Recent SLAM Research with HeRCULES 

SLAM Evaluation (in HeRCULES paper)

• For the SLAM baseline, we use Fast-LIO for LiDAR SLAM, 4DRadarSLAM for 4D radar SLAM, and ORORA for spinning radar SLAM. Fig. 1 shows the mapping results for the Sports Complex using ORORA. A comparison of these three baselines on Sports Complex and Library is shown in Fig. 8 and Table. IV.

•  Among these baselines, the odometry result of Fast-LIO was the most accurate, followed by ORORA and 4DRadarSLAM. 

• The result of 4DRadarSLAM is not as good because the point cloud from the Continental radar we used contains fewer points than the Oculii radar originally used in 4DRadarSLAM. 

• However, there is potential for improvement through preprocessing the raw point cloud. These findings validate that SLAM performance with 4D radar alone is limited on our dataset, highlighting the need for heterogeneous radar SLAM or radar-LiDAR fusion SLAM.