Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Alexandria Engineering Journal

Hybrid quantum transfer learning for supervised UAV aerial image classification

S. Abdel‐Khalek, Mariam Algarni, Suliman A. Alsuhibany, Mahmoud Abdel‐Aty

Published May 1, 2026• 2 citations
Publisher

Abstract

Unmanned Aerial Vehicles (UAVs) enable high-resolution remote-sensing for land-cover and object mapping, yet conventional deep learning (DL) models often remain sensitive to scale, viewpoint, and illumination variability in UAV imagery. This paper presents QTL-SAIC, a quantum transfer learning framework for supervised aerial image classification that integrates classical feature learning with quantum-enhanced decision making. A pre-trained Capsule Network (CapsNet) extracts compact, discriminative features, which are classified using a variational quantum neural network (QNN) within a hybrid quantum–classical pipeline. The quantum module is fully specified (qubit number, encoding, circuit architecture, measurements, optimization, and execution backend with shot settings) to support reproducibility. Hyperparameters are optimized via the Gannet Optimization Algorithm (GOA), and evaluation follows stratified splits, repeated multi-seed trials and statistical reporting, with class imbalance addressed through cost-sensitive and/or balanced sampling strategies. Results indicate that QTL-SAIC yields competitive and stable performance relative to resource-matched classical baselines. Although no formal quantum advantage is claimed, ablations on circuit depth and noise sensitivity delineate the conditions under which the quantum component contributes to predictive performance.

DOI
10.1016/j.aej.2026.04.034

Related papers

Towards Fair Benchmarking of Quantum Transfer Learning for Visual Classification

Nouhaila Innan, Saim Rehman, Muhammad Shafique

Quantum Transfer Learning (QTL) offers a promising approach for visual quantum machine learning under near-term constraints, where limited qubit counts, shallow circuit depths, and costly hybrid optimization restrict end-to-end quantum training. In this setting, pretrained classical backbones can extract high-level visual features, while compact quantum modules operate as trainable classification heads. However, existing QTL results are difficult to compare because they often differ in datasets, preprocessing, backbone settings, qubit budgets, circuit designs, optimization choices, and reporting protocols. This work presents a controlled benchmarking methodology for evaluating representative QTL methods under a unified transfer-learning pipeline. The benchmark compares DQN-QTL, QPIE-QTL, AE-CQTL, PVCQTL, and ED-QTL under shared preprocessing rules, frozen-backbone settings, training conditions, and reporting metrics. The evaluation focuses on Fashion-MNIST and Hymenoptera Ants vs. Bees as the two main datasets, while CIFAR-10 is used to provide additional configuration-level evidence on a harder natural-image task. Beyond predictive performance, the benchmark analyzes circuit size, trainable parameters, quantum parameters, training time, and architectural sensitivity to qubit count and circuit depth. The results show that no single QTL family dominates across all settings: performance depends on the dataset, encoding strategy, circuit design, and computational cost. These findings highlight the need for resource-aware QTL evaluation and provide guidance for selecting hybrid quantum-classical transfer models under near-term resource constraints.

quant-pharXiv:2605.19417•May 20, 2026•2 citations

Quantum Transfer Learning

Hamid D. Ismail

This chapter presents quantum transfer learning (QTL) as a practical and conceptually grounded framework for integrating quantum computation into modern machine learning. Building on the classical transfer learning paradigm, this chapter shows how pretrained classical models can be reused for large-scale representation learning while variational quantum circuits (VQCs) are deployed as compact downstream adaptation modules. Emphasis is placed on hybrid classical-quantum architectures in which quantum feature extractors operate on low-dimensional embeddings produced by convolutional networks, transformers, or graph neural networks, and their measurement-based outputs are processed by classical heads for prediction. This chapter develops the mathematical and architectural foundations of this approach, explains the role of variational quantum feature maps, and examines why quantum feature extractors may be useful in data-limited and structure-rich settings. Applications to text, image, and molecular data are discussed to illustrate the domain-independent nature of the framework. Practical implementation is addressed through PennyLane and PyTorch integration, highlighting differentiable quantum layers and end-to-end hybrid optimization. A hands-on example using encoded embeddings and VQCs demonstrates methodology, benchmarking, and model evaluation in a realistic supervised-learning setting. This chapter concludes with a critical discussion of limitations, including task dependence, noise, measurement cost, benchmarking challenges, and the continued dominance of large pretrained classical encoders.

