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

• URL: http://arxiv.org/abs/2511.10370v1
• Date: Fri, 14 Nov 2025 01:47:43 GMT
• Title: SHRUG-FM: Reliability-Aware Foundation Models for Earth Observation
• Abstract summary: Geospatial foundation models for Earth observation often fail to perform reliably in environments underrepresented during pretraining.<n>We introduce SHRUG-FM, a framework for reliability-aware prediction that integrates three complementary signals.<n>We show that OOD scores correlate with lower performance in specific environmental conditions, while uncertainty-based flags help discard many poorly performing predictions.
• Score: 2.6526177459233446
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Geospatial foundation models for Earth observation often fail to perform reliably in environments underrepresented during pretraining. We introduce SHRUG-FM, a framework for reliability-aware prediction that integrates three complementary signals: out-of-distribution (OOD) detection in the input space, OOD detection in the embedding space and task-specific predictive uncertainty. Applied to burn scar segmentation, SHRUG-FM shows that OOD scores correlate with lower performance in specific environmental conditions, while uncertainty-based flags help discard many poorly performing predictions. Linking these flags to land cover attributes from HydroATLAS shows that failures are not random but concentrated in certain geographies, such as low-elevation zones and large river areas, likely due to underrepresentation in pretraining data. SHRUG-FM provides a pathway toward safer and more interpretable deployment of GFMs in climate-sensitive applications, helping bridge the gap between benchmark performance and real-world reliability.

