A Unified Fractional Spectral Framework for Spatiotemporal Graph Signals: Bi-Fractional Transform and Geodesic Coupling
IEEE Transactions on Signal Processing · 2026 · 74 · 2069-2082
DOI: 10.1109/tsp.2026.3690762 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Signal Processing, 2026. Verfasst von Mingzhi Wang und 4 weiteren Autor:innen. Die Arbeit nennt 66 Literaturangaben. Gefördert von: Open Foundation of Hubei Key Laboratory of Applied Mathematics, Foundation of Key Laboratory of System Control and Information Processing, Ministry of Education, Startup Foundation for Introducing Talent of Nanjing Institute of Technology.
Abstract (Originalsprache)
Graph signal processing extends spectral analysis to data supported on irregular domains. Existing fractional transforms for two-dimensional graph signals, including the two-dimensional graph fractional Fourier transform (GFRFT), typically impose a shared fractional order across dimensions, which limits adaptivity to heterogeneous spatiotemporal spectra. To address this limitation, we propose the two-dimensional graph bi-fractional Fourier transform, which assigns independent fractional orders to the factor graphs of a Cartesian product, enabling decoupled spectral control while preserving separability, unitarity, and invertibility. To further resolve the basis ambiguity in temporal fractional analysis, we develop a geodesic-coupled GFRFT by constructing a coupling path along the principal geodesic on the unitary group, thereby unifying graph-induced and discrete temporal bases with guaranteed unitarity and a closed-form inverse. Building on these transforms, we derive a differentiable Wiener-type filtering framework with a hybrid optimization strategy: the fractional orders are learned end-to-end from data, while the coupling parameter is fixed as a structural regularizer. Experiments on real-world time-varying graph datasets and dynamic image restoration tasks demonstrate consistent gains over state-of-the-art fractional transforms and competitive learning-based baselines.
Förderung
- Open Foundation of Hubei Key Laboratory of Applied Mathematics
- Foundation of Key Laboratory of System Control and Information Processing, Ministry of Education
- Startup Foundation for Introducing Talent of Nanjing Institute of Technology
Zitieren
APA
Wang, M., Cui, M., Zhao, F., He, Y., & Zhang, Z. (2026). A Unified Fractional Spectral Framework for Spatiotemporal Graph Signals: Bi-Fractional Transform and Geodesic Coupling. IEEE Transactions on Signal Processing, 74, 2069-2082. https://doi.org/10.1109/tsp.2026.3690762
Vancouver
Wang M, Cui M, Zhao F, He Y, Zhang Z. A Unified Fractional Spectral Framework for Spatiotemporal Graph Signals: Bi-Fractional Transform and Geodesic Coupling. IEEE Transactions on Signal Processing. 2026;74:2069-2082. doi:10.1109/tsp.2026.3690762
BibTeX
@article{Wang2026,
author = {Wang, Mingzhi and Cui, Manjun and Zhao, Feiyue and He, Yangfan and Zhang, Zhichao},
title = {A Unified Fractional Spectral Framework for Spatiotemporal Graph Signals: Bi-Fractional Transform and Geodesic Coupling},
journal = {IEEE Transactions on Signal Processing},
year = {2026},
volume = {74},
pages = {2069--2082},
doi = {10.1109/tsp.2026.3690762}
}
EUVIMED