Vardeep S. Sandhu
Papers
1
Total Citations
16
H-Index
1
About
Vardeep S. Sandhu is a leading researcher in autonomous systems perception, specializing in radar-based scene understanding for robotics and self-driving vehicles. His work addresses the critical challenge of reliable moving object detection in dynamic environments where traditional sensors like LiDAR and cameras falter—particularly under adverse weather conditions. Sandhu’s most cited paper, "Radar Instance Transformer: Reliable Moving Instance Segmentation in Sparse Radar Point Clouds" (2023, 16 citations), introduces a novel transformer architecture that directly leverages radar’s unique motion-sensing capabilities to segment moving instances from sparse point clouds. This contribution is pivotal for collision avoidance in autonomous navigation, offering a robust alternative to optical sensors that lack direct motion information. By demonstrating how radar can overcome the limitations of LiDAR and cameras in rain, fog, or snow, Sandhu’s work has significant implications for safer, all-weather autonomous systems. His research bridges the gap between sparse radar data and high-level perception tasks, marking an important step toward more resilient robotic perception.
Research Focus
Key Achievements
Top Papers
- 1