Related papers

• FMMO: Detecting the Divergence Between Local Attribution and Global Drift [2.007262412327553]
  Post-deployment drift poses a critical risk to algorithmic accountability.<n>We demonstrate that popular local attribution methods can exhibit misleading stability even as model reliability collapses.<n>We propose a Framework for Model Monitoring and Observability (FMMO) designed to expose the divergence between local explanation stability and global distribution shifts.
  arXiv  Detail & Related papers  (2026-09-05T16:25:59Z)
• Exploring Zero-Shot Foundation Models for Multivariate Time Series Anomaly Detection [4.468174618435543]
  MTSAD is essential for reliability and safety in domains such as industrial process monitoring and financial risk management.<n>Foundation Models (FMs) are pre-trained on broad data with strong zero-shot generalization.<n>We investigate the zero-shot application of a FM to industrial MTSAD on the Secure Water Treatment (SWaT) benchmark.
  arXiv  Detail & Related papers  (2026-07-14T07:32:08Z)
• Scalable and Trustworthy Earth Observation Foundation Models [0.0]
  Foundation models (FMs) have transformed machine learning from isolated task-specific models to general-purpose models pretrained on broad data.<n>This chapter defines the FMs paradigm in remote sensing (RS), then synthesizes the current model landscape.<n>The chapter also incorporates recent benchmark evidence showing that no single geospatial foundation model is universally best.
  arXiv  Detail & Related papers  (2026-07-08T14:31:02Z)
• From Signals to Transfer: A Factorised Study of Probe-Based Uncertainty Estimation in Large Language Models [51.470637128942336]
  We propose a factorised study of probe-based uncertainty estimation (UE) under matched conditions.<n>Our results show that raw hidden states and attention features are difficult to outperform in-domain.<n>We train benchmark-based pretrained probes that transfer reasonably well to open-ended factual generation.
  arXiv  Detail & Related papers  (2026-06-26T03:23:15Z)
• Latent Dynamics-Aware OOD Monitoring for Trajectory Prediction with Provable Guarantees [8.920589816043298]
  In safety-critical Cyber-Physical Systems, accurate trajectory prediction provides vital guidance for downstream planning and control.<n> detecting such OOD events is challenging due to evolving traffic conditions and changing interaction patterns.<n>We extend the cumulative Maximum Discrepancy approach to enable detection without requiring explicit knowledge of the post-change distribution.<n>Experiments on three real-world driving datasets demonstrate reduced detection delay and robustness to heavy-tailed errors and unknown post-change conditions.
  arXiv  Detail & Related papers  (2026-03-15T20:56:57Z)
• RaUF: Learning the Spatial Uncertainty Field of Radar [2.6853734738584047]
  Millimeter-wave radar offers unique advantages in adverse weather but suffers from low spatial fidelity, severe azimuth ambiguity, and clutter-induced spurious returns.<n>We propose RaUF, a spatial uncertainty field learning framework that models radar measurements through their physically grounded anisotropic properties.<n>To resolve conflicting feature-to-label mapping, we design an anisotropic probabilistic model that learns fine-grained uncertainty.
  arXiv  Detail & Related papers  (2026-03-01T10:10:43Z)
• Learning to Explore: Policy-Guided Outlier Synthesis for Graph Out-of-Distribution Detection [51.93878677594561]
  In unsupervised graph-level OOD detection, models are typically trained using only in-distribution (ID) data.<n>We propose a Policy-Guided Outlier Synthesis framework that replaces statics with a learned exploration strategy.
  arXiv  Detail & Related papers  (2026-02-28T11:40:18Z)
• Forecasting Fails: Unveiling Evasion Attacks in Weather Prediction Models [60.728124907335]
  This work introduces Weather Adaptive Adversarial Perturbation Optimization (WAAPO), a novel framework for generating targeted adversarial perturbations.<n>WAAPO achieves this by incorporating constraints for channel sparsity, spatial localization, and smoothness, ensuring that perturbations remain physically realistic and imperceptible.<n>Our experiments highlight critical vulnerabilities in AI-driven forecasting models, where small perturbations to initial conditions can result in significant deviations.
  arXiv  Detail & Related papers  (2025-12-09T17:20:56Z)
• Synergistic Neural Forecasting of Air Pollution with Stochastic Sampling [50.3911487821783]
  Air pollution remains a leading global health and environmental risk, particularly in regions vulnerable to episodic air pollution spikes due to wildfires, urban haze and dust storms.<n>Here, we present SynCast, a high-resolution neural forecasting model that integrates meteorological and air composition data to improve predictions of both average and extreme pollution levels.
  arXiv  Detail & Related papers  (2025-10-28T01:18:00Z)
• Revisiting Multivariate Time Series Forecasting with Missing Values [65.30332997607141]
  Missing values are common in real-world time series.<n>Current approaches have developed an imputation-then-prediction framework that uses imputation modules to fill in missing values, followed by forecasting on the imputed data.<n>This framework overlooks a critical issue: there is no ground truth for the missing values, making the imputation process susceptible to errors that can degrade prediction accuracy.<n>We introduce Consistency-Regularized Information Bottleneck (CRIB), a novel framework built on the Information Bottleneck principle.
  arXiv  Detail & Related papers  (2025-09-27T20:57:48Z)
• Confidence-Filtered Relevance (CFR): An Interpretable and Uncertainty-Aware Machine Learning Framework for Naturalness Assessment in Satellite Imagery [3.846084066763095]
  Confidence-Filtered Relevance (CFR) is a data-centric framework that combines LRP Attention Rollout with Deep Deterministic Uncertainty estimation.<n>CFR partitions the dataset into subsets based on uncertainty thresholds, enabling systematic analysis of how uncertainty shapes explanations of naturalness in satellite imagery.<n>As uncertainty increases, the interpretability of relevance heatmaps declines and their entropy grows, indicating less selective and more ambiguous attributions.
  arXiv  Detail & Related papers  (2025-07-17T12:06:08Z)
• Reservoir-enhanced Segment Anything Model for Subsurface Diagnosis [24.223897705640105]
  Urban roads and infrastructure, vital to city operations, face growing threats from subsurface anomalies like cracks and cavities.<n>Ground Penetrating Radar (GPR) effectively visualizes underground conditions employing electromagnetic (EM) waves.<n>However, accurate anomaly detection via GPR remains challenging due to limited labeled data, varying subsurface conditions, and indistinct target boundaries.<n>We propose the Reservoir-enhanced Segment Anything Model (Res-SAM) to exploit both visual discernibility and wave-changing properties of GPR data.
  arXiv  Detail & Related papers  (2025-04-26T05:13:22Z)
• FOCUS on Contamination: A Geospatial Deep Learning Framework with a Noise-Aware Loss for Surface Water PFAS Prediction [3.499081366421979]
  Per- and polyfluoroalkyl substances (PFAS), chemicals found in products like non-stick cookware, are persistent environmental pollutants with severe health risks.<n>We introduce FOCUS, a deep learning framework with a label noise-aware loss function, to predict PFAS contamination in surface water over large regions.
  arXiv  Detail & Related papers  (2025-02-17T16:57:10Z)
• Adaptive Residual Transformation for Enhanced Feature-Based OOD Detection in SAR Imagery [5.63530048112308]
  The presence of unknown targets in real battlefield scenarios is unavoidable. Various feature-based out-of-distribution approaches have been developed to address this issue. We propose transforming feature-based OOD detection into a class-localized feature-residual-based approach.
  arXiv  Detail & Related papers  (2024-11-01T00:09:02Z)
• Energy-based Out-of-Distribution Detection for Graph Neural Networks [76.0242218180483]
  We propose a simple, powerful and efficient OOD detection model for GNN-based learning on graphs, which we call GNNSafe. GNNSafe achieves up to $17.0%$ AUROC improvement over state-of-the-arts and it could serve as simple yet strong baselines in such an under-developed area.
  arXiv  Detail & Related papers  (2023-02-06T16:38:43Z)
• Free Lunch for Generating Effective Outlier Supervision [46.37464572099351]
  We propose an ultra-effective method to generate near-realistic outlier supervision. Our proposed textttBayesAug significantly reduces the false positive rate over 12.50% compared with the previous schemes.
  arXiv  Detail & Related papers  (2023-01-17T01:46:45Z)
• Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines [65.0803400763215]
  This work critically examines how adversarial robustness guarantees change when state-of-the-art certifiably robust models encounter out-of-distribution data. We propose a novel data augmentation scheme, FourierMix, that produces augmentations to improve the spectral coverage of the training data. We find that FourierMix augmentations help eliminate the spectral bias of certifiably robust models enabling them to achieve significantly better robustness guarantees on a range of OOD benchmarks.
  arXiv  Detail & Related papers  (2021-12-01T17:11:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.