Fast and accurate prediction of pedestrian trajectory around vehicles can reduce or even prevent most traffic accidents and improve the safety of traffic participants. This involves the real-time interaction and fusion of various information, such as the vehicle’s motion characteristics, the pedestrian’s historical motion trajectory, and the motion relationship between people and cars. However, most of the existing algorithms use RNN as the skeleton to process the information prediction trajectory, which is weak in extracting the internal relationship between different information, and the running time of the algorithm is long. To solve these problems, we propose a Transformer based deep learning algorithm (CVTF) to complete the first-person perspective pedestrian trajectory prediction task. We have the following innovations about this model: 1: We use the stamp coding method for vehicle speed and pedestrian information to ensure we can learn the characteristics of different information sources. 2: Transformer structure is used, and its attention mechanism is improved (Maybe-self attention), which improves the running speed of the model and reduces the memory consumption. 3: Combined with Conditional Variational Autoencoder (CVAE), hidden variables are introduced to improve the prediction accuracy. Experiments on three pedestrian trajectory prediction benchmarks show that our model achieves the most advanced performance.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Predicting pedestrian tracks around moving vehicles based on conditional variational transformer


    Beteiligte:
    He, Youguo (Autor:in) / Yang, Yongxin (Autor:in) / Cai, Yingfeng (Autor:in) / Yuan, Chaochun (Autor:in) / Shen, Jie (Autor:in) / Tian, Liwei (Autor:in)


    Erscheinungsdatum :

    01.08.2024




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Learning and predicting on-road pedestrian behavior around vehicles

    Deo, Nachiket / Trivedi, Mohan M. | IEEE | 2017


    System for moving vehicles along two tracks

    FISCHER FLORIAN / WOHLGENANNT BERNHARD | Europäisches Patentamt | 2017

    Freier Zugriff

    SYSTEM FOR MOVING VEHICLES ALONG TWO TRACKS

    FISCHER FLORIAN / WOHLGENANNT BERNHARD | Europäisches Patentamt | 2015

    Freier Zugriff

    Pedestrian tracking in infrared from moving vehicles

    Jungling, K / Arens, M | IEEE | 2010


    Pedestrian tracking in infrared from moving vehicles

    Jüngling, K. / Arens, M. | Fraunhofer Publica | 2010

    Freier Zugriff