The UrbanV2X Datasets

Due to the limitations of single autonomous vehicles, Cellular Vehicle-to-Everything (C-V2X) technology opens a new window for achieving fully autonomous driving through sensor information sharing. However, real-world datasets supporting vehicle–infrastructure cooperative navigation in complex urban environments remain rare.

To address this gap, we present UrbanV2X, a comprehensive multisensory dataset collected from both vehicles and roadside infrastructure in the Hong Kong C-V2X testbed, designed to support research on smart mobility applications in dense urban areas.

Our onboard platform provides time-synchronized data from multiple industrial cameras, LiDARs, 4D radar, UWB, IMU, and high-precision GNSS-RTK/INS navigation systems. Meanwhile, our roadside infrastructure provides LiDAR, GNSS, and UWB measurements.

The entire vehicle–infrastructure platform is synchronized using the Precision Time Protocol (PTP), with sensor calibration data provided. We also benchmark various navigation algorithms to evaluate the collected cooperative data.

The contributions of our work can be summarized as follows:

UrbanV2X Dataset and Sensor Integration: Our sensor platform integrating various sensors like multiple cameras, GNSS/INS, 4D Radar, LiDAR and sky-pointing cameara with roadside infrastructure, including GNSS, LiDAR, and UWB.
UrbanV2X collects comprehensive data from both urban canyons and open areas. These equences are recorded under various conditions. Our dataset and benchmark results are publicly available on our website.

BibTeX

Please cite the following publication when using this benchmark in an academic context:

Qin, Q., Zhang, Z., Zhong, Y., Huang, F., Liu, X., Hu, R., Chen, H., Hu, W., Su, D., Zhang, J., Ng, H.-F., & Wen, W. (2025). UrbanV2X: A Multisensory Vehicle-Infrastructure Dataset for Cooperative Navigation in Urban Areas

Leveraging multiple sensors enhances complex environmental perception and increases resilience to varying luminance conditions and high-speed motion patterns, achieving precise localization and mapping. This paper proposes, UrbanV2X, an event-centric multisensory dataset containing 81 sequences and covering over 200 km of various challenging driving scenarios including high-speed motion, repetitive scenarios, dynamic objects, etc. UrbanV2X provides data from two sets of stereo event cameras with different resolutions (640×480, 346×260), stereo industrial cameras, an infrared camera, a top-installed mechanical LiDAR with two slanted LiDARs, two consumer-level GNSS receivers, and an onboard IMU. Meanwhile, the ground-truth of the vehicle was obtained using a centimeter-level high-accuracy GNSS-RTK/INS navigation system. All sensors are well-calibrated and temporally synchronized at the hardware level, with recording data simultaneously. We additionally evaluate several state-of-the-art SLAM algorithms for benchmarking visual and LiDAR SLAM and identifying their limitations.

License

This work is released under GPLv3 license. For commercial inquires, please contact Dr. Wen Weisong(welson.wen@polyu.edu.hk).

Acknowledgement

The authors thank Yuteng Wang, Shaoting Qiu from PolyU, Jiashi Feng, Alpamys Urtay and Siqiao from ASTRI for their kind support in this data evaluation. And the template from ECMD datasets.