Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866918442382131200 |
|---|---|
| author | Xiong, Xinmiao Liu, Bangya Wang, Hao Li, Dayou Chen, Nuo Feng, Andrew Ding, Mingyu Banerjee, Suman Zhou, Yang Fan, Zhiwen |
| author_facet | Xiong, Xinmiao Liu, Bangya Wang, Hao Li, Dayou Chen, Nuo Feng, Andrew Ding, Mingyu Banerjee, Suman Zhou, Yang Fan, Zhiwen |
| contents | Geometric Foundation Models (GFMs) have recently advanced monocular SLAM by providing robust, calibration-free 3D priors. However, deploying these models on dense video streams introduces significant computational redundancy. Current GFM-based SLAM systems typically rely on post hoc keyframe selection. Because of this, they must perform expensive dense geometric decoding simply to determine whether a frame contains novel geometry, resulting in late rejection and wasted computation. To mitigate this inefficiency, we propose LeanGate, a lightweight feed-forward frame-gating network. LeanGate predicts a geometric utility score to assess a frame's mapping value prior to the heavy GFM feature extraction and matching stages. As a predictive plug-and-play module, our approach bypasses over 90% of redundant frames. Evaluations on standard SLAM benchmarks demonstrate that LeanGate reduces tracking FLOPs by more than 85% and achieves a 5x end-to-end throughput speedup. Furthermore, it maintains the tracking and mapping accuracy of dense baselines. Project page: https://lean-gate.github.io/ |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_08718 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring Xiong, Xinmiao Liu, Bangya Wang, Hao Li, Dayou Chen, Nuo Feng, Andrew Ding, Mingyu Banerjee, Suman Zhou, Yang Fan, Zhiwen Computer Vision and Pattern Recognition Artificial Intelligence Robotics Geometric Foundation Models (GFMs) have recently advanced monocular SLAM by providing robust, calibration-free 3D priors. However, deploying these models on dense video streams introduces significant computational redundancy. Current GFM-based SLAM systems typically rely on post hoc keyframe selection. Because of this, they must perform expensive dense geometric decoding simply to determine whether a frame contains novel geometry, resulting in late rejection and wasted computation. To mitigate this inefficiency, we propose LeanGate, a lightweight feed-forward frame-gating network. LeanGate predicts a geometric utility score to assess a frame's mapping value prior to the heavy GFM feature extraction and matching stages. As a predictive plug-and-play module, our approach bypasses over 90% of redundant frames. Evaluations on standard SLAM benchmarks demonstrate that LeanGate reduces tracking FLOPs by more than 85% and achieves a 5x end-to-end throughput speedup. Furthermore, it maintains the tracking and mapping accuracy of dense baselines. Project page: https://lean-gate.github.io/ |
| title | Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Robotics |
| url | https://arxiv.org/abs/2604.08718 |