• Self-supervised, Risk-aware 3D Scene learning
• Traversability estimation of urban outdoor environments
• Uncertainty-aware 3D perception in autonomous navigation
• Learning from navigation experiences of mobile robots
• Uncertainty-aware localization in 3D point cloud maps
• Multi-sensor fusion for robust pose estimation
• Localization failure detection for enhanced reliability
• Global relocalization for pose recovery and initialization
• Finding collision-free paths for multiple agents.
• MAPF algorithms aim to optimize performance metrics.
• Executing the algorithm in the real world.