This paper proposes a quantum transfer learning (QTL) algorithm that combines a transfer learning (TL) algorithm with parameterized quantum circuits (PQC) to investigate the relationship between PQC and model performance in hybrid quantum classical neural networks (HQCNN). The QTL algorithm effectively improves training speed compared to current quantum algorithms, facilitating a more effective exploration of the relationship between PQC and HQCNN. Experimental results show that the expressibility of PQCs is closely related to circuit structure, overlap depth, and the number of qubits. The all-circuits and annular-circuits (APQC) demonstrate better expressibility. Additionally, the expressibility of annular-circuits strongly correlates with the accuracy of the QTL model. The paper also explores the impact of pre-trained models on QTL performance. Ultimately, the pre-trained model ResNet18 and the quantum classifier based on APQC are chosen to form the QTL model. In terms of training time, QTL completes 20 epochs in 4 hours, saving 90% of time compared to general QCHNN.
This paper proposes a quantum transfer learning (QTL) algorithm that combines a transfer learning (TL) algorithm with parameterized quantum circuits (PQC) to investigate the relationship between PQC and model performance in hybrid quantum classical neural networks (HQCNN). The QTL algorithm effectively improves training speed compared to current quantum algorithms, facilitating a more effective exploration of the relationship between PQC and HQCNN. Experimental results show that the expressibility of PQCs is closely related to circuit structure, overlap depth, and the number of qubits. The all-circuits and annular-circuits (APQC) demonstrate better expressibility. Additionally, the expressibility of annular-circuits strongly correlates with the accuracy of the QTL model. The paper also explores the impact of pre-trained models on QTL performance. Ultimately, the pre-trained model ResNet18 and the quantum classifier based on APQC are chosen to form the QTL model. In terms of training time, QTL completes 20 epochs in 4 hours, saving 90% of time compared to general QCHNN.
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.
Quantum Transfer Learning (QTL) recently gained popularity as a hybrid quantum-classical approach for image classification tasks by efficiently combining the feature extraction capabilities of large Convolutional Neural Networks with the potential benefits of Quantum Machine Learning (QML). Existing approaches, however, only utilize gate-based Variational Quantum Circuits for the quantum part of these procedures. In this work we present an approach to employ Quantum Annealing (QA) in QTL-based image classification. Specifically, we propose using annealing-based Quantum Boltzmann Machines as part of a hybrid quantum-classical pipeline to learn the classification of real-world, large-scale data such as medical images through supervised training. We demonstrate our approach by applying it to the three-class COVID-CT-MD dataset, a collection of lung Computed Tomography (CT) scan slices. Using Simulated Annealing as a stand-in for actual QA, we compare our method to classical transfer learning, using a neural network of the same order of magnitude, to display its improved classification performance. We find that our approach consistently outperforms its classical baseline in terms of test accuracy and AUC-ROC-Score and needs less training epochs to do this.
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.
Hybrid quantum neural networks (HQNNs) integrate parameterized quantum circuits (PQCs) within classical networks, where the behavior of the underlying PQCs is often the primary focus of analysis. In this context, expressibility and trainability are widely used to characterize PQC's performance and are commonly assumed to exhibit a trade-off, where highly expressive circuits are more susceptible to barren plateaus. However, the validity of this relationship in HQNNs remains unclear. In this paper, we systematically analyze the expressibility--trainability relationship in HQNNs across varying circuit depths, qubit counts, entanglement topologies. We consider different training configurations, including pure PQCs, quantum-only training in hybrid setting, and full end-to-end training of hybrid models. Our results show that pure PQCs exhibit only a weak and regime-dependent trade-off, while hybrid architectures increasingly disrupt and can eliminate this relationship under full hybrid training. This indicates that classical components reshape the optimization landscape, decoupling trainability from PQC expressibility. We further propose a multi-objective neural architecture search (NAS) framework that jointly optimizes expressibility, trainability, and task performance over a combined classical--quantum design space, revealing different Pareto-optimal solutions under full end-to-end and quantum only training in hybrid setting. different trainability definitions. Our results suggest that hybridization is not just an implementation detail, but a defining factor in the performance of quantum machine learning models.
The rapid increase in the volume of data increased the size and complexity of the deep learning models. These models are now more resource-intensive and time-consuming for training than ever. This paper presents a quantum transfer learning (QTL) based approach to significantly reduce the number of parameters of the classical models without compromising their performance, sometimes even improving it. Reducing the number of parameters reduces overfitting problems and training time and increases the models' flexibility and speed of response. For illustration, we have selected a surface anomaly detection problem to show that we can replace the resource-intensive and less flexible anomaly detection system (ADS) with a quantum transfer learning-based hybrid model to address the frequent emergence of new anomalies better. We showed that we could reduce the total number of trainable parameters up to 90 % of the initial model without any drop in performance.
The classical-quantum transfer learning (CQTL) method is introduced to address the challenge of training large-scale, high-resolution image data on a limited number of qubits (ranging from tens to hundreds) in the current Noisy Intermediate-Scale quantum (NISQ) era. existing CQTL frameworks have been demonstrate quantum advantages with a small number of parameters (around 50), but the performance of quantum neural networks is sensitive to the number of parameters. Currently, there is a lack of exploration into larger-scale quantum circuits with more parameters. This paper proposes an amplitude-encoding-based classical-quantum transfer learning (AE-CQTL) framework, accompanied by an effective learning algorithm. The AE-CQTL framework multiplies the parameters of quantum circuits by using multi-layer ansatz. Based on the AE-CQTL framework, we designed and implemented two CQTL neural network models: Transfer learning Quantum Neural Network (TLQNN) and Transfer Learning Quantum Convolutional Neural Network (TLQCNN). Both models significantly expand the parameter capacity of quantum circuits, elevating the parameter scale from a few dozen to over one hundred parameters. In cross-experiments with three benchmark datasets (MNIST, Fashion-MNIST and CIFAR10) and three source models (ResNet18, ResNet50 and DenseNet121), TLQNN and TLQCNN have exceeded the benchmark classical classifier in multiple performance metrics, including accuracy, convergence, stability, and generalization capability. Our work contributes to advancing the application of classical-quantum transfer learning on larger-scale quantum devices in future.