CRC Press eBooks•Nov 4, 2026•0 citations

Quantum Machine Learning for UAV Swarm Intrusion Detection

Kuan‐Cheng Chen, Samuel Yen-Chi Chen, Tai-Yue Li +2 more

Intrusion detection in unmanned-aerial-vehicle (UAV) swarms is complicated by high mobility, non-stationary traffic, and severe class imbalance. Leveraging a 120 k-flow simulation corpus that covers five attack types, we benchmark three quantum-machine-learning (QML) approaches - quantum kernels, variational quantum neural networks (QNNs), and hybrid quantum-trained neural networks (QT-NNs) - against strong classical baselines. All models consume an 8-feature flow representation and are evaluated under identical preprocessing, balancing, and noise-model assumptions. We analyse the influence of encoding strategy, circuit depth, qubit count, and shot noise, reporting accuracy, macro-F1, ROC-AUC, Matthews correlation, and quantum-resource footprints. Results reveal clear trade-offs: quantum kernels and QT-NNs excel in low-data, nonlinear regimes, while deeper QNNs suffer from trainability issues, and CNNs dominate when abundant data offset their larger parameter count. The complete codebase and dataset partitions are publicly released to enable reproducible QML research in network security.

arXiv:2509.01812•IEEE Globecom Workshops (GC Wkshps)•Dec 8, 2025•8 citations

Post-variational classical quantum transfer learning for binary classification

Kavitha Yogaraj, Brian Quanz, Tarun Vikas +2 more

We address the limitations of variational quantum circuits (VQCs) in hybrid classical-quantum transfer learning by introducing post-variational strategies, which reduce training overhead and mitigate optimization issues. Our approach Post Variational Classical Quantum Transfer Learning (PVCQTL) includes three designs: (1) modified observable construction, (2) a hybrid approach, and (3) a variational-post-variational combination. We evaluate these on pre-trained models (VGG19, ResNet50, ResNet18, MobileNet) for 4 and 8 qubits, with ResNet50 performing best in deepfake detection. Compared to classical models (MLP, ResNet50) and quantum baselines hybrid quantum classical neural network (HQCNN), classical-quantum transfer learning (CQTL). PVCQTL consistently achieves better accuracy. The modified observable variant reaches 85% accuracy for Deepfake dataset with lower computational cost. To evaluate generalizability, we tested PVCQTL on three additional binary classification datasets, observing improved accuracy on each. We conducted ablation studies to assess the effects of architectural choices on quantum component variations, including the choice of quantum gates, use of fixed ansatz circuits, and observable measurements. Robustness to input noise and sensitivity of the PVCQTL models were examined through ablation studies on learning rate, batch size, and number of qubits. These results demonstrate that PVCQTL offers a measurable improvement over traditional hybrid classical-quantum approaches.

Scientific Reports•Jul 2, 2025•8 citations

Remote Sensing Imagery Classification Techniques Using Quantum Deep Learning

Puneet Anchalia, G.S. Smrithy, D. Kavitha +1 more

Remote sensing imagery classification remains a formidable challenge, primarily attributed to the high dimensionality and complexity inherent in the data. While deep learning has demonstrated efficacy in this domain, its computational demands and reliance on extensive training data pose significant hurdles. This research investigates how quantum computing could be able to overcome these constraints and improve deep learning’s capacity for classifying remote sensing data. The study provides an extensive review of recent developments in quantum deep learning techniques specifically tailored for remote sensing applications. Various quantum algorithms designed for image feature extraction and classification are discussed, shedding light on their potential to revolutionize the field. As a practical demonstration, the chapter presents a case study focused on leveraging quantum deep learning for land cover classification using Sentinel-2 imagery. This chapter’s main focus is on hybrid quantum convolutional neural networks (QCNNs) based on circuits as effective tools for remote sensing picture classification. These hybrid models integrate quantum layers within traditional neural networks, introducing a novel paradigm for image analysis. To evaluate the efficacy of this approach, the proposed QCNN architecture is applied to land-use and land-cover classification, employing the Euro SAT dataset as a benchmark.

Auerbach Publications eBooks•Mar 3, 2025•0 citations

Comparing quantum and classical machine learning for radar-based drone classification: a like-for-like benchmark on noisy data

