用八百万对图像训练，实现快速准确的跨场景相机位姿定位
Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localization
- 直接回归相对位姿，搭配极简运动平均求绝对位姿
- 在六个公开数据集上实现实时高精度定位，跨场景泛化能力强
- 适合需要快速部署、跨场景适应的视觉定位应用
视觉定位旨在确定查询图像相对于一组带位姿图像数据库的相机位姿。近年来，直接回归相机位姿的深度神经网络因其快速推理能力而受到青睐。然而，现有方法在新场景下的泛化能力或位姿估计精度方面表现不佳。为此，我们提出 Reloc3r，一种简洁有效的视觉定位框架。它由一个精心设计的相对位姿回归网络和一个极简的运动平均模块组成，用于绝对位姿估计。在约八百万对带位姿图像上进行训练，Reloc3r 实现了出人意料的良好性能与泛化能力。我们在六个公开数据集上进行了广泛实验，持续验证了该方法的有效性与高效性。其可在实时条件下提供高质量的相机位姿估计，并能推广至新场景。代码已开源。
原文摘要 · Abstract (English)
Visual localization aims to determine the camera pose of a query image relative to a database of posed images. In recent years, deep neural networks that directly regress camera poses have gained popularity due to their fast inference capabilities. However, existing methods struggle to either generalize well to new scenes or provide accurate camera pose estimates. To address these issues, we present Reloc3r, a simple yet effective visual localization framework. It consists of an elegantly designed relative pose regression network, and a minimalist motion averaging module for absolute pose estimation. Trained on approximately eight million posed image pairs, Reloc3r achieves surprisingly good performance and generalization ability. We conduct extensive experiments on six public datasets, consistently demonstrating the effectiveness and efficiency of the proposed method. It provides high-quality camera pose estimates in real time and generalizes to novel scenes. Code: https://github.com/ffrivera0/reloc3r.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。