UniFlow: A Foundation Model for Unified Urban Spatio-Temporal Flow Prediction

TL;DR

UniFlow is a unified foundation model for urban spatio-temporal flow prediction that integrates grid and graph data using a novel transformer architecture and memory augmentation, outperforming existing models especially with limited data.

Contribution

The paper introduces UniFlow, a novel unified model combining grid and graph data for urban flow prediction, with a new spatio-temporal transformer and memory retrieval mechanism.

Findings

Outperforms existing models in both grid and graph scenarios.

Effective in low-data situations, maintaining high prediction accuracy.

Demonstrates broad applicability across different urban flow datasets.

Abstract

Urban spatio-temporal flow prediction, encompassing traffic flows and crowd flows, is crucial for optimizing city infrastructure and managing traffic and emergency responses. Traditional approaches have relied on separate models tailored to either grid-based data, representing cities as uniform cells, or graph-based data, modeling cities as networks of nodes and edges. In this paper, we build UniFlow, a foundational model for general urban flow prediction that unifies both grid-based and graphbased data. We first design a multi-view spatio-temporal patching mechanism to standardize different data into a consistent sequential format and then introduce a spatio-temporal transformer architecture to capture complex correlations and dynamics. To leverage shared spatio-temporal patterns across different data types and facilitate effective cross-learning, we propose SpatioTemporal Memory Retrieval Augmentation (ST-MRA). By creating structured memory modules to store shared spatio-temporal patterns, ST-MRA enhances predictions through adaptive memory retrieval. Extensive experiments demonstrate that UniFlow outperforms existing models in both grid-based and graph-based flow prediction, excelling particularly in scenarios with limited data availability, showcasing its superior performance and broad applicability. The datasets and code implementation have been released on https://github.com/YuanYuan98/UniFlow.