2026· E3S Web of Conferences· 0 citations· 9 references
TL;DR
A hybrid quantum-classical convolutional neural network model (HQ-CNN) is presented and its performance on binary image classification tasks using subsets of the MNIST and EMNIST datasets is evaluated.
Abstract
Machine learning (ML) has proven its efficacy in acquiring information from previous data and lever-aging that understanding to handle practical problems. The recent development of quantum computing, based on the principles of quantum physics, has made the integration of quantum computing and machine learning a prominent area of research. Nonetheless, the practical utilization of quantum computing in machine learning continues to present considerable obstacles, especially regarding implementation and the management of real-world tasks. This research presents a hybrid quantum-classical convolutional neural network model (HQ-CNN) and evaluates its performance on binary image classification tasks using subsets of the MNIST and EMNIST datasets. The study also presents an in-depth discussion of the challenges that arise from integrating quantum computing into machine learning, emphasizing the training procedure, the results obtained post-training, and the number of qubits used in the research. The experimental findings indicate that increasing the number of qubits generally enhances image classification efficacy. However, an increased number of qubits does not inherently ensure enhanced overall model performance.
The intersection of quantum processing and classical machine learning has spawned hybrid quantum-classical systems — practical system designs that attempt to use quantum potential in the limitations of modern Noisy Intermediate-Scale Quantum (NISQ) hardware. This paper introduces the development, deployment, and empirical analysis of a hybrid quantum-classical image classifier (which combines a Convolutional Neural Network (CNN) with an eight-qubit Variational Quantum Circuit (VQC)) in the binary classification of handwritten digits. This implementation, implemented in PennyLane and PyTorch, reaches a peak test accuracy of 99.85% across 1,984 test samples of the MNIST system, and only three errors are made. In addition to performance measures, the work presents an approach based on Quantum Software Engineering (QSE) by reporting major engineering issues, such as quantum-classical interface design, adjoint differentiation, feature-dimensionality reduction, and backend portability, and suggestion of seven quality-assurance practices of hybrid quantum software systems. The results demonstrate that it is possible to manufacture successfully hybrid quantum-classical architectures with the help of existing open-source tools and simulators, and provide future research and practice with hybrid QSE with effective advice.
Haider Ali, Muhammad Azeem Akbar, A. Khan et al.· SIGSOFT FSE Companion· 0 citations
Hybrid classical-quantum machine learning is a promising approach for image classification, but current methods often require too many qubits (quantum resources), creating a bottleneck for practical use. Current methodologies either constrain classification to basic binary decisions due to intricate quantum circuit design or depend on a Convolutional Neural Network (CNN) feature extractor that interfaces with a quantum layer necessitating a substantial quantity of qubits, frequently equivalent to the number of extracted features. To solve this, we propose a new Hybrid CNN-Quantum framework that dramatically reduces the required quantum resources. Our key innovation is an amplitude-encoding-inspired technique and a new activation function that together allow us to infer the final classification using only ⌈log2 (number of classes)⌉ readout qubits. We validated our framework on MNIST, Fashion-MNIST, and KMNIST, achieving accuracies of 0.9697, 0.8515, and 0.9258, respectively, using just 8 qubits. This matches or surpasses prior results that used over 1500 qubits, highlighting our method's competitive accuracy with drastically reduced quantum hardware requirements.
