I. Introduction
Place recognition plays a vital role in various fields [1], such as robotics, autonomous vehicles, augmented reality, and more. The main goal of place recognition is to identify previously visited locations based on sensor data and match them with a pre-built map database. For example, in Simultaneous Localization and Mapping (SLAM), place recognition plays an essential role in loop-closure detection, which helps correct the accumulated error in the robot's estimated trajectory. Moreover, in long-term navigation, place recognition assists with re-localization, enabling the robot to determine its position within the map after an extended period or after being temporarily lost.