TrustworthyQENN: A Quantum Evidential Neural Network Based on Complex-Valued Contrastive Learning for Uncertainty Pattern Classification

Out-of-distribution (OOD) detection requires accurately classifying in-distribution (ID) samples while effectively distinguishing anomalous OOD data. However, existing methodologies predominantly rely on real-valued magnitude features, neglecting the semantic richness embedded in phase information, and often lack a systematic theoretical framework for quantitively modeling uncertainty. To address this dual limitation of incomplete feature representation and insufficient uncertainty modeling, the trustworthy quantum evidence neural network (TrustworthyQENN) is proposed, a novel quantum-inspired framework bridging complex-valued representation learning with generalized quantum evidence theory (GQET). Specifically, supervised complex-valued contrastive learning (SCVCL) is proposed to synchronize amplitude distributions with phase correlations, thereby enforcing high intra-class compactness and inter-class separability for ID data. A quantum evidence generation mechanism based on GQET is subsequently devised, where the OOD state is formally grounded as the quantum empty set within a Hilbert space. Furthermore, the generalized quantum evidential combination rule (GQECR) is leveraged to fuse multi-view evidence, thereby achieving trustworthy inference. Extensive experiments on the MSTAR, EuroSAT, and FUSAR-Ship benchmarks substantiate the superiority of TrustworthyQENN, which achieves a peak AUROC of 95.94\% on the MSTAR dataset while consistently outperforming state-of-the-art methods across all evaluated scenarios.

Artificial intelligence systems are excellent at recognizing patterns they have been trained on, but they often struggle when encountering completely new or unexpected situations in the real world. Safely identifying these unfamiliar inputs—known as out-of-distribution data—is critical for the reliable deployment of AI. However, existing methods often miss subtle clues because they only look at basic data features, and they lack a rigorous mathematical way to express uncertainty. To solve this, we introduce the trustworthy quantum evidence neural network (TrustworthyQENN). Our approach improves AI reliability in two main ways. First, it analyzes complex-valued data, meaning it looks at both the size and the underlying rhythm (or phase) of a signal. This provides a much richer picture, which is especially important for things like radar and satellite imagery. Second, we use a concept inspired by quantum physics to explicitly teach the AI how to categorize an object as "unknown". By combining this deeper data analysis with a better system for handling uncertainty, our framework acts like a team of experts collaborating to make a decision. In tests using various radar and satellite images, TrustworthyQENN significantly outperformed current methods, proving it can confidently identify familiar objects while safely raising a red flag when it encounters something new. Ultimately, this helps pave the way for safer, more robust autonomous systems in unpredictable environments.