AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction

doi:10.1016/j.neucom.2021.03.024

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统管理与控制国家重点实验室 > 机器人应用与理论组

	AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction
	Zhou, Hao 1,4; Ren, Dongchun 2; Xia, Huaxia 2; Fan, Mingyu 2,3; Yang, Xu 4; Huang, Hai 1
刊名	NEUROCOMPUTING
	2021-07-20
卷号	445 页码:298-308
关键词	Pedestrian trajectory prediction Graph neural networks Spatio-temporal graph Graph attention
ISSN号	0925-2312
DOI	10.1016/j.neucom.2021.03.024
通讯作者	Yang, Xu(xu.yang@ia.ac.cn) ; Huang, Hai(haihus@163.com)
英文摘要	Predicting pedestrian trajectories in the future is a basic research topic in many real applications, such as video surveillance, self-driving cars, and robotic systems. There are two major challenges in this task, the complex interaction modeling among pedestrians and the unique motion pattern extraction for each pedestrian. Regarding the two challenges, an attention-based interaction-aware spatio-temporal graph neural network is proposed for predicting pedestrian trajectories. There are two components in the proposed method: spatial graph neural network for interaction modeling, and temporal graph neural network for motion feature extraction. Spatial graph neural network uses an attention mechanism to capture the spatial interactions among all the pedestrians at each time step. Meanwhile, temporal graph neural network uses an attention mechanism to capture the temporal motion pattern of each pedestrian. Finally, a time-extrapolator convolutional neural network is used in the temporal dimension of the aggregated graph features to predict the future trajectories. Experimental results on two benchmark pedestrian trajectory prediction datasets demonstrate the competitive performances of the proposed method in terms of both the final displace error and the average displacement error metrics as compared with state-of-the-art trajectory prediction methods. (C) 2021 Elsevier B.V. All rights reserved.
资助项目	Beijing Nova Program[Z201100006820046] ; National Naturral Science Foundation of China[61973301] ; National Naturral Science Foundation of China[61972020] ; National Naturral Science Foundation of China[61633009] ; National Naturral Science Foundation of China[61772373] ; National Naturral Science Foundation of China[51579053] ; National Naturral Science Foundation of China[U1613213] ; National Key R&D Program of China[2016YFC0300801] ; National Key R&D Program of China[2017YFB1300202] ; Field Fund of the 13th FiveYear Plan for Equipment Preresearch Fund[61403120301] ; Key Basic Research Project of Shanghai Science and Technology Innovation Plan[15JC1403300] ; Beijing Science and Technology Project[Z181100008918018] ; Meituan Open RD Fund
WOS研究方向	Computer Science
语种	英语
出版者	ELSEVIER
WOS记录号	WOS:000652811700007
资助机构	Beijing Nova Program ; National Naturral Science Foundation of China ; National Key R&D Program of China ; Field Fund of the 13th FiveYear Plan for Equipment Preresearch Fund ; Key Basic Research Project of Shanghai Science and Technology Innovation Plan ; Beijing Science and Technology Project ; Meituan Open RD Fund
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/45211]
专题	自动化研究所_复杂系统管理与控制国家重点实验室_机器人应用与理论组
通讯作者	Yang, Xu; Huang, Hai
作者单位	1.Harbin Engn Univ, Natl Key Lab Sci & Technol Underwater Vehicle, Harbin, Peoples R China 2.Meituan, Beijing, Peoples R China 3.Wenzhou Univ, Coll Comp Sci & Artificial Intelligence, Wenzhou, Peoples R China 4.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Zhou, Hao,Ren, Dongchun,Xia, Huaxia,et al. AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction[J]. NEUROCOMPUTING,2021,445:298-308.
APA	Zhou, Hao,Ren, Dongchun,Xia, Huaxia,Fan, Mingyu,Yang, Xu,&Huang, Hai.(2021).AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction.NEUROCOMPUTING,445,298-308.
MLA	Zhou, Hao,et al."AST-GNN: An attention-based spatio-temporal graph neural network for Interaction-aware pedestrian trajectory prediction".NEUROCOMPUTING 445(2021):298-308.