With over 13 hours of highly dynamic urban scenarios from autonomous agents and expert drivers, our team is working on research problems ranging from robust sensor fusion, trajectory prediction, dynamic planning and automatic calibration. To learn more about our data and releases, please see the sections below.
In recent years, High Definition (HD) maps have become the core dependency for various state-of-the-art autonomous driving stacks; however, these often require centimeter-level annotations defined manually in order for an autonomous vehicle to be able to identify feasible paths. To motivate research directions using light-weight and fully automated map representations, we present NominalScenes.
This data augments the NominalScenes dataset by incorporating intersection specific scenarios with various configurations including three-way and four-way intersections.
With 942, 161 individually labeled frames with 2D and 3D centerline and lane-level definitions, this dataset provides a train, validation, and test split across three front facing cameras. The approach leverages an automatic labeling pipeline to process over 1,000 unique driving scenarios.
More details coming soon
At AVL, we leverage road furniture for automatic camera calibration. This dataset contains over 3,100 stop signs that have been used to develop our automatic intrinsic calibration pipeline.
© 2020 Autonomous Vehicle Laboratory