Over The past decade, quantum computing and machine learning have each witnessed tremendous growth in their respective fields. While quantum computing promises a completely novel paradigm for computation, based on the principles of superposition, entanglement, and interference, machine learning has firmly cemented its place in the sphere of data-driven decision-making. This paper seeks to provide a practical solution to this question, implementing a quantum classical learning system that is capable of performing binary classification tasks. We achieve this through the creation of a Quantum Kernel Support Vector Classifier (QSVC) using the Qiskit framework. While the classical feature set from the Iris dataset is scaled appropriately for the quantum system, the feature set is then represented in the quantum state through the ZZFeatureMap circuit. The QSVC utilizes the kernel-based similarity between quantum states. The results were promising, with the QSVC obtaining a classification accuracy of 92% over a test set, with the model trained in 6.32 seconds on a simulated quantum backend. The decision boundaries learned by the model were complex, and the quantum kernel representation of the feature set was able to capture the underlying class structure, which would have been difficult to achieve in a classical low-dimensional feature space. Our method, with quantum circuits relegated to state preparation and measurement while leaving the optimization to classical processes, is designed to operate within these limits. This makes our method not only theoretically interesting but also immediately applicable to existing simulators and quantum hardware.
Hersh Kumar, Eslavath Hemanth Naik, Gadige Hemanth et al.· International Conference Com...· 0 citations
Quantum Machine Learning (QML) combines quantum computing principles and traditional machine learning principles to provide new ways to solve difficult classification problems through the use of quantum technology. The goal of this paper is to implement a Quantum Kernel Support Vector Machine (QKSVM) using PennyLane and evaluate it against a standard image classification dataset, Fashion-MNIST. The Fashion-MNIST images are represented as 28x28 grayscale pixel arrays; they are normalized using a min-max normalizer with output values between zero and one. The 28x28 arrays are then dimensionally reduced to a size of four principal components (approximately 85% cumulative explained variance), and the QKSVM will utilize a quantum circuit limited to four qubits which is consistent with today's Noisy, Intermediate-Scale Quantum (NISQ) capabilities for performing simulations. Each input feature is represented as a qubit via an RY rotation and the qubits are then entangled using a CNOT circuit to generate the quantum kernel matrix, which is computed as the pairwise fidelity of each input image's state, which is subsequently used by a classical SVM classifier implementing a one-versus-all classification method. The QKSVM achieved approximately 40% accuracy and a macro-F1 score of 0.37 on a 100 sample stratified subset of the Fashion-MNIST test dataset, compared with 60% and 91.2%, respectively, using a classical Radial Basis Function Support Vector Machine (RBF-SVM) classifier with the same four principal components of input features and a Convolutional Neural Network (CNN) with full input images. With a primary bottleneck of qubit count, circuit depth, and shot noise, this work serves as a fully documented and reproducible baseline for future research on quantifying the performance of hybrid classical and quantum image classifiers across standardized image dataset benchmarks.
M. Devi, S.Sravanthi, V. Chaithanya et al.· 2026 7th International Confe...· 0 citations
A systematic comparison of four classical machine learning architectures, support vector machines, artificial neural networks, convolutional neural networks, and long short-term memory networks against their quantum counterparts against their quantum counterparts characterize the trade-offs between classical and quantum approaches under realistic, resource-constrained conditions and provide a benchmark for future studies on actual quantum hardware.
Tariq Mahmood, Z. Abidin, Itzel Luviano Soto et al.· 0 citations
This thesis investigates the application of machine-learning methods in the context of quantum computing and neutrino physics, with particular emphasis on the construction of effective representations for complex, high-dimensional data. The first part of the work is devoted to Quantum Extreme Learning Machines (QELMs), a hybrid quantum--classical framework in which classical data are encoded into quantum states and processed through fixed quantum dynamics, while learning is performed by a classical readout layer. Within this framework, we analyze the role of encoding strategies, feature-reduction methods, Hamiltonian structure, and measurement, with particular focus on the relationship between quantum dynamics, expressivity, entanglement, and classical simulability. The second part of the thesis concerns the application of deep learning to the analysis of images produced by water Cherenkov detectors in neutrino physics. Convolutional architectures, including residual networks, are developed for the classification of complex events in realistic simulated datasets, showing that such models can effectively extract relevant information from detector data. Taken together, these results highlight the potential of machine learning, in both its classical and quantum forms, as a powerful framework for the analysis of complex data in fundamental physics, while also outlining relevant challenges and directions for future research.
