Robust Perception and Mobile Robotics Lab (RPM Robotics Lab) focuses on perceptual simultaneous localization and mapping (SLAM). Main research interest and detailed robotics topics include perception-based environment mapping, intelligent sensor fusion, decision making and control of the robotic agents, robotic operation, and navigation in GPS-denied environments (e.g., underwater, urban, and indoor environments).
Datasets (click to expand):
- ๐ Complex Urban Dataset : Complex Urban Dataset with Multi-level Sensors from Highly Diverse Urban Environments;
- ๐ MulRan : Multimodal Range Dataset for Urban Place Recognition;
- ๐ STheReO Dataset : Stereo Thermal Dataset for Research in Odometry and Mapping;
- ๐ TRansPose : Large-Scale Multispectral Dataset for Transparent Object;
- ๐ HeLiPR Dataset : Heterogeneous LiDAR Dataset for inter-LiDAR Place Recognition;
LiDAR SLAM for Enhanced Resilience (click to expand):
- ๐ Scan Context (IROS'18) : Egocentric Spatial Descriptor for Place Recognition within 3D Point Cloud Map;
- ๐ Scan Context++ (T-RO'22) : Structural Place Recognition Robust to Rotation and Lateral Variations in Urban Environments;
- ๐ MA-LIO (RA-L'23) : Asynchronous Multiple LiDAR-Inertial Odometry using Point-wise Inter-LiDAR Uncertainty Propagation;
Radar in Robot Navigation (click to expand):
- ๐ PhaRaO (ICRA'20) : Direct Radar Odometry using Phase Correlation;
- ๐ RaPlace (IROS'23) : Place Recognition for Imaging Radar using Radon Transform and Mutable Threshold;
Thermal Camera for SLAM beyond the Visible Spectrum (click to expand):
- ๐ sRGB-TIR (ICRA'23) : Edge-guided Multi-domain RGB-to-TIR image Translation for Training Vision Tasks with Challenging Labels;
Vision-based Robot Manipulation (click to expand):
- ๐ PrimA6D (RA-L'20) : Rotational Primitive Reconstruction for Enhanced and Robust 6D Pose Estimation;
- ๐ PrimA6D++ (RA-L'22) : Ambiguity-Aware Multi-Object Pose Optimization for Visually-Assisted Robot Manipulation;