Towards Efficient Traffic Incident Detection Via Explicit Edge-Level Incident Modeling
Chen Liu,Jiming Chen,Haoyu Liu,Shizhong Li,Shibo He
DOI: https://doi.org/10.1109/jiot.2024.3371482
IF: 10.6
2024-01-01
IEEE Internet of Things Journal
Abstract:Traffic incident detection is a critical task within traffic monitoring systems, enabling on-the-fly alerts for emergency actions. Numerous efforts have been made to detect and localize traffic incidents using data recorded by inductive loop detectors. However, they only focus on the node-level incidents that happen within the surveillance areas and ignore the edge-level ones that take place outside of these areas. In this article, we propose to detect both kinds of incidents simultaneously based on the sparsely distributed sensors. An important challenge is how to explicitly model the edge status and detect this kind of incident. Additionally, capturing complex relationships among traffic dynamics, road locations, and temporal information is nontrivial. In this article, we first describe the traffic dynamics by a fine-grained graph where the sensor range is designed as a hyperparameter to control the coverage boundaries. Then, we propose an edge- and node-aware dual autoencoder (ENDAE), where the correlations are decoupled into internodes, interseries, and interattribute parts, which are further captured via node encoder, temporal encoder, and attribute encoder, respectively. Furthermore, the reconstruction errors are calculated for node-level and edge-level event detection separately. The overall method is evaluated based on two real-world data sets from the Bay Area and Los Angeles in California. ENDAE surpasses all the state-of-the-art methods in both kinds of incidents, with at least a 12.5% improvement in recall and an 18.5% decrease in delay. Notably, for edge-level incidents, ENDAE achieves double the recall of the previous SOTA methods.