The UrbanV2X Datasets

Data Format

We provide various sensor data in the rosbag file, including 4D Radar, sky-pointing camera, 7 surround cameras, IMU, LiDAR, UWB, and GNSS. Additionally, a text file provides ground truth data as well as data and parameters from various roadside equipment.

/Target_Radar_1 (sensor_msgs/PointCloud2): 3D point clouds from 4D Radar GPAL-Ares-R7861.
/Target_Radar_1_extreme (sensor_msgs/PointCloud2): Extreme 3D point clouds from 4D Radar GPAL-Ares-R7861.
/camera/image_color/compressed (sensor_msgs/CompressedImage): Color images from the sky-pointing camera.
/hikcamera/image_0/compressed (sensor_msgs/CompressedImage): Color images from the right-back camera.
/hikcamera/image_1/compressed (sensor_msgs/CompressedImage): Color images from the front camera (right).
/hikcamera/image_2/compressed (sensor_msgs/CompressedImage): Color images from the front camera (left).
/hikcamera/image_3/compressed (sensor_msgs/CompressedImage): Color images from the left-front camera.
/hikcamera/image_4/compressed (sensor_msgs/CompressedImage): Color images from the right-front camera.
/hikcamera/image_5/compressed (sensor_msgs/CompressedImage): Color images from the left-back camera.
/hikcamera/image_6/compressed (sensor_msgs/CompressedImage): Color images from the rear camera.
/imu/data (sensor_msgs/Imu): 9-axis IMU from Xsens Mti-30.
/lidar_points (sensor_msgs/PointCloud2): Point clouds from the front-installed LiDAR.
/nlink_linktrack_nodeframe3 (nlink_parser/LinktrackNodeframe3): UWB data from Nooploop P-B.
/ublox_driver/ephem (gnss_comm/GnssEphemMsg): Events from the right DVXplorer.
/ublox_driver/glo_ephem (gnss_comm/GnssGloEphemMsg): 6-axis IMU from the right DVXplorer.
/ublox_driver/iono_params (gnss_comm/StampedFloat64Array): Color images from the left standard cameras.
/ublox_driver/range_meas (gnss_comm/GnssMeasMsg): Color images from the right standard cameras.
/ublox_driver/receiver_lla (sensor_msgs/NavSatFix): Infrared images from the hikrobot infrared camera.
/ublox_driver/receiver_pvt (gnss_comm/GnssPVTSolnMsg): Point clouds from the left slanted LiDAR.
/velodyne_points (sensor_msgs/PointCloud2): Point clouds from the top-installed LiDAR.
/imu/data (sensor_msgs/Imu): 9-axis IMU from the Xsens Mti-30.
gt.txt (TUM format): The ground truth signals from centimeter-level GNSS-RTK/INS system.

The following content is divided into two sections:

1. Visualization.

2. Summary Table.

Visualization of each sequence

Small Loop Scenario (Download Link)

small_loop

Rosbag (Ego): small_loop.bag (34.92 G)

Ground Truth: gt_small_loop.txt (30.04 MB)

Roadside LiDAR (RSI2): small_loop-RSI2_LiDAR (1.27 GB)

Roadside GNSS (RSG0: HKSTP West) Format: RINEX 3.02, PPP Pose: 22.42516252,114.21146463,3.0557 : west_rsi_gnss_small_loop.o

Roadside GNSS (RSG3: HKSTP East High) Format: RINEX 3.02, PPP Pose: 22.42744535, 114.21107245, 11.2765 : east_high_small_loop.o

Large Loop Scenario (Download Link)

large_loop

Rosbag (Ego): large_loop.bag (62.21 G)

Ground Truth: gt_large_loop.txt (77.36 MB)

Roadside LiDAR (RSI2): large_loop-RSI2_LiDAR (1.16 GB)

Roadside GNSS (RSG0: HKSTP West) Format: RINEX 3.02, PPP Pose: 22.42516252,114.21146463,3.0557: west_rsi_gnss_large_loop.o

Roadside GNSS (RSG3: HKSTP East High) Format: RINEX 3.02, PPP Pose: 22.42744535, 114.21107245, 11.2765 : east_high_large_loop.o

Summarization

Scene	Duration	Total Size
Small Loop Scenario (Download Link)	309s	34.9 GB
Large Loop Scenario (Download Link)	600s	62.2 GB