Argos: Adapt Rich Geometric Priors for Generalizable Online Scene-Change-Detection

2026-10-07

AI Summary

This study addresses the limitations of robot scene change detection in dynamic environments, where performance is constrained by viewpoint variations, occlusions, and cross-domain generalization. To overcome these challenges, this work proposes Argos, a framework that leverages implicit 3D priors from geometric foundation models (GFMs) to jointly optimize scene change detection and 3D reconstruction. Furthermore, it constructs a large-scale benchmark comprising mixed synthetic and real-world data to enhance generalization capabilities, and develops a real-time Argos-SLAM system for online change perception and 4D mapping. Experimental results demonstrate that the proposed method significantly outperforms existing baselines across multiple benchmarks, improving change IoU and F1 scores by 42.01% and 27.91%, respectively, thereby exhibiting strong potential for scalable deployment.

Abstract

Robots operating in dynamic environments require reliable detection of how their surroundings change over time. Existing learning-based methods largely rely on pairwise 2D image features, which struggle under large viewpoint changes and occlusions, are sensitive to noise, and show limited generalization across domains, while explicit 3D approaches typically require costly offline optimization. We show that the implicit 3D knowledge of Geometric Foundation Models (GFMs) provides a strong basis for addressing these limitations. We introduce Argos, which adapts GFM features for joint scene change detection and 3D reconstruction. To address data scarcity and take a step toward a foundation model for scene change detection, we introduce a large-scale benchmark comprising two synthetic datasets and one real-world dataset, and train jointly across diverse datasets to improve cross-domain generalization. We further introduce Argos-SLAM, a real-time system designed for robotics, which performs online change detection and change-aware 4D mapping. Across benchmarks, our framework substantially outperforms existing baselines, with gains of up to 42.01% in change IoU and 27.91% in F1, while supporting scalable deployment in changing real-world environments.

Problem

• Scene Change Detection
• Dynamic Environments
• Cross-domain Generalization
• 3D Reconstruction
• Data Scarcity

Innovation

• Geometric Foundation Models
• Scene Change Detection
• 3D Reconstruction
• SLAM
• Cross-domain Generalization