Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction

TLDR

Forecasting crowd flow is critical for traffic management and public safety, yet it is difficult because it depends on complex factors such as inter‑region traffic, events, and weather. This study introduces ST‑ResNet, a deep‑learning model that jointly predicts the inflow and outflow of crowds in every city region. ST‑ResNet is an end‑to‑end residual network that models temporal closeness, period, and trend through separate branches of residual convolutional units, dynamically aggregates their outputs with learned weights, and incorporates external variables like weather and day of the week. Experiments on Beijing and New York City data show that ST‑ResNet outperforms six state‑of‑the‑art baselines.

Abstract

Forecasting the flow of crowds is of great importance to traffic management and public safety, and very challenging as it is affected by many complex factors, such as inter-region traffic, events, and weather. We propose a deep-learning-based approach, called ST-ResNet, to collectively forecast the inflow and outflow of crowds in each and every region of a city. We design an end-to-end structure of ST-ResNet based on unique properties of spatio-temporal data. More specifically, we employ the residual neural network framework to model the temporal closeness, period, and trend properties of crowd traffic. For each property, we design a branch of residual convolutional units, each of which models the spatial properties of crowd traffic. ST-ResNet learns to dynamically aggregate the output of the three residual neural networks based on data, assigning different weights to different branches and regions. The aggregation is further combined with external factors, such as weather and day of the week, to predict the final traffic of crowds in each and every region. Experiments on two types of crowd flows in Beijing and New York City (NYC) demonstrate that the proposed ST-ResNet outperforms six well-known methods.