The City that Never Settles: Simulation-based LiDAR Dataset for Long-Term Place Recognition Under Extreme Structural Changes

2025-05-08

AI Summary

Long-term place recognition (PR) suffers severe performance degradation under extreme urban structural changes caused by large-scale construction and demolition. To address this, we introduce CNS, the first synthetic LiDAR dataset specifically designed for outdoor environments undergoing drastic spatiotemporal evolution. Generated using CARLA, CNS comprises multi-temporal, multi-map, and multi-stage point cloud sequences. We propose a novel symmetric change metric, TCR_sym, to quantitatively characterize the intensity of structural evolution. Experiments demonstrate that CNS exhibits significantly higher change intensity than existing real-world benchmarks (e.g., Oxford, KAIST) and systematically evaluates mainstream LiDAR-PR algorithms, exposing their robustness bottlenecks under severe environmental transformations. This work fills a critical gap in data support for modeling and evaluating long-term outdoor environmental evolution, providing a reproducible, quantitative benchmark platform to drive algorithmic innovation in robust place recognition.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Computer Vision: Visual Reasoning & Symbolic Representations

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Sustainability and carbon-aware systems for Web, mobile, and WoT

Abstract

Large-scale construction and demolition significantly challenge long-term place recognition (PR) by drastically reshaping urban and suburban environments. Existing datasets predominantly reflect limited or indoor-focused changes, failing to adequately represent extensive outdoor transformations. To bridge this gap, we introduce the City that Never Settles (CNS) dataset, a simulation-based dataset created using the CARLA simulator, capturing major structural changes-such as building construction and demolition-across diverse maps and sequences. Additionally, we propose TCR_sym, a symmetric version of the original TCR metric, enabling consistent measurement of structural changes irrespective of source-target ordering. Quantitative comparisons demonstrate that CNS encompasses more extensive transformations than current real-world benchmarks. Evaluations of state-of-the-art LiDAR-based PR methods on CNS reveal substantial performance degradation, underscoring the need for robust algorithms capable of handling significant environmental changes. Our dataset is available at https://github.com/Hyunho111/CNS_dataset.

Problem

• Addresses long-term place recognition under extreme urban structural changes
• Overcomes limitations of existing datasets lacking extensive outdoor transformations
• Proposes new metric for consistent measurement of structural changes

Innovation

• Simulation-based LiDAR dataset for extreme changes
• Symmetric TCR metric for consistent change measurement
• CARLA simulator for diverse structural transformations