Beyond Controlled Environments: 3D Camera Re-Localization in Changing Indoor Scenes

TL;DR

This paper introduces RIO10, a new benchmark for long-term indoor camera re-localization that accounts for appearance changes, highlighting the challenges and performance gaps of current methods in dynamic indoor environments.

Contribution

The paper adapts the existing 3RScan dataset to create RIO10, proposes new evaluation metrics, and analyzes how scene changes impact re-localization performance, revealing it remains an unsolved problem.

Findings

State-of-the-art methods struggle with long-term indoor re-localization.

Scene changes significantly degrade re-localization accuracy.

The benchmark and tools are publicly available for further research.

Abstract

Long-term camera re-localization is an important task with numerous computer vision and robotics applications. Whilst various outdoor benchmarks exist that target lighting, weather and seasonal changes, far less attention has been paid to appearance changes that occur indoors. This has led to a mismatch between popular indoor benchmarks, which focus on static scenes, and indoor environments that are of interest for many real-world applications. In this paper, we adapt 3RScan - a recently introduced indoor RGB-D dataset designed for object instance re-localization - to create RIO10, a new long-term camera re-localization benchmark focused on indoor scenes. We propose new metrics for evaluating camera re-localization and explore how state-of-the-art camera re-localizers perform according to these metrics. We also examine in detail how different types of scene change affect the performance of different methods, based on novel ways of detecting such changes in a given RGB-D frame. Our results clearly show that long-term indoor re-localization is an unsolved problem. Our benchmark and tools are publicly available at waldjohannau.github.io/RIO10