In this paper, we propose the use of semantic segmentation to incorporate scene information for better understanding of human motion in crowded environments. Our proposed SSeg-LSTM method leverages SegNet, which is a semantic segmentation encoder-decoder architecture, to extract semantically meaningful scene features. We then train the Social Scene LSTM (SS-LSTM) model with the contextual information regarding dynamics, social neighborhood, and scene semantics to predict future trajectory points of pedestrians. Experimental evaluation on public datasets show better performance for SSeg-LSTM than SS-LSTM which highlights the utility of semantic encoding for trajectory prediction.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    SSeg-LSTM: Semantic Scene Segmentation for Trajectory Prediction


    Contributors:


    Publication date :

    2019-06-01


    Size :

    3475711 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Learning Semantic Scene Models by Trajectory Analysis

    Wang, X. / Tieu, K. / Grimson, E. | British Library Conference Proceedings | 2006


    Pedstrian Trajectory Prediction Based on LSTM

    Li, Shaosong / Jiang, Junchen / Zhou, Qingbin et al. | IEEE | 2023


    An LSTM network for highway trajectory prediction

    Altche, Florent / de La Fortelle, Arnaud | IEEE | 2017


    An LSTM Network for Highway Trajectory Prediction

    Altché, Florent / de La Fortelle, Arnaud | ArXiv | 2018

    Free access

    Ship Trajectory Prediction with Social History LSTM

    Zhao, Wenfeng / Zhang, Xudong | IEEE | 2023