Overcoming Catastrophic Forgetting in Continual Fine-Grained Urban Flow Inference

Citywide fine-grained urban flow inference (FUFI) problem aims to infer the high-resolution flow maps from the coarse-grained ones, which plays an important role in sustainable and economic urban computing and intelligent traffic management. Previous models tackle this problem from spatial constraint, external factors, and memory cost. However, utilizing the new urban flow maps to calibrate the learned model is very challenging due to the "catastrophic forgetting" problem and is still under-explored. In this article, we make the first step in FUFI and present CUFAR-Continual Urban Flow inference with augmented Adaptive knowledge Replay-a novel framework for inferring the fine-grained citywide traffic flows. Specifically, (1) we design a spatial-temporal inference network that can extract better flow map features from both local and global levels; (2) then, we present an augmented adaptive knowledge replay (AKR) training algorithm to selectively replay the learned knowledge to facilitate the learning process of the model on new knowledge without forgetting. We apply several data augmentation techniques to improve the generalization capability of the learning model, gaining additional performance improvements. We also propose a knowledge discriminator to avoid the "negative replaying" issue introduced by noisy urban flow maps. Extensive experiments on two large-scale real-world FUFI datasets demonstrate that our proposed model consistently outperforms strong baselines and effectively mitigates the forgetting problem.