Kehui Ma

Papers

Total Citations

H-Index

About

Kehui Ma is a leading researcher in autonomous off-road navigation, specializing in learning-based perception and path planning for unstructured environments. Their seminal work, "Learning-Based Traversability Costmap for Autonomous Off-Road Navigation" (2025), has already garnered 4 citations, establishing a foundation for robust terrain assessment in robotics. Ma’s primary contributions lie in developing deep learning models that generate real-time traversability costmaps, enabling vehicles to safely navigate challenging terrains like forests, rocky slopes, and muddy trails. By fusing data from LiDAR, cameras, and inertial sensors, their approach significantly improves the accuracy of obstacle detection and path selection compared to traditional geometric methods. This work has direct applications in agricultural robotics, search-and-rescue missions, and planetary exploration. Ma’s research is notable for bridging the gap between simulation-trained models and real-world deployment, addressing critical issues of domain shift and computational efficiency. Their achievements include pioneering costmap learning frameworks that adapt to dynamic environmental conditions, a key step toward fully autonomous off-road systems. With growing recognition in the robotics community, Kehui Ma continues to push boundaries in safe, intelligent navigation for complex outdoor settings.