Boosting Fine-Grained Urban Flow Inference via Lightweight Architecture and Focalized Optimization

2025-11-10

AI Summary

To address high computational costs from model over-parameterization and the inadequacy of conventional loss functions in handling highly skewed urban flow distributions, this paper proposes a lightweight PLGF architecture and a DualFocal Loss optimization mechanism. PLGF employs a Progressive Local-Global Fusion strategy to jointly capture fine-grained local patterns and long-range spatial dependencies, enhanced by dual-space supervision and hard-sample focusing to improve modeling robustness. DualFocal Loss adaptively modulates class-wise weights and sample-wise gradient magnitudes, effectively mitigating skewness-induced bias. Evaluated on four real-world urban datasets, our method achieves state-of-the-art performance: it reduces model size by up to 97% compared to the best-performing baseline, and under equal parameter budgets, lowers MAE by over 10%. The approach thus delivers both high accuracy and exceptional deployment efficiency, offering a scalable, lightweight solution for urban planning and intelligent transportation systems.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Learning to Search
• Computer Vision: Learning & Optimization for CV

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Fine-grained urban flow inference is crucial for urban planning and intelligent transportation systems, enabling precise traffic management and resource allocation. However, the practical deployment of existing methods is hindered by two key challenges: the prohibitive computational cost of over-parameterized models and the suboptimal performance of conventional loss functions on the highly skewed distribution of urban flows. To address these challenges, we propose a unified solution that synergizes architectural efficiency with adaptive optimization. Specifically, we first introduce PLGF, a lightweight yet powerful architecture that employs a Progressive Local-Global Fusion strategy to effectively capture both fine-grained details and global contextual dependencies. Second, we propose DualFocal Loss, a novel function that integrates dual-space supervision with a difficulty-aware focusing mechanism, enabling the model to adaptively concentrate on hard-to-predict regions. Extensive experiments on 4 real-world scenarios validate the effectiveness and scalability of our method. Notably, while achieving state-of-the-art performance, PLGF reduces the model size by up to 97% compared to current high-performing methods. Furthermore, under comparable parameter budgets, our model yields an accuracy improvement of over 10% against strong baselines. The implementation is included in the https://github.com/Yasoz/PLGF.

Problem

• Reducing computational costs of over-parameterized urban flow inference models
• Improving performance on skewed urban flow distributions via adaptive optimization
• Developing lightweight architecture to capture fine-grained and global dependencies

Innovation

• Progressive Local-Global Fusion lightweight architecture
• DualFocal Loss with dual-space supervision mechanism
• Difficulty-aware focusing on hard-to-predict regions