Aurora Innovation, Inc. (NASDAQ: AUR) today publicly released the Aurora Multi-Sensor Dataset, a large-scale multi-sensor dataset with localization ground truth, in collaboration with the University of Toronto.
The dataset contains rich metadata such as semantic segmentation and spans weather patterns during all four seasons, including rain, snow, overcast and sunny days, as well as different times of day, and a variety of traffic conditions. It is between one and two orders of magnitude larger than other publicly available localization datasets and can be used to develop and evaluate large-scale, long-term approaches to autonomous vehicle localization.
This dataset was previously introduced as PIT30M at the International Conference on Intelligent Robots and Systems (IROS) 2020, one of the world’s top robotics publication venues, in a paper that was nominated as a Finalist for Best Application Paper. Martinez, J., Doubov, S., Fan, J., Wang, S., Máttyus, G., & Urtasun, R. (2020, October). Pit30M: A benchmark for global localization in the age of self-driving cars. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 4477-4484). IEEE.
In making this data broadly available to the academic community, Aurora hopes to contribute to meaningful engineering research and development that will spur continued progress in the autonomous systems field. The Aurora Multi-Sensor Dataset’s size and diversity also make it suitable for a wide range of research areas such as 3D reconstruction, HD map construction, and map compression, among others.
The Aurora Multi-Sensor Dataset was captured between January 2017 and February 2018 in the metropolitan area of Pittsburgh, PA, by Uber Advanced Technologies Group (ATG), which was acquired by Aurora in January 2021. The data was captured using a 64-beam Velodyne HDL-64E LiDAR sensor and seven 1920x1200-pixel resolution cameras including a forward-facing stereo pair and five wide-angle lenses covering a 360º view around the vehicle.
The Aurora Multi-Sensor Dataset is hosted on Amazon Simple Storage Service (S3), made available through the Open Data Sponsorship Program, and is intended for non-commercial academic use. It is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC 4.0).
To access the Aurora Multi-Sensor Dataset, click here.
Aurora (Nasdaq: AUR) is delivering the benefits of self-driving technology safely, quickly, and broadly to make transportation safer, increasingly accessible, and more reliable and efficient than ever before. The Aurora Driver is a self-driving system designed to operate multiple vehicle types, from freight-hauling trucks to ride-hailing passenger vehicles, and underpins Aurora Horizon and Aurora Connect, its driver-as-a-service products for trucking and ride-hailing. Aurora is working with industry leaders across the transportation ecosystem, including Toyota, FedEx, Volvo Trucks, PACCAR, Uber, Uber Freight, U.S. Xpress, Werner, Covenant, Schneider, and Ryder. For Aurora’s latest news, visit aurora.tech and @aurora_inno on Twitter.