Dual Encoding U-Net for Spatio-Temporal Domain Shift Frame Prediction

TL;DR

This paper introduces a lightweight Dual-Encoding U-Net with novel skip-connections for accurate traffic frame prediction amid changing city mobility patterns due to COVID-19, effectively modeling spatio-temporal domain shifts.

Contribution

It presents a novel dual-encoding U-Net architecture with skip-connections tailored for spatio-temporal domain shift in traffic prediction tasks.

Findings

Effective modeling of post-COVID mobility behavior.

Lightweight model with only 12 convolutional layers.

Handles domain shifts between pre- and post-pandemic data.

Abstract

The landscape of city-wide mobility behaviour has altered significantly over the past 18 months. The ability to make accurate and reliable predictions on such behaviour has likewise changed drastically with COVID-19 measures impacting how populations across the world interact with the different facets of mobility. This raises the question: "How does one use an abundance of pre-covid mobility data to make predictions on future behaviour in a present/post-covid environment?" This paper seeks to address this question by introducing an approach for traffic frame prediction using a lightweight Dual-Encoding U-Net built using only 12 Convolutional layers that incorporates a novel approach to skip-connections between Convolutional LSTM layers. This approach combined with an intuitive handling of training data can model both a temporal and spatio-temporal domain shift (gitlab.com/alchera/alchera-traffic4cast-2021).