Fractional diffusion bridge models
File(s)
Author(s)
Type
Conference Paper
Abstract
We present Fractional Diffusion Bridge Models (FDBM), a novel generative diffusion bridge framework driven by an approximation of the rich and non-Markovian fractional Brownian motion (fBM). Real stochastic processes exhibit a degree of memory effects (correlations in time), long-range dependencies, roughness and anomalous diffusion phenomena that are not captured in standard diffusion or bridge modeling due to the use of Brownian motion (BM). As a remedy, leveraging a recent Markovian approximation of fBM (MA-fBM), we construct FDBM that enable tractable inference while preserving the non-Markovian nature of fBM. We prove the existence of a coupling-preserving generative diffusion bridge and leverage it for future state prediction from paired training data. We then extend our formulation to the Schrödinger bridge problem and derive a principled loss function to learn the unpaired data translation. We evaluate FDBM on both tasks: predicting future protein conformations from aligned data, and unpaired image translation. In both settings, FDBM achieves superior performance compared to the Brownian baselines, yielding lower root mean squared deviation (RMSD) of Cₐ atomic positions in protein structure prediction and lower Fréchet Inception Distance (FID) in unpaired image translation.
Date Issued
2026-08-01
Date Acceptance
2025-11-30
Citation
Advances in Neural Information Processing Systems, 2026, 38, pp.126988-127038
ISBN
9798331338275
ISSN
1049-5258
Publisher
Neural Information Processing Systems Foundation, Inc. (NeurIPS)
Start Page
126988
End Page
127038
Journal / Book Title
Advances in Neural Information Processing Systems
Volume
38
Copyright Statement
© 2024 Neural Information Processing Systems Foundation, Inc. (NeurIPS).
Source
Advances in Neural Information Processing Systems 38
Publication Status
Published
Start Date
2025-11-30
Finish Date
2025-12-05
Coverage Spatial
Mexico City, Mexico
Date Publish Online
2026-08-01