Stefan Klug, Stefan Pickl, Maximilian Moll

Abstract Reliable radar-based classification of small unmanned aerial vehicles (UAVs, drones) is hampered by low signal-to-noise ratio (SNR), non-Gaussian clutter, and site-specific shifts. This motivates interest in Quantum Machine Learning (QML) as a potentially more noise-resilient alternative to classical methods, yet reproducible advantages over strong classical baselines remain unclear. This study presents a controlled, like-for-like comparison between feed-forward neural networks and pure QML models without a preceding classical dimensionality-reduction encoder. All models operate on the same binned Fourier features. The QML variants span two data encodings (angle, amplitude), two variational circuit families, and an optional Quantum Fourier Transform, and are trained either with Adam or the gradient-free Asexual Reproduction Optimization (ARO) under matched evaluation budgets.Across three regimes (noiseless, additive white Gaussian noise (AWGN), and colored AR(1) complex Gaussian noise (cAR(1)-GN)), QML achieves competitive accuracy with only tens of parameters and, in two settings (noiseless and AWGN), slightly exceeds the best classical baselines, although these differences are not statistically significant. Under temporally correlated cAR(1) Gaussian noise, the tuned classical network significantly outperforms the best QML configuration. Overall, QML provides parameter-efficient performance parity in two of three conditions but no definitive advantage. The results highlight a strong dependence on the noise regime and point to angle-encoded, ARO-trained circuits as the most promising QML candidates for noisy radar-based drone classification.

