Salvato in:
Dettagli Bibliografici
Autori principali: Rey, Melanie, Mnih, Andriy, Neumann, Maxim, Overlan, Matt, Purves, Drew
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.19586
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915569986437120
author Rey, Melanie
Mnih, Andriy
Neumann, Maxim
Overlan, Matt
Purves, Drew
author_facet Rey, Melanie
Mnih, Andriy
Neumann, Maxim
Overlan, Matt
Purves, Drew
contents This paper investigates methods for estimating uncertainty in semantic segmentation predictions derived from satellite imagery. Estimating uncertainty for segmentation presents unique challenges compared to standard image classification, requiring scalable methods producing per-pixel estimates. While most research on this topic has focused on scene understanding or medical imaging, this work benchmarks existing methods specifically for remote sensing and Earth observation applications. Our evaluation focuses on the practical utility of uncertainty measures, testing their ability to identify prediction errors and noise-corrupted input image regions. Experiments are conducted on two remote sensing datasets, PASTIS and ForTy, selected for their differences in scale, geographic coverage, and label confidence. We perform an extensive evaluation featuring several models, such as Stochastic Segmentation Networks and ensembles, in combination with a number of neural architectures and uncertainty metrics. We make a number of practical recommendations based on our findings.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19586
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Uncertainty evaluation of segmentation models for Earth observation
Rey, Melanie
Mnih, Andriy
Neumann, Maxim
Overlan, Matt
Purves, Drew
Computer Vision and Pattern Recognition
Machine Learning
This paper investigates methods for estimating uncertainty in semantic segmentation predictions derived from satellite imagery. Estimating uncertainty for segmentation presents unique challenges compared to standard image classification, requiring scalable methods producing per-pixel estimates. While most research on this topic has focused on scene understanding or medical imaging, this work benchmarks existing methods specifically for remote sensing and Earth observation applications. Our evaluation focuses on the practical utility of uncertainty measures, testing their ability to identify prediction errors and noise-corrupted input image regions. Experiments are conducted on two remote sensing datasets, PASTIS and ForTy, selected for their differences in scale, geographic coverage, and label confidence. We perform an extensive evaluation featuring several models, such as Stochastic Segmentation Networks and ensembles, in combination with a number of neural architectures and uncertainty metrics. We make a number of practical recommendations based on our findings.
title Uncertainty evaluation of segmentation models for Earth observation
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2510.19586