 citations · 2023
- 10