Abstract Hybrid Quantum-Classical (HQC) models have been shown to offer a promising strategy for leveraging near-term quantum computing resources for machine learning tasks. While initial studies have shown promise on small datasets, it remains uncertain how HQC models would scale and what their architectures would look like for a more complex problem, such as CIFAR-10. This paper poses a fundamental question with regard to the emplacement of a Parameterized Quantum Circuit (PQC) in a deep learning architecture to ensure most effectiveness in a deep learning architecture. In this experiment, a comprehensive evaluation of HQC models using four different architectures has been done, where a PQC has been used either as a front-end feature extractor or a back-end classifier. Our findings show a striking point where models using a PQC as a classifier always outperform models using a PQC as a feature extractor, which are computationally expensive and create information bottlenecks. Notably, our proposed HQC-ViT model, which combines a classical ViT with a quantum classifier, outperforms a purely classical ViT model, proposing a valuable architectural blueprint for hybrid models under NISQ constraints. Ultimately, optimizing PQC placement in vision models provides important architectural guidelines for deploying resource-efficient, quantum-assisted diagnostic tools in future digital health and medical imaging applications.
Abstract Hybrid Quantum-Classical (HQC) models have been shown to offer a promising strategy for leveraging near-term quantum computing resources for machine learning tasks. While initial studies have shown promise on small datasets, it remains uncertain how HQC models would scale and what their architectures would look like for a more complex problem, such as CIFAR-10. This paper poses a fundamental question with regard to the emplacement of a Parameterized Quantum Circuit (PQC) in a deep learning architecture to ensure most effectiveness in a deep learning architecture. In this experiment, a comprehensive evaluation of HQC models using four different architectures has been done, where a PQC has been used either as a front-end feature extractor or a back-end classifier. Our findings show a striking point where models using a PQC as a classifier always outperform models using a PQC as a feature extractor, which are computationally expensive and create information bottlenecks. Notably, our proposed HQC-ViT model, which combines a classical ViT with a quantum classifier, outperforms a purely classical ViT model, proposing a valuable architectural blueprint for hybrid models under NISQ constraints. Ultimately, optimizing PQC placement in vision models provides important architectural guidelines for deploying resource-efficient, quantum-assisted diagnostic tools in future digital health and medical imaging applications.
Hybrid Quantum-Classical Neural Networks (HQC-NNs) integrate Parameterized Quantum Circuits (PQCs) with classical deep learning. We systematically compare Early Fusion (quantum preprocessing) versus Late Fusion (quantum classification) across nine classification tasks. Using 6-qubit circuits evaluated over 30 independent trials per task with Bonferroni-corrected statistical validation, Early Fusion achieves superior performance on 6/9 tasks, significantly outperforming Late Fusion on problems requiring global feature correlations: Parity Problem (91.5% vs 83.1%), Entangled XOR (89.3% vs 83.5%), and multiclass digit recognition (92.9% vs 88.1%). Late Fusion demonstrates nominal advantages on structured visual data with geometric patterns (Fashion-MNIST: 92.0% vs 90.0%). We characterize performance differences through an information-theoretic embedding bottleneck framework. Gradient flow analysis reveals 2.1× faster convergence (in terms of training epochs) for Early Fusion on complex tasks, while noise robustness analysis indicates advantages persist under representative NISQ noise levels (p ≤ 0.02). The observed performance gains are attributed to architectural inductive bias differences rather than provable quantum computational speedups, providing principled, empirically validated task-dependent architectural guidelines for quantum machine learning practitioners.
The objective of this research is to develop a Hybrid Quantum-Classical Convolutional Neural Network (QC-CNN) framework that integrates parameterized quantum circuits (PQCs) within classical CNN architectures to enhance image classification performance.The proposed model leverages quantum principles such as superposition and entanglement for high-dimensional feature representation, achieving superior accuracy and reduced computational complexity.Experiments on MNIST and CIFAR-10 datasets demonstrated improved classification accuracy of 98.7% and 82.5%, respectively, surpassing traditional CNNs while reducing training time by 33% and parameters by 45%.Statistical analysis confirmed the significance of these improvements.Visualizations using t-SNE revealed enhanced class separability, and noise perturbation tests validated the model's robustness.The results highlight the hybrid QC-CNN's potential for efficient and scalable quantum-enhanced deep learning applications.Extended evaluations, including quantum-layer-depth analysis, robustness testing, and t-SNE visualization, empirically support the hybrid QC-CNN's effectiveness across varied scenarios.
