Crowd flow prediction in high density urban scenes is involved in a wide range of intelligent transportation and smart city applications, and it has become a significant topic in urban computing. In this letter, a CNN-based framework called Pyramidal Spatio-Temporal Network (PSTNet) for crowd flow prediction is proposed. Spatial encoding is employed for spatial representation of external factors, while prior pyramid enhances feature dependence of spatial scale distances and temporal spans, after that, post pyramid is proposed to fuse the heterogeneous spatio-temporal features of multiple scales. Experimental results based on TaxiBJ and MobileBJ demonstrate that proposed PSTNet outperforms the state-of-the-art methods.
Enze YANG
China Academy of Railway Sciences
Shuoyan LIU
China Academy of Railway Sciences
Yuxin LIU
China Academy of Railway Sciences
Kai FANG
China Academy of Railway Sciences
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Enze YANG, Shuoyan LIU, Yuxin LIU, Kai FANG, "PSTNet: Crowd Flow Prediction by Pyramidal Spatio-Temporal Network" in IEICE TRANSACTIONS on Information,
vol. E104-D, no. 10, pp. 1780-1783, October 2021, doi: 10.1587/transinf.2020EDL8111.
Abstract: Crowd flow prediction in high density urban scenes is involved in a wide range of intelligent transportation and smart city applications, and it has become a significant topic in urban computing. In this letter, a CNN-based framework called Pyramidal Spatio-Temporal Network (PSTNet) for crowd flow prediction is proposed. Spatial encoding is employed for spatial representation of external factors, while prior pyramid enhances feature dependence of spatial scale distances and temporal spans, after that, post pyramid is proposed to fuse the heterogeneous spatio-temporal features of multiple scales. Experimental results based on TaxiBJ and MobileBJ demonstrate that proposed PSTNet outperforms the state-of-the-art methods.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2020EDL8111/_p
Copy
@ARTICLE{e104-d_10_1780,
author={Enze YANG, Shuoyan LIU, Yuxin LIU, Kai FANG, },
journal={IEICE TRANSACTIONS on Information},
title={PSTNet: Crowd Flow Prediction by Pyramidal Spatio-Temporal Network},
year={2021},
volume={E104-D},
number={10},
pages={1780-1783},
abstract={Crowd flow prediction in high density urban scenes is involved in a wide range of intelligent transportation and smart city applications, and it has become a significant topic in urban computing. In this letter, a CNN-based framework called Pyramidal Spatio-Temporal Network (PSTNet) for crowd flow prediction is proposed. Spatial encoding is employed for spatial representation of external factors, while prior pyramid enhances feature dependence of spatial scale distances and temporal spans, after that, post pyramid is proposed to fuse the heterogeneous spatio-temporal features of multiple scales. Experimental results based on TaxiBJ and MobileBJ demonstrate that proposed PSTNet outperforms the state-of-the-art methods.},
keywords={},
doi={10.1587/transinf.2020EDL8111},
ISSN={1745-1361},
month={October},}
Copy
TY - JOUR
TI - PSTNet: Crowd Flow Prediction by Pyramidal Spatio-Temporal Network
T2 - IEICE TRANSACTIONS on Information
SP - 1780
EP - 1783
AU - Enze YANG
AU - Shuoyan LIU
AU - Yuxin LIU
AU - Kai FANG
PY - 2021
DO - 10.1587/transinf.2020EDL8111
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E104-D
IS - 10
JA - IEICE TRANSACTIONS on Information
Y1 - October 2021
AB - Crowd flow prediction in high density urban scenes is involved in a wide range of intelligent transportation and smart city applications, and it has become a significant topic in urban computing. In this letter, a CNN-based framework called Pyramidal Spatio-Temporal Network (PSTNet) for crowd flow prediction is proposed. Spatial encoding is employed for spatial representation of external factors, while prior pyramid enhances feature dependence of spatial scale distances and temporal spans, after that, post pyramid is proposed to fuse the heterogeneous spatio-temporal features of multiple scales. Experimental results based on TaxiBJ and MobileBJ demonstrate that proposed PSTNet outperforms the state-of-the-art methods.
ER -