VGGT-SLAM++

TL;DR

VGGT-SLAM++ is an advanced visual SLAM system that improves large-scale mapping accuracy and stability by integrating transformer-based visual geometry, dense DEMs, and frequent local optimizations.

Contribution

It introduces a novel SLAM pipeline combining VGGT outputs, dense DEMs, and a back-end for high-cadence local bundle adjustment, outperforming prior transformer-based SLAM methods.

Findings

Achieves state-of-the-art accuracy on standard SLAM benchmarks.

Reduces short-term pose drift significantly.

Maintains global consistency with efficient DEM tiles.

Abstract

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.