SARA: Scene-Aware Reconstruction Accelerator

TL;DR

SARA introduces a geometry-first pair selection method for SfM that significantly reduces matching complexity and improves accuracy by scoring reconstruction informativeness before matching.

Contribution

The paper proposes a novel geometry-driven pair selection module that outperforms traditional visual similarity-based methods in efficiency and accuracy.

Findings

Reduces rotation errors by 46.5%

Achieves up to 50x speedup in pair matching

Maintains reconstruction quality within 3% of baseline

Abstract

We present SARA (Scene-Aware Reconstruction Accelerator), a geometry-driven pair selection module for Structure-from-Motion (SfM). Unlike conventional pipelines that select pairs based on visual similarity alone, SARA introduces geometry-first pair selection by scoring reconstruction informativeness - the product of overlap and parallax - before expensive matching. A lightweight pre-matching stage uses mutual nearest neighbors and RANSAC to estimate these cues, then constructs an Information-Weighted Spanning Tree (IWST) augmented with targeted edges for loop closure, long-baseline anchors, and weak-view reinforcement. Compared to exhaustive matching, SARA reduces rotation errors by 46.5+-5.5% and translation errors by 12.5+-6.5% across modern learned detectors, while achieving at most 50x speedup through 98% pair reduction (from 30,848 to 580 pairs). This reduces matching complexity from quadratic to quasi-linear, maintaining within +-3% of baseline reconstruction metrics for 3D Gaussian Splatting and SVRaster.