🤖 AI Summary
This paper addresses the cross-environment robust localization challenge in SLAM—specifically for loop closure detection and long-term navigation. It presents a systematic survey of scene recognition methods based on CNNs, Transformers, and cross-modal (vision/LiDAR/text) representations. For the first time, it unifies the evolutionary trajectories of these three paradigms and establishes a standardized benchmarking framework encompassing major datasets and evaluation metrics. The work identifies key future research directions: domain adaptation, real-time inference, and continual learning. Under a unified experimental protocol, it conducts the most comprehensive comparative evaluation of state-of-the-art methods to date and releases an open-source, reproducible codebase. The contributions provide theoretical foundations, standardized evaluation protocols, and practical engineering guidance for scene recognition in autonomous driving systems.
📝 Abstract
Place recognition is a cornerstone of vehicle navigation and mapping, which is pivotal in enabling systems to determine whether a location has been previously visited. This capability is critical for tasks such as loop closure in Simultaneous Localization and Mapping (SLAM) and long-term navigation under varying environmental conditions. In this survey, we comprehensively review recent advancements in place recognition, emphasizing three representative methodological paradigms: Convolutional Neural Network (CNN)-based approaches, Transformer-based frameworks, and cross-modal strategies. We begin by elucidating the significance of place recognition within the broader context of autonomous systems. Subsequently, we trace the evolution of CNN-based methods, highlighting their contributions to robust visual descriptor learning and scalability in large-scale environments. We then examine the emerging class of Transformer-based models, which leverage self-attention mechanisms to capture global dependencies and offer improved generalization across diverse scenes. Furthermore, we discuss cross-modal approaches that integrate heterogeneous data sources such as Lidar, vision, and text description, thereby enhancing resilience to viewpoint, illumination, and seasonal variations. We also summarize standard datasets and evaluation metrics widely adopted in the literature. Finally, we identify current research challenges and outline prospective directions, including domain adaptation, real-time performance, and lifelong learning, to inspire future advancements in this domain. The unified framework of leading-edge place recognition methods, i.e., code library, and the results of their experimental evaluations are available at https://github.com/CV4RA/SOTA-Place-Recognitioner.