Towards Good Practices of U-Net for Traffic Forecasting

TL;DR

This paper explores using U-Net for traffic forecasting by leveraging city roadmaps and a fine-tuning strategy, achieving improved predictions in the Traffic4Cast Challenge and discussing future enhancements.

Contribution

It introduces a U-Net based approach with a novel roadmap generation method and validation-based fine-tuning for better traffic prediction accuracy.

Findings

Effective traffic flow predictions using U-Net with prior city roadmaps

Fine-tuning on validation set reduces overfitting and improves results

Discussion of future directions like seasonality and knowledge transfer

Abstract

This technical report presents a solution for the 2020 Traffic4Cast Challenge. We consider the traffic forecasting problem as a future frame prediction task with relatively weak temporal dependencies (might be due to stochastic urban traffic dynamics) and strong prior knowledge, \textit{i.e.}, the roadmaps of the cities. For these reasons, we use the U-Net as the backbone model, and we propose a roadmap generation method to make the predicted traffic flows more rational. Meanwhile, we use a fine-tuning strategy based on the validation set to prevent overfitting, which effectively improves the prediction results. At the end of this report, we further discuss several approaches that we have considered or could be explored in future work: (1) harnessing inherent data patterns, such as seasonality; (2) distilling and transferring common knowledge between different cities. We also analyze the validity of the evaluation metric.