mlsanityData to 14 Oct 2026

Pattern-Guided Diffusion Models

arXiv 2512.13945 ↗First posted 15 Dec 2025cs.LGcited by 0

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

Diffusion models have shown promise in forecasting future data from multivariate time series. However, few existing methods account for recurring structures, or patterns, that appear within the data. We present Pattern-Guided Diffusion Models (PGDM), which leverage inherent patterns within temporal data for forecasting future time steps. PGDM first extracts patterns using archetypal analysis and estimates the most likely next pattern in the sequence. By guiding predictions with this pattern estimate, PGDM makes more realistic predictions that fit within the set of known patterns. We additionally introduce a novel uncertainty quantification technique based on archetypal analysis, and we dynamically scale the guidance level based on the pattern estimate uncertainty. We apply our method to two well-motivated forecasting applications, predicting visual field measurements and motion capture frames. On both, we show that pattern guidance improves PGDM's performance (MAE / CRPS) by up to 40.67% / 56.26% and 14.12% / 14.10%, respectively. PGDM also outperforms baselines by up to 65.58% / 84.83% and 93.64% / 92.55%.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 18 of its 27 references that are in the map. First posted on arXiv 15 Dec 2025.

Builds on

18 of 27 references.

Closest ancestors · 6

Same direction, newest first

  1. 2024Auto-Regressive Moving Diffusion Models for Time Series Forecasting12 Dec
  2. 2024A Survey on Diffusion Models for Time Series and Spatio-Temporal Data29 Apr
  3. 2023Towards Unifying Diffusion Models for Probabilistic Spatio-Temporal Graph Learning26 Oct
  4. 2023DiffSTG: Probabilistic Spatio-Temporal Graph Forecasting with Denoising Diffusion Models31 Jan
  5. 2021CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation7 Jul
  6. 2021Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting28 Jan

Other references · 12

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  3. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  4. 2022Classifier-Free Diffusion Guidancecited by 5,847
  5. 2021AI Choreographer: Music Conditioned 3D Dance Generation with AIST++cited by 485
  6. 2022Stochastic Trajectory Prediction via Motion Indeterminacy Diffusioncited by 185
  7. 2023Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglementcited by 90
  8. 2022Your ViT is Secretly a Hybrid Discriminative-Generative Diffusion Modelcited by 58
  9. 20223D Human Motion Prediction: A Surveycited by 30
  10. 2023A Transformer-based Diffusion Probabilistic Model for Heart Rate and Blood Pressure Forecasting in Intensive Care Unitcited by 14
  11. 2024Diffusion-Based Environment-Aware Trajectory Predictioncited by 14
  12. 2024EgoNav: Egocentric Scene-aware Human Trajectory Predictioncited by 9

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Diffusion Models for Time Series Forecasting: A Survey
  2. 2023Diffusion Models for Time Series Applications: A Survey
  3. 2024The Rise of Diffusion Models in Time-Series Forecasting
  4. 2026Differencing the Diffusion Trajectory toward Uncertain Components for Time Series Forecasting
  5. 2024Stochastic Diffusion: A Diffusion Probabilistic Model for Stochastic Time Series Forecasting
  6. 2026GARDiff: Graph-Aligned Residual Diffusion for Probabilistic Multivariate Time-Series Forecasting
  7. 2026DiTS: Multimodal Diffusion Transformers Are Time Series Forecasters
  8. 2026DynG-Diff: A State-Aware Dynamic Guidance Diffusion Framework for Probabilistic Time Series Forecasting