Repository logo
Log in(current)
Repository logoMIT Open ScholarshipDSpace@MIT
  1. Home
  2. MIT Open Access Articles
  3. MIT Open Access Articles
  4. Leveraging the Spatial Hierarchy: Coarse-to-fine Trajectory Generation via Cascaded Hybrid Diffusion

Leveraging the Spatial Hierarchy: Coarse-to-fine Trajectory Generation via Cascaded Hybrid Diffusion

Thumbnail Image
Download
Name

3770854.3780191.pdf

Size

6.87 MB

Format

Adobe PDF

Checksum (MD5)

711a78f9a1dd13b9aa2268697849887c

Download all files submitted through automated deposit
3770854.3780191.zip (6.39 MB)
Author(s)
Guo, Baoshen
•
Hong, Zhiqing
•
Li, Junyi
•
Wang, Shenhao
•
Zhao, Jinhua
Date Issued
August 9, 2026
Publisher
Association for Computing Machinery
Citation
Baoshen Guo, Zhiqing Hong, Junyi Li, Shenhao Wang, and Jinhua Zhao. 2026. Leveraging the Spatial Hierarchy: Coarse-to-fine Trajectory Generation via Cascaded Hybrid Diffusion. In Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1 (KDD ’26), August 09–13, 2026, Jeju Island, Republic of Korea. ACM, New York, NY, USA, 12 pages.
Version
Final published version
Abstract
Urban mobility data has significant connections with economic growth and plays an essential role in various smart-city applications. Due to privacy concerns and substantial data collection costs, fine-grained human mobility trajectories are challenging to make publicly available on a large scale. A promising solution to address this issue is trajectory synthesizing, which generates synthetic trajectories that preserve aggregate spatiotemporal distributions. However, existing works often neglect the road network structural constraints or rely on instance-level external supervision, thus limiting their scalability in generating fine-grained and high-fidelity trajectories. In this paper, we propose Cardiff, a coarse-to-fine Cascaded hybrid diff usion-based framework for fine-grained and structure-plausible trajectory generation. By leveraging the hierarchical nature of urban mobility, Cardiff decomposes the generation process into two cascaded levels, i.e., discrete road segment-level and continuous fine-grained GPS-level: (i) At the segment level, to reduce computational costs and redundancy in raw trajectories, we first encode the discrete road segments into low-dimensional latent embeddings and design a diffusion transformer-based latent denoising network for segment-level synthesis. (ii) Taking the first stage of generation as conditions, we then design a fine-grained GPS-level conditional denoising network with a noise augmentation mechanism to achieve road-network-constrained and finegrained generation. The cascaded progressive generation yields high-fidelity fine-grained trajectories while adhering to road geometry and topology constraints. Experimental results on three large real-world trajectory datasets demonstrate that our method outperforms state-of-the-art baselines in various metrics. The code is available at https://github.com/urban-mobility-generation/Cardiff.
Description
KDD ’26, Jeju Island, Republic of Korea
Terms of Use
Creative Commons Attribution
https://creativecommons.org/licenses/by/4.0/
Persistent DSpace Link
https://dspace.mit.edu/handle/1721.1/171087
DOI of Published Version
https://doi.org/10.1145/3770854.3780191
Repository logo
PrivacyPermissionsAccessibilityContact us
Repository logo
Notify us about copyright concerns.