Map-Relative Pose Regression for Visual Re-Localization

TL;DR

The paper introduces marepo, a scene-agnostic pose regression method that uses scene-specific maps to predict camera pose relative to the map, enabling rapid adaptation and superior accuracy across diverse environments.

Contribution

It proposes a novel map-relative pose regression approach that generalizes across scenes, reducing data requirements and enabling quick fine-tuning for improved accuracy.

Findings

Outperforms previous pose regression methods on public datasets

Can be applied to new map representations immediately or after minimal fine-tuning

Works effectively for both indoor and outdoor environments

Abstract

Pose regression networks predict the camera pose of a query image relative to a known environment. Within this family of methods, absolute pose regression (APR) has recently shown promising accuracy in the range of a few centimeters in position error. APR networks encode the scene geometry implicitly in their weights. To achieve high accuracy, they require vast amounts of training data that, realistically, can only be created using novel view synthesis in a days-long process. This process has to be repeated for each new scene again and again. We present a new approach to pose regression, map-relative pose regression (marepo), that satisfies the data hunger of the pose regression network in a scene-agnostic fashion. We condition the pose regressor on a scene-specific map representation such that its pose predictions are relative to the scene map. This allows us to train the pose regressor across hundreds of scenes to learn the generic relation between a scene-specific map representation and the camera pose. Our map-relative pose regressor can be applied to new map representations immediately or after mere minutes of fine-tuning for the highest accuracy. Our approach outperforms previous pose regression methods by far on two public datasets, indoor and outdoor. Code is available: https://nianticlabs.github.io/marepo