计算机应用研究2024，Vol.41Issue(11)：3303-3310,8.DOI:10.19734/j.issn.1001-3695.2024.04.0103
融合距离阈值和双向TCN的时空注意力行人轨迹预测模型
Fusion of distance threshold and Bi-TCN for spatio-temporal attention pedestrian trajectory prediction model
摘要
Abstract
In order to solve the problems such as insufficient interaction modeling and low prediction accuracy due to the lack of partial pedestrian modeling ideas,the lack of global vision in time dimension,and the neglect of the diversity of pedestrian interaction modes,this paper proposed an improved model STG-DTBTA based on Social-STGCNN.Firstly,the model con-structed PPM module,and pruned the pedestrians inter links that were not meet constraints such as distance threshold for de-noising.Secondly,the model introduced the spatio-temporal attention mechanism.Spatial attention dynamically assigned in-teractive weights,and set up multiple attention heads to deal with the interaction diversity problem.Temporal attention cap-tured temporal dependencies of temporal data.Finally,the model used Bi-TCN to increase global perspective to capture dy-namic patterns and trends in trajectory data,and used gating mechanism to incorporate the bidirectional features.The experi-mental results on the datasets ETH and UCY show that compared with Social-STGCNN,ADE and FDE are decreased by an ave-rage of 8％and 16％respectively when the number of parameters and the inference time kept close to it.The STG-DTBTA has good interactive modeling ability,model performance and prediction effect.关键词
行人轨迹预测/部分行人建模/距离阈值/时空注意力机制/双向TCN/门控机制Key words
pedestrian trajectory prediction/partial pedestrian model/distance threshold/spatio-temporal attention mecha-nism/bidirectional temporal convolutional network(Bi-TCN)/gating mechanism分类
信息技术与安全科学引用本文复制引用
王红霞,聂振凯,钟强..融合距离阈值和双向TCN的时空注意力行人轨迹预测模型[J].计算机应用研究,2024,41(11):3303-3310,8.基金项目
辽宁省自然科学基金指导计划资助项目(2022-MS-276) （2022-MS-276）