•Feb 26, 2026•2 citations
Unmanned Aerial Vehicles (UAVs) enable high-resolution remote-sensing for land-cover and object mapping, yet conventional deep learning (DL) models often remain sensitive to scale, viewpoint, and illumination variability in UAV imagery. This paper presents QTL-SAIC, a quantum transfer learning framework for supervised aerial image classification that integrates classical feature learning with quantum-enhanced decision making. A pre-trained Capsule Network (CapsNet) extracts compact, discriminative features, which are classified using a variational quantum neural network (QNN) within a hybrid quantum–classical pipeline. The quantum module is fully specified (qubit number, encoding, circuit architecture, measurements, optimization, and execution backend with shot settings) to support reproducibility. Hyperparameters are optimized via the Gannet Optimization Algorithm (GOA), and evaluation follows stratified splits, repeated multi-seed trials and statistical reporting, with class imbalance addressed through cost-sensitive and/or balanced sampling strategies. Results indicate that QTL-SAIC yields competitive and stable performance relative to resource-matched classical baselines. Although no formal quantum advantage is claimed, ablations on circuit depth and noise sensitivity delineate the conditions under which the quantum component contributes to predictive performance.
Quantum Transfer Learning (QTL) offers a promising approach for visual quantum machine learning under near-term constraints, where limited qubit counts, shallow circuit depths, and costly hybrid optimization restrict end-to-end quantum training. In this setting, pretrained classical backbones can extract high-level visual features, while compact quantum modules operate as trainable classification heads. However, existing QTL results are difficult to compare because they often differ in datasets, preprocessing, backbone settings, qubit budgets, circuit designs, optimization choices, and reporting protocols. This work presents a controlled benchmarking methodology for evaluating representative QTL methods under a unified transfer-learning pipeline. The benchmark compares DQN-QTL, QPIE-QTL, AE-CQTL, PVCQTL, and ED-QTL under shared preprocessing rules, frozen-backbone settings, training conditions, and reporting metrics. The evaluation focuses on Fashion-MNIST and Hymenoptera Ants vs. Bees as the two main datasets, while CIFAR-10 is used to provide additional configuration-level evidence on a harder natural-image task. Beyond predictive performance, the benchmark analyzes circuit size, trainable parameters, quantum parameters, training time, and architectural sensitivity to qubit count and circuit depth. The results show that no single QTL family dominates across all settings: performance depends on the dataset, encoding strategy, circuit design, and computational cost. These findings highlight the need for resource-aware QTL evaluation and provide guidance for selecting hybrid quantum-classical transfer models under near-term resource constraints.
This chapter presents quantum transfer learning (QTL) as a practical and conceptually grounded framework for integrating quantum computation into modern machine learning. Building on the classical transfer learning paradigm, this chapter shows how pretrained classical models can be reused for large-scale representation learning while variational quantum circuits (VQCs) are deployed as compact downstream adaptation modules. Emphasis is placed on hybrid classical-quantum architectures in which quantum feature extractors operate on low-dimensional embeddings produced by convolutional networks, transformers, or graph neural networks, and their measurement-based outputs are processed by classical heads for prediction. This chapter develops the mathematical and architectural foundations of this approach, explains the role of variational quantum feature maps, and examines why quantum feature extractors may be useful in data-limited and structure-rich settings. Applications to text, image, and molecular data are discussed to illustrate the domain-independent nature of the framework. Practical implementation is addressed through PennyLane and PyTorch integration, highlighting differentiable quantum layers and end-to-end hybrid optimization. A hands-on example using encoded embeddings and VQCs demonstrates methodology, benchmarking, and model evaluation in a realistic supervised-learning setting. This chapter concludes with a critical discussion of limitations, including task dependence, noise, measurement cost, benchmarking challenges, and the continued dominance of large pretrained classical encoders.
Intrusion detection in unmanned-aerial-vehicle (UAV) swarms is complicated by high mobility, non-stationary traffic, and severe class imbalance. Leveraging a 120 k-flow simulation corpus that covers five attack types, we benchmark three quantum-machine-learning (QML) approaches - quantum kernels, variational quantum neural networks (QNNs), and hybrid quantum-trained neural networks (QT-NNs) - against strong classical baselines. All models consume an 8-feature flow representation and are evaluated under identical preprocessing, balancing, and noise-model assumptions. We analyse the influence of encoding strategy, circuit depth, qubit count, and shot noise, reporting accuracy, macro-F1, ROC-AUC, Matthews correlation, and quantum-resource footprints. Results reveal clear trade-offs: quantum kernels and QT-NNs excel in low-data, nonlinear regimes, while deeper QNNs suffer from trainability issues, and CNNs dominate when abundant data offset their larger parameter count. The complete codebase and dataset partitions are publicly released to enable reproducible QML research in network security.
We address the limitations of variational quantum circuits (VQCs) in hybrid classical-quantum transfer learning by introducing post-variational strategies, which reduce training overhead and mitigate optimization issues. Our approach Post Variational Classical Quantum Transfer Learning (PVCQTL) includes three designs: (1) modified observable construction, (2) a hybrid approach, and (3) a variational-post-variational combination. We evaluate these on pre-trained models (VGG19, ResNet50, ResNet18, MobileNet) for 4 and 8 qubits, with ResNet50 performing best in deepfake detection. Compared to classical models (MLP, ResNet50) and quantum baselines hybrid quantum classical neural network (HQCNN), classical-quantum transfer learning (CQTL). PVCQTL consistently achieves better accuracy. The modified observable variant reaches 85% accuracy for Deepfake dataset with lower computational cost. To evaluate generalizability, we tested PVCQTL on three additional binary classification datasets, observing improved accuracy on each. We conducted ablation studies to assess the effects of architectural choices on quantum component variations, including the choice of quantum gates, use of fixed ansatz circuits, and observable measurements. Robustness to input noise and sensitivity of the PVCQTL models were examined through ablation studies on learning rate, batch size, and number of qubits. These results demonstrate that PVCQTL offers a measurable improvement over traditional hybrid classical-quantum approaches.
Remote sensing imagery classification remains a formidable challenge, primarily attributed to the high dimensionality and complexity inherent in the data. While deep learning has demonstrated efficacy in this domain, its computational demands and reliance on extensive training data pose significant hurdles. This research investigates how quantum computing could be able to overcome these constraints and improve deep learning’s capacity for classifying remote sensing data. The study provides an extensive review of recent developments in quantum deep learning techniques specifically tailored for remote sensing applications. Various quantum algorithms designed for image feature extraction and classification are discussed, shedding light on their potential to revolutionize the field. As a practical demonstration, the chapter presents a case study focused on leveraging quantum deep learning for land cover classification using Sentinel-2 imagery. This chapter’s main focus is on hybrid quantum convolutional neural networks (QCNNs) based on circuits as effective tools for remote sensing picture classification. These hybrid models integrate quantum layers within traditional neural networks, introducing a novel paradigm for image analysis. To evaluate the efficacy of this approach, the proposed QCNN architecture is applied to land-use and land-cover classification, employing the Euro SAT dataset as a benchmark.
Abstract Reliable radar-based classification of small unmanned aerial vehicles (UAVs, drones) is hampered by low signal-to-noise ratio (SNR), non-Gaussian clutter, and site-specific shifts. This motivates interest in Quantum Machine Learning (QML) as a potentially more noise-resilient alternative to classical methods, yet reproducible advantages over strong classical baselines remain unclear. This study presents a controlled, like-for-like comparison between feed-forward neural networks and pure QML models without a preceding classical dimensionality-reduction encoder. All models operate on the same binned Fourier features. The QML variants span two data encodings (angle, amplitude), two variational circuit families, and an optional Quantum Fourier Transform, and are trained either with Adam or the gradient-free Asexual Reproduction Optimization (ARO) under matched evaluation budgets.Across three regimes (noiseless, additive white Gaussian noise (AWGN), and colored AR(1) complex Gaussian noise (cAR(1)-GN)), QML achieves competitive accuracy with only tens of parameters and, in two settings (noiseless and AWGN), slightly exceeds the best classical baselines, although these differences are not statistically significant. Under temporally correlated cAR(1) Gaussian noise, the tuned classical network significantly outperforms the best QML configuration. Overall, QML provides parameter-efficient performance parity in two of three conditions but no definitive advantage. The results highlight a strong dependence on the noise regime and point to angle-encoded, ARO-trained circuits as the most promising QML candidates for noisy radar-based drone classification.