Multi-Source Urban Traffic Prediction using Drone and Loop Detector Data

Abstract

Traffic forecasting has been a fundamental task in transportation research, with many methods and datasets mainly based on highway loop detector data. In recent years, drones are becoming a favorable choice for urban traffic monitoring, due to their flexibility, high data quality and larger spatial coverage. However, the lack of public datasets has made the joint use of drone and loop detector data fairly under-explored. Therefore, we create a novel simulated multi-source dataset SimBarca for urban traffic prediction, featuring speed measurements from both drones and loop detectors. We provide a graph-based baseline model HiMSNet to handle multiple input modalities and evaluate it along with two benchmark predictors. Our analysis shows that HiMSNet achieves good performance for regional speed prediction, but the road segment-level prediction still requires more in-depth efforts.