TL;DR
This study presents the first comprehensive evaluation framework systematically assessing XAI robustness under natural image corruptions encountered in production environments and establishes the first evidence-based XAI robustness ranking under natural corruptions, providing actionable guidance for practitioners selecting methods in real-world applications where input quality cannot be guaranteed.
This review explores the key techniques for explainability in deep visual recognition, including model-agnostic methods such as LIME and SHAP, model-specific approaches like saliency maps and feature visualization, and intrinsically interpretable models like decision trees and rule-based systems.
Model robustness against natural image corruptions is essential for safety-critical applications. While existing methods primarily focus on implicit representation learning, we provide the first systematic exploration of computational pathways to explicitly characterize internal robustness. We identify a progressive decay of robust features across network layers and establish a functional dependency between the prevalence of these features and model performance. To exploit these insights, we propose Suppress and Diversify (S\&D), a non-intrusive refinement approach that enhances robustness by dynamically selecting robust pathways and diversifying them through symmetry-preserving transformations. S\&D is architecture-agnostic, parameter-free, and incurs zero test-time overhead. Extensive evaluations across eight benchmarks demonstrate that S\&D consistently improves performance across multiple vision tasks, diverse backbones, and complex real-world scenarios, highlighting its broad efficacy and scalability.
Visual-language models (VLMs) frequently struggle with robustness issues in real-world situations due to low- or varying-quality input images. In this paper, we aim at analyzing VLMs'robustness by applying perturbations and distortions to the input images, such as blur or low contrast. Toward this goal, we propose BRUCE (Benchmarking Robustness Under Corruption Escalation), a multimodal reasoning fragility framework for scientific vision-language reasoning. State-of-the-art evaluation frameworks/studies primarily focus on clean-task accuracy and rarely analyze how reasoning stability degrades across robustness dimensions. Besides varying over a wide-range of input perturbations, BRUCE employs two novel metrics -- Robustness Corruption Index (RCI) and Traversal-RCI (T-RCI) -- to quantify how rapidly multimodal reasoning performance deteriorates in VLMs as visual corruption severity increases under progressive perturbation scaling. We evaluate BRUCE across chemistry and mathematical reasoning tasks for multiple datasets, while analyzing corruption-induced prediction failures in terms of four high-level reasoning domains: OCR-dependent reasoning, spatial reasoning, symbolic reasoning, and semantic failures, with each containing fine-grained corruption specific failure subtypes, thereby enabling an interpretable failure analysis.
Feature-robust Augmentation is introduced, which comprises diversified degradation-aware augmentation strategies, and a supervised contrastive learning pattern paired with a mean-teacher architecture that stabilizes features against augmentations through consistency constraints that wins the first place in ACM Multimedia 2026 Explainable Deepfake Detection Challenge.
It is concluded that explainability is a necessary, though not sufficient, condition for trustworthy Al, and concrete directions for future research are outlined, including standardised benchmarks, human-centred evaluation, and explainability for large generative models.
