Kechen Song
Papers
15
Total Citations
411
H-Index
11
About
Kechen Song is a prominent researcher specializing in robotic visual perception, salient object detection, and intelligent grasping systems. His work sits at the intersection of computer vision and robotics, with a particular focus on enabling robots to reliably detect and grasp objects in complex, real-world environments. Song's most influential contribution is his development of a novel visible-depth-thermal image dataset for salient object detection in robotic grasping applications (2022, 100 citations), which has become a foundational resource for the research community. Complementing this, his comprehensive review of data-driven robotic visual grasping detection (59 citations) and his lightweight RGB-D generative grasping framework (46 citations) have significantly advanced the field's understanding of efficient, intelligent manipulation strategies. His research portfolio also demonstrates impressive breadth: from asymptotically optimal motion planning algorithms to semantic segmentation-guided grasping for weakly textured objects, and industrial defect detection in seamless steel pipes. His work addressing transparent and reflective objects — notoriously difficult for standard vision systems — reflects a commitment to solving practically challenging problems. With over 370 cumulative citations across his most notable papers, Song has established himself as a highly impactful voice in intelligent robotic perception and industrial automation research.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10