This dataset was collected as part of the U.S. Department of Transportation (U.S. DOT) Intersection Safety Challenge (hereafter, “the Challenge”) for Stage 1B: System Assessment and Virtual Testing. Multi-sensor data was collected at a controlled test roadway intersection within the Federal Highway Administration (FHWA) Turner-Fairbank Highway Research Center (TFHRC) Smart Intersection facility in McLean, VA from October 2023 through March 2024. The data includes potential conflict-based and non-conflict-based experimental scenarios between vulnerable road users (e.g., pedestrians, bicyclists) and vehicles during both daytime and nighttime conditions. Note that no actual human vulnerable road users were put at risk of being involved in a collision during the data collection efforts. The collected data (hereafter, “the Challenge Dataset”) was intended to be used for intersection safety system algorithm training, refinement, tuning, and/or validation, but may have additional uses. The Challenge Dataset is divided into training data (unlabeled, without associated ground truth) and validation data (labeled, with associated ground truth). This particular data asset provides the validation data, and is complimentary to the training data component of the Challenge Dataset found on the ITS DataHub at: Intersection Safety Challenge Stage 1B Training Data. For a summary of the Stage 1B data collection effort, please see this video: https://youtu.be/csirVHFa2Cc.

All data within the Challenge Dataset (i.e., both the training data and the validation data) includes data collected at a single, signalized four-way intersection from 20 roadside sensors and traffic control devices, including eight closed-circuit television (CCTV) visual cameras, five thermal cameras, two light detection and ranging (LiDAR) sensors, and four radar sensors. Intrinsic calibration was performed for all visual and thermal cameras. Extrinsic calibration was performed for specific pairs of roadside sensors. Additionally, the traffic signal phase and timing data and vehicle and/or pedestrian calls to the traffic signal controller (if any) are also provided in addition to the ground truth position, rotation, and conflict vs. non-conflict labels.

This validation dataset also includes two sets of ground truth: 1) three dimensional (3D) bounding boxes for each road user present at each timestamp within a run; 2) a binary classification label of conflict or no conflict for each run. See the metadata document in the attachments for additional information on the ground truth sets.

The total number of unique runs in the provided Challenge Dataset validation data is 41, bringing the total size of the dataset to approximately 30 GB. A sample of 5 unique runs from the validation data is provided here along with the entire 41 unique runs for download, inspection, and use. It is recommended that the sample be downloaded first for inspection before committing the download time and storage space to the full dataset download.

For more details about the data collection, supplemental files, organization and dictionary, and sensor calibration, see the attached “U.S. DOT ISC Stage 1B ITS DataHub Metadata_v2.0.pdf” document. For more information on the background of the Intersection Safety Challenge Stage 1B, please visit: https://www.its.dot.gov/research-areas/intersection-safety/challenge/.