Supplementary Materials

Was there a specific task in mind? Was there a specific gap that needed to be filled? The dataset was created to address the challenge of early and accurate detection of anomalous events on freeways, such as accidents. The specific task in mind was freeway traffic anomaly detection at the lane level, which involves detecting unusual incidents like vehicle accidents, malfunctions, or severe weather conditions that could affect traffic flow. The dataset fills a significant gap by providing a large-scale, lane-level freeway traffic dataset designed explicitly for anomaly detection, addressing the limitations of existing traffic datasets that either lack anomaly information or have low granularity in data collection. This dataset, called FT-AED (Freeway Traffic Anomalous Event Detection), includes traffic data collected from radar detection sensors and official crash reports from the Nashville Traffic Management Center. The dataset aims to facilitate future research in machine learning and traffic management by providing high-fidelity traffic measurements and ground truth anomaly labels, allowing for the development and benchmarking of new anomaly detection models.