Advances in SLAM and 3D Reconstruction

The field of Simultaneous Localization and Mapping (SLAM) and 3D reconstruction is rapidly advancing with the introduction of new methodologies and techniques. Recent developments have focused on improving the accuracy and robustness of SLAM systems, particularly in dynamic and densely populated environments. The integration of semantic information with 3D Gaussian representations has shown promising results in enhancing both localization and mapping accuracy. Additionally, the use of hybrid representation models and structural supervision has improved scene representation capacity and global consistency. Noteworthy papers in this area include 4D Gaussian Splatting SLAM, which proposes an efficient architecture for incrementally tracking camera poses and constructing Gaussian radiance fields in dynamic scenes, and STAMICS, which integrates semantic information with 3D Gaussian representations for dense RGB-D SLAM. Other notable papers include GI-SLAM, which introduces an IMU-enhanced camera tracking module and a realistic 3D Gaussian-based scene representation, and HS-SLAM, which proposes a hybrid encoding network and structural supervision for improved dense SLAM.

Sources

4D Gaussian Splatting SLAM

Targetless 6DoF Calibration of LiDAR and 2D Scanning Radar Based on Cylindrical Occupancy

Autonomous Exploration-Based Precise Mapping for Mobile Robots through Stepwise and Consistent Motions

ColabSfM: Collaborative Structure-from-Motion by Point Cloud Registration

Splat-LOAM: Gaussian Splatting LiDAR Odometry and Mapping

GI-SLAM: Gaussian-Inertial SLAM

Ground Penetrating Radar-Assisted Multimodal Robot Odometry Using Subsurface Feature Matrix

From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splatting

MM-LINS: a Multi-Map LiDAR-Inertial System for Over-Degenerate Environments

Lidar-only Odometry based on Multiple Scan-to-Scan Alignments over a Moving Window

STAMICS: Splat, Track And Map with Integrated Consistency and Semantics for Dense RGB-D SLAM

InteractionMap: Improving Online Vectorized HDMap Construction with Interaction

HS-SLAM: Hybrid Representation with Structural Supervision for Improved Dense SLAM

Robust simultaneous UWB-anchor calibration and robot localization for emergency situations

Built with on top of