Sensor Configuration
4D Radar (Continental ARS548)
Spinning Radar(Navtech RAS6)
FMCW LiDAR (Aeva Aeries II)
Stereo Camera (FLIR Blackfly S)
IMU (Xsens MTi-300)
RTK-GPS (SPAN-CPT7)
Sensor Specifications
Data Format
4D Radar & FMCW LiDAR
Binary file
Spinning Radar
PNG format image
Stereo camera
PNG format image
Individual Sensor Groundtruth
TXT format with [time, x, y, z, qx, qy, qz, qw]
Calibration parameter
TXT format with rotation matrix and translation vector
Inertial data
CSV format for raw measurement from IMU, INS
4D Radar Data
4D Radar was acquired at a frequency of 20Hz and saved as a bin file in the format. The filename of each binary file corresponds to the acquisition timestamp, and these timestamps are also archived in stamp.csv. The file player releases the topics grounded on this timestamp file.
Spinning Radar Data
Spinning radar data were collected at 4 Hz. For each sweep we provide both a polar-domain image and a Cartesian-domain image, saved in PNG format. The polar image resolution is 400 × 3424 (width × height). This dual representation enables users to work either directly with the native polar measurements or with a pre-projected Cartesian view for downstream perception tasks.
LiDAR Data
FMCW LiDAR was acquired at a frequency of 10Hz and saved as a bin file in the format. The filename of each binary file corresponds to the acquisition timestamp, and these timestamps are also archived in stamp.csv. The file player releases the topics grounded on this timestamp file.
Stereo Camera Data
Stereo camera data were recorded at 15 Hz and saved as PNG images at 1440 × 1080 (width × height). To protect privacy, faces and vehicle license plates were anonymized using the EgoBlur system from Project Aria (see: https://www.projectaria.com/tools/egoblur/). The anonymization is applied consistently across frames to preserve temporal coherence while removing personally identifiable information.
IMU Data
The data structure for the IMU is organized in the following order: [timestamp, qx, qy, qz, qw, eul x, eul y, eul z, gyr x, gyr y, gyr z, acc x, acc y, acc z, mag x, mag y, mag z]. This information is stored in the 'xsens_imu.csv' file. Additionally, it is broadcasted as a ROS topic under the name '/imu/data_raw', which publishes data in the format [timestamp, quaternion, angular_velocity, linear_acceleration] at a frequency of 100Hz.
INS Data
The INS data is stored in the 'inspva.csv' file, organized in the following sequence: [timestamp, latitude, longitude, height, north velocity, east velocity, up velocity, roll, pitch, azimuth, status]. This data serves as the foundation for creating the ground truth and was acquired at a frequency of 50Hz.
Individual Ground Truth in the HeRCULES Dataset
A notable contribution of the HeRCULES dataset is its provision of individual sensor ground truth. Given that every sensor has distinct acquisition timestamp and installation points, the scanning positions differ due to both spatial and temporal variations.
Since place recognition evaluations are predominantly decided by range, the precise positioning of these data points holds paramount importance.
Recognizing this, the HeRCULES Dataset offers individual sensor groundtruth. These ground truths are formulated based on the extrinsic calibration of each sensor and the B-Spline Interpolation, which is tailored to each sensor's unique acquisition timestamp.
Sensor Calibration of the HeRCULES Dataset
Extrinsic Calibration of LiDAR - Spinning Radar: We employ the method used in the Boreas dataset.
Extrinsic Calibration of LiDAR - 4D Radar - Camera: We utilize the calibration tool for cameras, LiDAR, and radar.
Extrinsic Calibration of LiDAR - IMU: We initialize the system using the method proposed by Zhu et al.