Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
GeoGS-SLAM integrates 3D Gaussian Splatting with learned geometric priors to enhance online monocular dense reconstruction without relying on external depth sensors. By utilizing a feed-forward visual geometry model to predict camera and scene priors, the system effectively samples Gaussian primitives from RGB inputs and geometric data, optimizing both camera poses and the scene map through a coarse-to-fine strategy. Experimental results show that GeoGS-SLAM outperforms state-of-the-art methods in rendering quality and tracking accuracy while maintaining real-time performance across various benchmarks.
Superior rendering quality and tracking accuracy achieved in real-time monocular reconstruction without external depth sensors.
SLAM methods based on 3D Gaussian Splatting (3DGS) have demonstrated impressive tracking and mapping performance, but typically require additional geometric information from external depth sensors. Meanwhile, recent SLAM systems that leverage geometric priors from pre-trained feed-forward models enable real-time dense reconstruction, yet often discard original RGB information during optimization, thus degrading overall reconstruction quality. We present GeoGS-SLAM, an online monocular dense reconstruction system that combines the 3DGS-based map representation with learned geometric priors. Given uncalibrated RGB input, we first employ a feed-forward visual geometry model to predict camera and scene priors. The Gaussian scene map is then expanded by directly sampling Gaussian primitives from both RGB input and geometric priors. Camera poses and the scene map are jointly optimized through a coarse-to-fine strategy that minimizes both photometric and geometric losses. To ensure global consistency, we further incorporate online loop closure detection and pose graph optimization. Extensive experiments across indoor and outdoor benchmarks demonstrate that GeoGS-SLAM achieves superior rendering quality and tracking accuracy compared to state-of-the-art methods while maintaining online real-time performance. Project page: https://rlgao.github.io/geogs_slam.