arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.08718cs.CVcs.AIcs.RO
收藏

通过几何效用评分加速基于Transformer的单目SLAM

Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring

UW–Madison(威斯康星大学麦迪逊分校);Texas A&M(德州农工大学);USC(南加州大学);UNC Chapel Hill(北卡罗来纳大学教堂山分校)

Xinmiao Xiong, Bangya Liu, Hao Wang, Dayou Li, Nuo Chen, Andrew Feng, Mingyu Ding, Suman Banerjee, Yang Zhou, Zhiwen Fan

AI总结:

本文提出LeanGate,通过预测几何效用评分在重特征提取前筛选帧,减少冗余计算,提升SLAM效率与准确性。

摘要:

几何基础模型(GFMs)通过提供稳健且无需校准的3D先验,最近在单目SLAM中取得了进展。然而,在密集视频流中部署这些模型会引入显著的计算冗余。当前基于GFMs的SLAM系统通常依赖事后关键帧选择。由于此原因,它们必须执行昂贵的密集几何解码仅为了确定一帧是否包含新几何,导致后期拒绝和计算浪费。为缓解这种低效,我们提出了LeanGate,一个轻量级的前馈帧门控网络。LeanGate预测一个几何效用评分以在重特征提取和匹配阶段之前评估帧的映射价值。作为预测的即插即用模块,我们的方法绕过了超过90%的冗余帧。在标准SLAM基准测试中,评估显示LeanGate将跟踪FLOPs减少超过85%,并实现5倍的端到端吞吐量提升。此外,它保持了密集基线的跟踪和映射准确性。项目页面:https://lean-gate.github.io/

英文摘要:

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/

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