Chamelion: Reliable Change Detection for Long-Term LiDAR Mapping in Transient Environments

2026-02-09

AI Summary

This work addresses the challenge of robustly detecting structural changes and continuously updating long-term LiDAR maps in dynamic environments—such as construction sites or reconfigurable indoor spaces—where existing methods struggle due to occlusions, spatiotemporal variations, and scarce labeled data. The authors propose a dual-head deep network architecture that enables online change detection alongside persistent map maintenance. To overcome the reliance on extensive manual annotations, they introduce an innovative cross-scenario structural element synthesis strategy for effective training. Experiments in real-world construction and office settings demonstrate that the method efficiently and accurately updates maps while exhibiting strong generalization across diverse scenes.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Life-Long and Continual Learning
• Search and Optimization: Learning to Search

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
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Online change detection is crucial for mobile robots to efficiently navigate through dynamic environments. Detecting changes in transient settings, such as active construction sites or frequently reconfigured indoor spaces, is particularly challenging due to frequent occlusions and spatiotemporal variations. Existing approaches often struggle to detect changes and fail to update the map across different observations. To address these limitations, we propose a dual-head network designed for online change detection and long-term map maintenance. A key difficulty in this task is the collection and alignment of real-world data, as manually registering structural differences over time is both labor-intensive and often impractical. To overcome this, we develop a data augmentation strategy that synthesizes structural changes by importing elements from different scenes, enabling effective model training without the need for extensive ground-truth annotations. Experiments conducted at real-world construction sites and in indoor office environments demonstrate that our approach generalizes well across diverse scenarios, achieving efficient and accurate map updates.\resubmit{Our source code and additional material are available at: https://chamelion-pages.github.io/.

Problem

• change detection
• LiDAR mapping
• transient environments
• map maintenance
• dynamic environments

Innovation

• online change detection
• LiDAR mapping
• data augmentation
• dual-head network
• transient environments