Traffic Forecasting on Traffic Moving Snippets

TL;DR

This paper introduces a patch-based traffic forecasting method using a Unet++ architecture, improving accuracy and generalization to unseen cities by processing traffic data in overlapping snippets.

Contribution

The novel approach of predicting small city snippets instead of full city rasters enhances model robustness and accuracy in traffic forecasting, especially for unseen locations.

Findings

Patch-wise prediction improves accuracy over full-city models.

Unet++ architecture enhances traffic prediction performance.

Processing multiple patches per sample increases robustness.

Abstract

Advances in traffic forecasting technology can greatly impact urban mobility. In the traffic4cast competition, the task of short-term traffic prediction is tackled in unprecedented detail, with traffic volume and speed information available at 5 minute intervals and high spatial resolution. To improve generalization to unknown cities, as required in the 2021 extended challenge, we propose to predict small quadratic city sections, rather than processing a full-city-raster at once. At test time, breaking down the test data into spatially-cropped overlapping snippets improves stability and robustness of the final predictions, since multiple patches covering one cell can be processed independently. With the performance on the traffic4cast test data and further experiments on a validation set it is shown that patch-wise prediction indeed improves accuracy. Further advantages can be gained with a Unet++ architecture and with an increasing number of patches per sample processed at test time. We conclude that our snippet-based method, combined with other successful network architectures proposed in the competition, can leverage performance, in particular on unseen cities. All source code is available at https://github.com/NinaWie/NeurIPS2021-traffic4cast.