Introduction
Room reidentification (ReID) is a challenging yet essential task with numerous applications in fields such as augmented reality (AR) and homecare robotics. Existing visual place recognition (VPR) methods, which typically rely on global descriptors or aggregate local features, often struggle in cluttered indoor environments densely populated with man-made objects. These methods tend to overlook the crucial role of object-oriented information. To address this, we propose AirRoom, an object-aware pipeline that integrates multi-level object-oriented information—from global context to object patches, object segmentation, and keypoints—utilizing a coarse-to-fine retrieval approach. Extensive experiments on four newly constructed datasets—MPReID, HMReID, GibsonReID, and ReplicaReID—demonstrate that AirRoom outperforms state-of-the-art (SOTA) models across nearly all evaluation metrics, with improvements ranging from 6% to 80%. Moreover, AirRoom exhibits significant flexibility, allowing various modules within the pipeline to be substituted with different alternatives without compromising overall performance. It also shows robust and consistent performance under diverse viewpoint variations.
We introduce AirRoom, an object-aware room ReID pipeline with two novel modules: the Receptive Field Expander and Object-Aware Scoring, effectively leveraging multi-level object-oriented information to overcome the limitations observed in previous methods. And we have curated four comprehensive room reidentification datasets—MPReID, HMReID, GibsonReID, and ReplicaReID—providing diverse benchmarks for evaluating room reidentification methods.
Method Overview
The pipeline begins with the Global Feature Extractor, which captures global context features to retrieve the top-5 reference images. Instance segmentation then generates object masks, followed by the Receptive Field Expander, which extracts object patches. The Object Feature Extractor processes both object and patch features. The Object-Aware Scoring module narrows the selection to the top-2 candidates, and Fine-Grained Retrieval identifies the most suitable reference image.
Room ReID Datasets
Several high-quality indoor 3D datasets—such as Matterport3D, Habitat-Matterport3D, the Gibson Database of 3D Spaces, and Replica—offer real-world indoor scenes. Building on these resources and utilizing the interactive Habitat Simulator, we created four new datasets: MPReID, HMReID, GibsonReID, and ReplicaReID.
Publication
-
AirRoom: Objects Matter in Room Reidentification.IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025.