The integration of quantum computing principles with classical deep learning architectures has emerged as a promising direction for advancing image recognition systems beyond the limitations of purely classical convolutional neural networks (CNNs). This study proposes and analyzes a hybrid quantum–classical deep learning framework based on quanvolutional neural network (QNN) principles, where quantum circuits are embedded within convolutional feature extraction layers to enhance representation capability. The research investigates how quantum-enhanced filters, particularly quanvolutional layers, contribute to improved feature diversity, non-linear transformation capacity, and classification accuracy in image recognition tasks. Building upon established convolutional architectures and quantum machine learning theories, the proposed framework integrates quantum random circuit encoding with classical deep neural networks to evaluate performance gains in comparison to conventional CNN models. The study further incorporates theoretical insights from hybrid quantum-classical learning systems, emphasizing the role of parameterized quantum circuits in feature mapping and optimization stability. The methodology is grounded in NISQ-era quantum computing constraints and leverages hybrid optimization strategies to ensure computational feasibility. Experimental evaluation demonstrates that quanvolution-based models can provide improved robustness in feature extraction, particularly in high-dimensional image spaces, while maintaining competitive computational efficiency. However, scalability limitations and noise sensitivity in quantum circuits remain key challenges. Overall, this work contributes to the growing field of quantum-enhanced machine learning by providing a structured performance analysis of hybrid quanvolutional architectures and highlighting their potential role in next-generation intelligent visual recognition systems.
The notion of the expressivity of hybrid models is often confused with a heavy reliance on pre-trained classical backbones even in regard to the development of Quantum Machine Learning (QML) as an aspect of the Noisy Intermediate-Scale Quantum (NISQ) regime. As shown in this paper, an end-to-end Hybrid Convolutional-Quantum Neural Network (CNN-QNN) can be trained on a complex dataset (CIFAR-10) in full form via a strict iso-parameter analysis. By inhibiting a fully classical control model and the proposed hybrid architecture to a very small scale of slightly less than 357,000 parameters we can isolate the architectural effect of an 8-qubit Variational Quantum Circuit (VQC) which is the network bottleneck. Empirical data suggests that the suggested hybrid architecture has exceptionally small model footprint and has a competitive image classification performance. The hybrid CNN-QNN reaches the highest testing accuracy of CIFAR-10 of 84.06% with a stringent parameter requirements demonstrating the validity of quantum circuits constructed on a small scale. It is completely end-to-end trained and this does not utilize any large pre-trained classical networks, which further demonstrates that our system can be easily explained by Explainable AI (XAI) tools, which can be used to understand the manner in which this classical-to-quantum transition of feature representations occurs. Further, we package the trained weights of the optimum model as a tiny Graphical User Interface (GUI) which can execute run time inference on the standard CPU machines, which shows the feasibility of NISQ-era deployability. The work offers a deployable and interpretable reference on hybrid quantum vision models, that are parameter efficient as well as not being trained using transfer learning.
This paper reviews hybrid quantum–classical (HQC) architectures for COVID-19-related respiratory medical-image analysis. To address the heterogeneity of existing studies, we propose an architecture-centric taxonomy based on the functional role and placement of the quantum module. Reviewed models are grouped into three archetypes: Archetype A, where quantum circuits act as patch-level quanvolutional preprocessors; Archetype B, where classical feature extractors are coupled with quantum classifier heads; and Archetype C, where quantum circuits generate intermediate features for downstream classical classifiers. Ten peer-reviewed journal studies were selected through a PRISMA-inspired search and analyzed across architecture, diagnostic performance, quantum resource reporting, validation rigor, computational scalability, and deployment feasibility. The review shows that HQC models often report promising binary COVID-19 screening results on CT or chest X-ray images, but multiclass respiratory classification remains less stable. Key limitations include simulator-dominated evaluation, limited external validation, unclear patient-wise splitting, incomplete reporting of qubit counts, circuit depth, and shots, and insufficient comparison with strong classical baselines. Overall, current HQC models should be viewed as exploratory quantum-augmented classical pipelines rather than clinically validated diagnostic systems. No conclusive task-level quantum advantage has yet been demonstrated for COVID-19 medical imaging. Future progress requires standardized benchmarking, transparent quantum-resource reporting, patient-wise and multi-center validation, hardware-aware evaluation, and interpretable hybrid designs compatible with NISQ-era constraints.
Hybrid quantum classical neural networks integrate parameterized quantum circuits (PQCs) with established deep learning architectures, but their performance depends strongly on the choice of quantum circuit architecture, a choice that remains largely manual. Most existing approaches rely on hand-designed or fixed circuit ansätze, requiring circuit structure, gate composition, and qubit connectivity to be specified in advance with no guarantee that they suit the task. This limitation is especially acute in image classification, where quantum circuits must transform features extracted by classical networks while remaining compact enough for practical training, requirements that generic, task-agnostic ansätze are unlikely to satisfy simultaneously. We extend EXAQC, an evolutionary framework for automated quantum circuit discovery, to image classification. EXAQC evolves PQCs as intermediate processing modules while retaining classical feature-extraction and prediction layers. On MNIST, Fashion-MNIST, and CIFAR-10, EXAQC achieves 98.42%, 90.62%, and 85.47% accuracy, respectively, while using comparable gate counts to other quantum architecture-search methods. Against classical networks, evolved hybrid models maintain comparable accuracy with substantially fewer trainable parameters, reaching 85.68% on CIFAR-10 with over 25× fewer parameters than a 10-layer CNN. Encoding choice also matters: rotation-based encodings (RX, RY, U3) outperform amplitude encoding by 22-25 points on CIFAR-10. These results demonstrate that automated circuit discovery yields compact quantum modules that can replace larger classical components in vision architectures while retaining competitive accuracy.