ROS 2 underwater localization for real AUV/ROV data, validated on public ocean datasets.
Kalman filter state estimation for robotics from a DSP angle — tracking + position/IMU fusion with sensor-outage robustness
2D SLAM back-end: least-squares pose-graph optimisation on the SE(2) manifold.
Research prototype for semantic visual SLAM with front-view observations, BEV geometry, and pose-graph optimization.
2D LiDAR SLAM from scratch in NumPy/SciPy for ROS 2 Jazzy + Gazebo: ICP, point-to-line, robust losses, correlative matching, pose graph, with 42 measured experiments
A 3D testbed that verifies SLAM: a ground-truth generator + geometric oracle that grades pose-graph / VO / full-SLAM algorithms on a hidden trajectory. Built on the Touchstone verifier.
Stereo visual odometry and SLAM on KITTI sequence 00 — feature tracking, PnP-RANSAC odometry, GTSAM bundle adjustment, pose-graph optimization, and loop closure, plus a deep-learning front-end study that halves open-loop drift.
To associate your repository with the pose-graph topic, visit your repo's landing page and select "manage topics."