A Bi-LSTM approach for modelling movement uncertainty of crowdsourced human trajectories under complex urban environments

Yue Yu, Yepeng Yao, Zhewei Liu, Zhenlin An, Bi Yu Chen, Liang Chen, Ruizhi Chen

Abstract: Highlights•A Bi-LSTM framework is proposed for modelling the uncertainty of crowdsourced human trajectories.•The pedestrian motion detection characteristic and GNSS originated location are combined in the model.•The proposed model outperforms baseline methods in achieving more adaptive and accurate movement uncertainty prediction.