DLLib DLLib is a modular toolkit for image-based deep learning that pairs baseline models with supervision, uncertainty, anomaly detection, and ensemble mechanisms to improve robustness, explainability, and reliability. Highlights - Uncertainty Inference: Aleatoric- and parallel-based transformations to produce diverse, semantically consistent predictions. - Supervision & Explainability: Lightweight checks to surface potential issues during model execution. - Anomaly Detection (VAEs): Input/output/activation-level outlier detection against the training distribution. - Ensembling: Combines model predictions with supervisor signals for stronger, more reliable outcomes. - Surrogate Model: A lightweight, feature-driven alternative to deep detectors. Components:SEMDRLIB: a dedicated DL library that generates diverse redundant versions of image-based DL models and applies several user-selected transformations in input images to perform multiple diverse inferences intended to provide semantically-identical, yet not bit-identical, results. DLETLIB: dedicated DL Explainable and Traceable library, incorporating a strongly structured and layered software architectural design that allows for the development of DL components following the requirements from functional safety standards like ISO 26262, ISO 21448 (SOTIF), IEC 61508, and others.