M^3: Dense Matching Meets Multi-View Foundation Models for Monocular Gaussian Splatting SLAM
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912971226087424 |
|---|---|
| author | Ren, Kerui Li, Guanghao Jiang, Changjian Xu, Yingxiang Lu, Tao Xu, Linning Dong, Junting Pang, Jiangmiao Yu, Mulin Dai, Bo |
| author_facet | Ren, Kerui Li, Guanghao Jiang, Changjian Xu, Yingxiang Lu, Tao Xu, Linning Dong, Junting Pang, Jiangmiao Yu, Mulin Dai, Bo |
| contents | Streaming reconstruction from uncalibrated monocular video remains challenging, as it requires both high-precision pose estimation and computationally efficient online refinement in dynamic environments. While coupling 3D foundation models with SLAM frameworks is a promising paradigm, a critical bottleneck persists: most multi-view foundation models estimate poses in a feed-forward manner, yielding pixel-level correspondences that lack the requisite precision for rigorous geometric optimization. To address this, we present M^3, which augments the Multi-view foundation model with a dedicated Matching head to facilitate fine-grained dense correspondences and integrates it into a robust Monocular Gaussian Splatting SLAM. M^3 further enhances tracking stability by incorporating dynamic area suppression and cross-inference intrinsic alignment. Extensive experiments on diverse indoor and outdoor benchmarks demonstrate state-of-the-art accuracy in both pose estimation and scene reconstruction. Notably, M^3 reduces ATE RMSE by 64.3% compared to VGGT-SLAM 2.0 and outperforms ARTDECO by 2.11 dB in PSNR on the ScanNet++ dataset. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_16844 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | M^3: Dense Matching Meets Multi-View Foundation Models for Monocular Gaussian Splatting SLAM Ren, Kerui Li, Guanghao Jiang, Changjian Xu, Yingxiang Lu, Tao Xu, Linning Dong, Junting Pang, Jiangmiao Yu, Mulin Dai, Bo Computer Vision and Pattern Recognition Streaming reconstruction from uncalibrated monocular video remains challenging, as it requires both high-precision pose estimation and computationally efficient online refinement in dynamic environments. While coupling 3D foundation models with SLAM frameworks is a promising paradigm, a critical bottleneck persists: most multi-view foundation models estimate poses in a feed-forward manner, yielding pixel-level correspondences that lack the requisite precision for rigorous geometric optimization. To address this, we present M^3, which augments the Multi-view foundation model with a dedicated Matching head to facilitate fine-grained dense correspondences and integrates it into a robust Monocular Gaussian Splatting SLAM. M^3 further enhances tracking stability by incorporating dynamic area suppression and cross-inference intrinsic alignment. Extensive experiments on diverse indoor and outdoor benchmarks demonstrate state-of-the-art accuracy in both pose estimation and scene reconstruction. Notably, M^3 reduces ATE RMSE by 64.3% compared to VGGT-SLAM 2.0 and outperforms ARTDECO by 2.11 dB in PSNR on the ScanNet++ dataset. |
| title | M^3: Dense Matching Meets Multi-View Foundation Models for Monocular Gaussian Splatting SLAM |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.16844 |