Uncertainty-Calibrated Spatiotemporal Field Diffusion with Sparse Supervision

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Valencia, Kevin, Luo, Xihaier, Yoo, Shinjae, Park, David Keetae
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910041465946112
author Valencia, Kevin
Luo, Xihaier
Yoo, Shinjae
Park, David Keetae
author_facet Valencia, Kevin
Luo, Xihaier
Yoo, Shinjae
Park, David Keetae
contents Physical fields are typically observed only at sparse, time-varying sensor locations, making forecasting and reconstruction ill-posed and uncertainty-critical. We present SOLID, a mask-conditioned diffusion framework that learns spatiotemporal dynamics from sparse observations alone: training and evaluation use only observed target locations, requiring no dense fields and no pre-imputation. Unlike prior work that trains on dense reanalysis or simulations and only tests under sparsity, SOLID is trained end-to-end with sparse supervision only. SOLID conditions each denoising step on the measured values and their locations, and introduces a dual-masking objective that (i) emphasizes learning in unobserved void regions while (ii) upweights overlap pixels where inputs and targets provide the most reliable anchors. This strict sparse-conditioning pathway enables posterior sampling of full fields consistent with the measurements, achieving up to an order-of-magnitude improvement in probabilistic error and yielding calibrated uncertainty maps (\r{ho} > 0.7) under severe sparsity.
format Preprint
id arxiv_https___arxiv_org_abs_2603_04431
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Uncertainty-Calibrated Spatiotemporal Field Diffusion with Sparse Supervision
Valencia, Kevin
Luo, Xihaier
Yoo, Shinjae
Park, David Keetae
Machine Learning
Artificial Intelligence
Physical fields are typically observed only at sparse, time-varying sensor locations, making forecasting and reconstruction ill-posed and uncertainty-critical. We present SOLID, a mask-conditioned diffusion framework that learns spatiotemporal dynamics from sparse observations alone: training and evaluation use only observed target locations, requiring no dense fields and no pre-imputation. Unlike prior work that trains on dense reanalysis or simulations and only tests under sparsity, SOLID is trained end-to-end with sparse supervision only. SOLID conditions each denoising step on the measured values and their locations, and introduces a dual-masking objective that (i) emphasizes learning in unobserved void regions while (ii) upweights overlap pixels where inputs and targets provide the most reliable anchors. This strict sparse-conditioning pathway enables posterior sampling of full fields consistent with the measurements, achieving up to an order-of-magnitude improvement in probabilistic error and yielding calibrated uncertainty maps (\r{ho} > 0.7) under severe sparsity.
title Uncertainty-Calibrated Spatiotemporal Field Diffusion with Sparse Supervision
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2603.04431