---
title: "Coarse-to-Fine Latent Guidance: A Multi-Scale Diffusion Transformer for Traffic Flow Forecasting"
authors:
  - "Zetao Li"
  - "Silin Zhou"
  - "Zheng Hu"
  - "Shimin Cai"
  - "Tao Zhou"
doi: "10.24963/ijcai.2026/311"
url: "https://ijcai.org/proceedings/2026/311"
pdf: "https://ijcai.org/proceedings/2026/0311.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "2798-2806"
track: "Main Track"
keywords:
  - "Data Mining: Mining spatial and/or temporal data"
---

# Coarse-to-Fine Latent Guidance: A Multi-Scale Diffusion Transformer for Traffic Flow Forecasting

Zetao Li, Silin Zhou, Zheng Hu, Shimin Cai, Tao Zhou

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 2798-2806. DOI: [10.24963/ijcai.2026/311](https://doi.org/10.24963/ijcai.2026/311)

## Abstract

Accurate traffic flow prediction is fundamental to Intelligent Transportation Systems (ITS). However, traffic dynamics exhibit inherent multi-scale heterogeneity, where stable global trends are often masked by stochastic local fluctuations. Existing methods struggle to reconcile these conflicting resolutions, leading to sub-optimal forecasting. To address this, we propose the Multi-Scale Spatial-Temporal Diffusion Transformer (MS-STDT). Deviating from previous diffusion approaches that treat generation as a monolithic task, we reframe the forward diffusion process as an intrinsic temporal coarse-graining operation. By leveraging the data's multi-scale hierarchy as a structural anchor, we introduce a coarse-to-fine latent guidance strategy that enables the model to reconstruct stable global trends before refining fine-grained details. This ensures physical consistency and generative stability without requiring external labels. Extensive experiments across six real-world datasets confirm that MS-STDT performs the best, demonstrating significant improvements in predictive accuracy and zero-shot robustness against sensor failures. We provide code and data at https://github.com/ZetaoLiPhD/MS-STDT.