Towards Global Localization using Multi-Modal Object-Instance Re-Identification

TL;DR

This paper introduces a novel multimodal transformer architecture for robust object-instance re-identification using RGB and depth data, significantly improving localization and perception in cluttered and varying illumination scenes.

Contribution

It proposes a dual-path transformer model for multimodal ReID and a ReID-based localization framework, validated on custom and public RGB-D datasets, advancing robotic perception capabilities.

Findings

ReID accuracy achieved mAP of 75.18

Localization success rate of 83% on TUM-RGBD

Depth data enhances ReID robustness in challenging scenes

Abstract

Re-identification (ReID) is a critical challenge in computer vision, predominantly studied in the context of pedestrians and vehicles. However, robust object-instance ReID, which has significant implications for tasks such as autonomous exploration, long-term perception, and scene understanding, remains underexplored. In this work, we address this gap by proposing a novel dual-path object-instance re-identification transformer architecture that integrates multimodal RGB and depth information. By leveraging depth data, we demonstrate improvements in ReID across scenes that are cluttered or have varying illumination conditions. Additionally, we develop a ReID-based localization framework that enables accurate camera localization and pose identification across different viewpoints. We validate our methods using two custom-built RGB-D datasets, as well as multiple sequences from the open-source TUM RGB-D datasets. Our approach demonstrates significant improvements in both object instance ReID (mAP of 75.18) and localization accuracy (success rate of 83% on TUM-RGBD), highlighting the essential role of object ReID in advancing robotic perception. Our models, frameworks, and datasets have been made publicly available.