AMB3R：带后端的前馈式精确度量尺度三维重建
AMB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend
用稀疏紧凑的体积场景表示作为后端，把多视图前馈网络零优化地扩展到未标定的视觉里程计与任意数量图像的 SfM，可直接复用于需要度量尺度稠密重建与位姿估计的视觉管线。
原文依据 employ a sparse, yet compact, volumetric scene representation as our backend, enabling geometric reasoning with spatial compactness
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 AMB3R，一个面向度量尺度稠密三维重建的多视图前馈模型，可应对多种三维视觉任务。其核心思想是采用一种稀疏却紧凑的体积场景表示作为后端，从而在空间紧凑性下进行几何推理。我们进一步引入 AMB3R-VO 与 AMB3R-SfM 两条无需训练、与模型无关的流水线，将多视图网络扩展到未标定的视觉里程计（在线）或适用于任意数量图像的运动恢复结构，且无需测试时优化。与先前的三维基础模型相比，AMB3R 在相机位姿、深度、度量尺度估计与三维重建上均达到最优性能，而 AMB3R-VO 与 AMB3R-SfM 在常见基准上甚至超越了带有稠密重建先验的基于优化的 SLAM 与 SfM 系统。
英文原文摘要 818 字符
We present AMB3R, a multi-view feed-forward model for metric-scale dense 3D reconstruction that addresses diverse 3D vision tasks. The key idea is to employ a sparse, yet compact, volumetric scene representation as our backend, enabling geometric reasoning with spatial compactness. We further introduce AMB3R-VO and AMB3R-SfM, two training-free, model-agnostic pipelines that extend multi-view networks to uncalibrated visual odometry (online) or structure from motion for any number of images without test-time optimization. Compared to prior 3D foundation models, AMB3R achieves state-of-the-art performance in camera pose, depth, metric-scale estimation, and 3D reconstruction, while AMB3R-VO and AMB3R-SfM even surpass optimization-based SLAM and SfM systems with dense reconstruction priors on common benchmarks.