TestifAI, a deep learning testing framework for efficient and accurate estimation of robustness against combinations of perturbations, is proposed and partial model tomography is introduced, a novel approach to reconstructing model behaviour in a multi-perturbation space from tests that apply only a small number of perturbations.
Abstract
As AI systems are increasingly deployed in safety-critical application domains (e.g., autonomous driving), associated risks increase too. Deep learning models underlying modern AI systems, therefore, must undergo thorough testing to ensure their correct behaviour. A single robustness test involves thousands of inferences to empirically verify if a model's outputs remain stable under a bounded perturbation of its inputs. However, existing testing frameworks lack the means to systematically explore and summarise robustness across a combinatorial space of perturbations. We propose TestifAI, a deep learning testing framework for efficient and accurate estimation of robustness against combinations of perturbations. TestifAI enables users to specify operational conditions as structured spaces of semantic input perturbations (e.g., image blur, brightness and zoom) and discrete severity levels (e.g., low, medium and high). Users can query model robustness for any combination (e.g.,"low blur, high brightness, and medium zoom"). To achieve efficiency and accuracy, TestifAI introduces partial model tomography, a novel approach to reconstructing model behaviour in a multi-perturbation space from tests that apply only a small number of perturbations (lower-order projections). To estimate robustness against at least three perturbations, TestifAI trains an auxiliary model on the results of tests involving up to two perturbations only, avoiding execution of an exponential number of tests. Our experiments on five image and language classification tasks show that TestifAI can predict higher-order (3 and 4 perturbations) test outcomes from low-order (1 and 2 perturbations) observations with an aggregate robustness estimation error of less than 7%, while reducing the number of inferences by 60-80%.
TESTNAV prioritises severe yet realistic failures by formulating robustness testing as bi-objective optimisation: maximise performance degradation while preserving input fidelity measured by modality-specific metrics.
Arooj Arif, T. Hartung, E. Botoeva et al.· 0 citations
The paradigm shift from testing for correct instruction execution to testing for robust behavior under open‑world conditions is analyzed and practical recommendations for integrating metamorphic testing and formal robustness verification into existing validation pipelines for mPNT systems are concluded.
O. Sieliukov, S. Ponomarenko· International Scientific Uni...· 0 citations
This work proposes a self-supervised masking scheme that simulates common sensor failures and explicitly trains the model to recover the original signal, and demonstrates that the resulting representations significantly improve the robustness of predictions to seen and unseen sensor failures on a vehicle dynamics dataset.
Jens U. Brandt, N. Pütz, M. Greiff et al.· Neural Information Processin...· 1 citation
A data maturity-deployment scale framework is introduced, arguing that engineering AI should be understood not as a search for a single best model, but as an evolving intelligence strategy that co-adapts with system maturity, operational context, and deployment scale.
Jay Lee, Hanqi Su, Ruoxin Wang et al.· Machine Learning: Engineerin...· 0 citations
Deep neural networks often suffer significant accuracy degradation when exposed to real-world image corruptions and distribution shifts. To overcome the limitations of fixed, input-agnostic test-time augmentation (TTA), an adaptive framework is proposed that learns per-sample transformations via reinforcement learning. Augmentation selection is cast as a Markov decision process and proximal policy optimization (PPO) agents are trained to choose sample-specific transforms under a composite reward combining classifier confidence gains with a self-consistency KL-divergence penalty on the model's own softmax outputs, thereby preserving overall belief stability. On clean CIFAR-10 (1 000 samples), the adaptive ensemble raises accuracy from 88.5% (baseline) and 87.3% (static TTA) to 90.0% (+1.5 pp). On CIFAR-10-C (15 corruptions × 5 severities; 1 000 images per condition), pooled top-1 accuracy improves from 75.7% (baseline) and 74.3% (static TTA) to 76.4% (+0.7 pp), and exceeds a TENT entropy-minimization baseline (75.9%) while operating in a strictly label-free regime that updates no model weights. Per-corruption gains are consistently positive across noise, blur, weather, and compression distortions, with the adaptive policy outperforming TENT on texture and compression corruptions where input-space transforms are most effective. These findings demonstrate that learned, per-sample augmentation policies improve robustness and reliability of deep vision models under diverse image conditions, against a strong baseline classifier.
Tushar Mittal, A. Dubey, Dharmender Saini et al.· Scientific Reports· 0 citations
Deep visual recognition has achieved remarkable success across various domains, including medical imaging, autonomous vehicles, and security systems. However, the black-box nature of deep learning models poses challenges in terms of transparency and trust, especially in critical applications where human understanding is essential. Explainable AI (XAI) seeks to address these concerns by providing human-interpretable explanations for model predictions. 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. We also discuss the evaluation of explainability through metrics like fidelity, consistency, and stability, and explore the challenges of balancing model performance with interpretability. Furthermore, we examine applications of XAI in medical imaging, autonomous driving, security and surveillance, agriculture, satellite imagery and remote sensing, industrial inspection, and visual forensics, highlighting how domain-specific data and operational constraints affect the required form and validation of explanations. Finally, we address current research gaps and propose future directions for enhancing the robustness and human–AI interaction in explainable visual recognition systems. As AI continues to be integrated into safety-critical domains, the development of explainable, transparent, and trustworthy AI systems will be crucial for their widespread adoption and ethical use.
