REACT: Real-time Efficient Attribute Clustering and Transfer for Updatable 3D Scene Graph

2025-03-05

AI Summary

To address the challenge of real-time object node relocalization in dynamic environments for 3D scene graphs (3DSGs), this paper proposes the first attribute-driven real-time relocation framework. Methodologically, we design a triplet-loss-based embedding model to enable robust cross-temporal object instance matching; develop an attribute-aware clustering and node migration mechanism; and integrate a lightweight incremental update strategy. Our key contribution lies in breaking the static assumption inherent in conventional 3DSGs, enabling millisecond-level (<10 ms) online updates during long-term operation. Experimental results demonstrate significant improvements in object relocalization accuracy under dynamic conditions—achieving an average 23.6% gain over baseline methods. This work establishes a foundation for sustainable, high-level semantic mapping essential for autonomous robotic systems operating in evolving environments.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Modern-day autonomous robots need high-level map representations to perform sophisticated tasks. Recently, 3D scene graphs (3DSGs) have emerged as a promising alternative to traditional grid maps, blending efficient memory use and rich feature representation. However, most efforts to apply them have been limited to static worlds. This work introduces REACT, a framework that efficiently performs real-time attribute clustering and transfer to relocalize object nodes in a 3DSG. REACT employs a novel method for comparing object instances using an embedding model trained on triplet loss, facilitating instance clustering and matching. Experimental results demonstrate that REACT is able to relocalize objects while maintaining computational efficiency. The REACT framework's source code will be available as an open-source project, promoting further advancements in reusable and updatable 3DSGs.

Problem

• Real-time attribute clustering for 3D scene graphs
• Efficient object relocalization in dynamic environments
• Open-source framework for updatable 3D scene graphs

Innovation

• Real-time attribute clustering and transfer
• Embedding model with triplet loss
• Open-source updatable 3D scene graphs