Search papers, labs, and topics across Lattice.
VGGT-SLAM++ is a visual SLAM system that integrates a Visual Geometry Grounded Transformer (VGGT) front-end with a novel spatially corrective back-end to improve accuracy and efficiency. The back-end constructs a dense Digital Elevation Map (DEM) from VGGT submaps, partitions it into patches, and uses DINOv2 embeddings for covisibility graph construction and visual place recognition. By restoring high-cadence local bundle adjustment, VGGT-SLAM++ achieves state-of-the-art accuracy on standard SLAM benchmarks, reducing short-term drift and accelerating graph convergence.
Transformer-based SLAM can achieve state-of-the-art accuracy and efficiency by integrating a spatially corrective back-end that restores high-cadence local bundle adjustment.
We introduce VGGT-SLAM++, a complete visual SLAM system that leverages the geometry-rich outputs of the Visual Geometry Grounded Transformer (VGGT). The system comprises a visual odometry (front-end) fusing the VGGT feed-forward transformer and a Sim(3) solution, a Digital Elevation Map (DEM)-based graph construction module, and a back-end that jointly enable accurate large-scale mapping with bounded memory. While prior transformer-based SLAM pipelines such as VGGT-SLAM rely primarily on sparse loop closures or global Sim(3) manifold constraints - allowing short-horizon pose drift - VGGT-SLAM++ restores high-cadence local bundle adjustment (LBA) through a spatially corrective back-end. For each VGGT submap, we construct a dense planar-canonical DEM, partition it into patches, and compute their DINOv2 embeddings to integrate the submap into a covisibility graph. Spatial neighbors are retrieved using a Visual Place Recognition (VPR) module within the covisibility window, triggering frequent local optimization that stabilizes trajectories. Across standard SLAM benchmarks, VGGT-SLAM++ achieves state-of-the-art accuracy, substantially reducing short-term drift, accelerating graph convergence, and maintaining global consistency with compact DEM tiles and sublinear retrieval.
