HeRCULES: Heterogeneous Radar Dataset in Complex Urban Environment for Multi-session Radar SLAM

• DOI: 10.48550/arxiv.2502.01946
• Publication Date: 2025-02-03
• Authors: Hanjun Kim; Minwoo Jung; Chiyun Noh; Sangwoo Jung; Haiyu Song; Wooseong Yang; Hyesu Jang; Ayoung Kim
• Keywords: FOS: Computer and information sciences; Computer Science - Robotics; Computer Vision and Pattern Recognition (cs.CV); Computer Science - Computer Vision and Pattern Recognition; Robotics (cs.RO)

Recently, radars have been widely featured in robotics for their robustness challenging weather conditions. Two commonly used radar types are spinning and phased-array radars, each offering distinct sensor characteristics. Existing datasets typically feature only a single type of radar, leading to the development algorithms limited that specific kind. In this work, we highlight combining different offers complementary advantages, which can be leveraged through heterogeneous dataset. Moreover, new dataset fosters research multi-session multi-robot scenarios where robots equipped with radars. context, introduce HeRCULES dataset, comprehensive, multi-modal FMCW LiDAR, IMU, GPS, cameras. This is first integrate 4D alongside unparalleled localization, mapping, place recognition capabilities. The covers diverse lighting conditions range urban traffic scenarios, enabling comprehensive analysis across various environments. sequence paths multiple revisits ground truth pose enhance its suitability research. We expect facilitate odometry, recognition, fusion tools available at https://sites.google.com/view/herculesdataset.