The invention provides an end-to-end multi-modal trajectory prediction method based on dynamic graph convolution, the prediction method is realized through a DG-Trajector network model, the DG-Trajector network model comprises a DGANet feature extraction module, an end point generation module and a trajectory prediction module, the DGANet feature extraction module carries out feature coding and feature extraction on vectorized features, and the end point generation module carries out feature extraction on the vectorized features; and inputting the extracted features and the anchor points into an end point generation module, further generating position information of track end points, and finally predicting a group of tracks by utilizing the generated track end point information through a track prediction module. Besides, the DG-Trajector model comprehensively considers endpoint loss, prediction loss and prediction score loss, target prediction and trajectory prediction tasks can be comprehensively optimized, and the performance and generalization ability of the model are improved, so that the DG-Trajector model can better adapt to various complex actual scenes.

    本发明提供一种基于动态图卷积的端到端多模态轨迹预测方法,该预测方法通过DG‑Trajector网络模型实现,DG‑Trajector网络模型包括DGANet特征提取模块、端点生成模块和轨迹预测模块,其中,通过DGANet特征提取模块对矢量化特征进行特征编码和特征提取,将提取的特征与锚点输入端点生成模块,进一步生成轨迹端点的位置信息,最后,通过轨迹预测模块来利用生成的轨迹端点信息预测一组轨迹。另外,DG‑Trajector模型通过综合考虑端点损失、预测损失和预测评分损失,能够全面优化目标预测和轨迹预测任务,提高模型的性能和泛化能力,从而更好地适应各种复杂的实际场景。


    Access

    Download


    Export, share and cite



    Title :

    End-to-end multi-modal trajectory prediction method based on dynamic graph convolution


    Additional title:

    一种基于动态图卷积的端到端多模态轨迹预测方法


    Contributors:
    LI HAISHENG (author) / MA HUIJIA (author)

    Publication date :

    2024-11-01


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06V



    Multi-modal vehicle trajectory prediction method based on graph interaction mechanism

    GAO JIANPING / YANG YU / LIU PAN et al. | European Patent Office | 2024

    Free access

    Multi-modal trajectory prediction method

    JIANG WENJUAN / JIN ZHI / WANG REN et al. | European Patent Office | 2023

    Free access

    Traffic prediction method based on dynamic graph convolution

    FAN JIN / WENG WENCHAO / TIAN HAO et al. | European Patent Office | 2023

    Free access

    Dynamic graph convolution traffic speed prediction method

    LIU QILIANG / YUAN HAOTAO / YANG LIU et al. | European Patent Office | 2020

    Free access

    Multi-modal multi-agent trajectory prediction

    SUN PEI / ZHAO HANG / MCCAULEY ALEXANDER et al. | European Patent Office | 2024

    Free access