mlsanityData to 14 Oct 2026

SHRUG-FM: Reliability-Aware Foundation Models for Earth Observation

arXiv 2511.10370 ↗First posted 13 Nov 2025cs.CV cs.AI cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

Geospatial foundation models (GFMs) for Earth observation often fail to perform reliably in environments underrepresented during pretraining. We introduce SHRUG-FM, a framework for reliability-aware prediction that enables GFMs to identify and abstain from likely failures. Our approach integrates three complementary signals: geophysical out-of-distribution (OOD) detection in the input space, OOD detection in the embedding space, and task-specific predictive uncertainty. We evaluate SHRUG-FM across three high-stakes rapid-mapping tasks: burn scar segmentation, flood mapping, and landslide detection. Our results show that SHRUG-FM consistently reduces prediction risk on retained samples, outperforming established single-signal baselines like predictive entropy. Crucially, by utilizing a shallow "glass-box" decision tree for signal fusion, SHRUG-FM provides interpretable abstention thresholds. It builds a pathway toward safer and more interpretable deployment of GFMs in climate-sensitive applications, bridging the gap between benchmark performance and real-world reliability.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 34 of its 53 references that are in the map. First posted on arXiv 13 Nov 2025.

Builds on

34 of 53 references.

References · 34

  1. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  2. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  3. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  4. 2017On Calibration of Modern Neural Networkscited by 4,960
  5. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,448
  6. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  7. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
  8. 2017What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?cited by 2,352
  9. 2014Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Imagescited by 776
  10. 2021A Survey of Uncertainty in Deep Neural Networkscited by 680
  11. 2022Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learningcited by 308
  12. 2023SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagerycited by 205
22 more
  1. 2023Foundation Models for Generalist Geospatial Artificial Intelligencecited by 149
  2. 2024Prithvi-EO-2.0: A Versatile Multi-Temporal Foundation Model for Earth Observation Applicationscited by 144
  3. 2025AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label datacited by 137
  4. 2018Bias-Reduced Uncertainty Estimation for Deep Neural Classifierscited by 113
  5. 2021Machine Learning with a Reject Option: A surveycited by 109
  6. 2025TerraMind: Large-Scale Generative Multimodality for Earth Observationcited by 104
  7. 2024PANGAEA: A Global and Inclusive Benchmark for Geospatial Foundation Modelscited by 68
  8. 2025TESSERA: Temporal Embeddings of Surface Spectra for Earth Representation and Analysiscited by 39
  9. 2025REOBench: Benchmarking Robustness of Earth Observation Foundation Modelscited by 15
  10. 2023EarthPT: a time series foundation model for Earth Observationcited by 14
  11. 2024Distribution Shifts at Scale: Out-of-distribution Detection in Earth Observationcited by 10
  12. 2023Analyzing Local Representations of Self-supervised Vision Transformerscited by 8
  13. 2025Spatially-Aware Evaluation of Segmentation Uncertaintycited by 6
  14. 2022Decision boundaries and convex hulls in the feature space that deep learning functions learn from imagescited by 5
  15. 2025How Does the Spatial Distribution of Pre-training Data Affect Geospatial Foundation Models?cited by 5
  16. 2025ExEBench: Benchmarking Foundation Models on Extreme Earth Eventscited by 5
  17. 2025Deep Ensembles Secretly Perform Empirical Bayescited by 4
  18. 2024NCDD: Nearest Centroid Distance Deficit for Out-Of-Distribution Detection in Gastrointestinal Visioncited by 3
  19. 2023An Ambiguity Measure for Recognizing the Unknowns in Deep Learningcited by 1
  20. 2024Segmentation Re-thinking Uncertainty Estimation Metrics for Semantic Segmentationcited by 1
  21. 2025CARE: Confidence-Aware Regression Estimation of building density fine-tuning EO Foundation Modelscited by 1
  22. 2025Uncertainty evaluation of segmentation models for Earth observationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Spatial Representation Learning Beyond Pixels: Unifying Raster Data and Vector Semantics for Human-Centric Geospatial Foundation Modelscited by 1
  2. 2026Beyond Accuracy: Assessing Calibration of Geospatial Foundation Models and Their Sensitivity to Distribution Shiftscited by 1
  3. 2026Transferability and operational reliability of a Prithvi crop classification foundation model under phenological and geographic shift across three continentscited by 0