Nowadays, many cities are equipped with surveillance systems and traffic control centers to monitor vehicular traffic for road safety and efficiency. The monitoring process is mostly done manually which is inefficient and expensive. In recent years, several data-driven solutions have been proposed in the literature to automatically analyze traffic flow data using machine learning techniques. However, existing solutions require large and comprehensive datasets for training which are not readily available, thus limiting their application. In this paper, we develop a traffic anomaly detection system, referred to as DeepFlow, based on Siamese neural networks, which are suitable in scenarios where only small datasets are available for training. Our model can detect abnormal traffic flows by analyzing the trajectory data collected from the vehicles in a fleet. To evaluate DeepFlow, we use realistic vehicular traffic simulations in SUMO. Our results show that DeepFlow detects abnormal traffic patterns with an F1 score of 78%, while outperforming other existing approaches including: Dynamic Time Warping (DTW), Global Alignment Kernels (GAK), and iForest.


    Access

    Download


    Export, share and cite



    Title :

    DeepFlow: Abnormal Traffic Flow Detection Using Siamese Networks


    Contributors:

    Publication date :

    2021


    Remarks:

    7 pages, 12 figures, 3 tables


    Type of media :

    Preprint


    Type of material :

    Electronic Resource


    Language :

    English





    Traffic abnormal flow causal detection method and device

    SONG XUAN / GAO HAOTIAN / FAN ZIPEI et al. | European Patent Office | 2023

    Free access

    SIAMESE-RESNET: IMPLEMENTING LOOP CLOSURE DETECTION BASED ON SIAMESE NETWORK

    Qiu, Kai / Ai, Yunfeng / Tian, Bin et al. | British Library Conference Proceedings | 2018


    Detection of Abnormal Traffic Flow with the Hurst Parameter for Traffic Management

    Karelina, E. A. / Stepanov, A. A. / Smirnov, P. I. et al. | IEEE | 2023


    Siamese-ResNet: Implementing Loop Closure Detection based on Siamese Network

    Qiu, Kai / Ai, Yunfeng / Tian, Bin et al. | IEEE | 2018