We perform image classification on the Honda Scenes Dataset on Quantinuum's H-2 and IBM's Heron chips utilizing up to 72 qubits and thousands of two-qubit gates. For data loading, we extend the hierarchical learning to the task of approximate amplitude encoding and block amplitude encoding for commercially relevant images up to 2 million pixels. Hierarchical learning enables the training of variational circuits with shallow enough resources to fit within the classification pipeline. For comparison, we also study how classifier performance is affected by using piecewise angle encoding. At the end of the VQC, we employ a fully-connected layer between measured qubits and the output classes. Some deployed models are able to achieve above 90\% accuracy even on test images. In comparing with classical models, we find we are able to achieve close to state of the art accuracy with relatively few parameters. These results constitute the largest quantum experiment for image classification to date.
We perform image classification on the Honda Scenes Dataset on Quantinuum's H-2 and IBM's Heron chips utilizing up to 72 qubits and thousands of two-qubit gates. For data loading, we extend the hierarchical learning to the task of approximate amplitude encoding and block amplitude encoding for commercially relevant images up to 2 million pixels. Hierarchical learning enables the training of variational circuits with shallow enough resources to fit within the classification pipeline. For comparison, we also study how classifier performance is affected by using piecewise angle encoding. At the end of the VQC, we employ a fully-connected layer between measured qubits and the output classes. Some deployed models are able to achieve above 90% accuracy even on test images. In comparing with classical models, we find we are able to achieve close to state of the art accuracy with relatively few parameters. These results constitute the largest quantum experiment for image classification to date.
Given the excitement for the potential of quantum computing for machine learning methods, a natural subproblem is how to load classical data into a quantum state. Leveraging insights from [GST24] where certain qubits play an outsized role in the amplitude encoding, we extend the hierarchical learning framework to encode images into quantum states. We successfully load digits from the MNIST dataset as well as road scenes from the Honda Scenes dataset. Additionally, we consider the use of block amplitude encoding, where different parts of the image are encoded in a tensor product of smaller states. The simulations and overall orchestration of workflows was done on the BlueQubit platform. Finally, we deploy our learned circuits on both IBM and Quantinuum hardware and find that these loading circuits are sufficiently shallow to fit within existing noise rates.
Image data grows in volume and size more quickly than classical processing power. Thanks to its entanglement and super-position qualities, quantum computing, which is based on quantum physics, has a lot of potential for speed and processing capacity. For high power challenges, it is therefore highly common to try to utilize quantum computing units rather than classical computing units. In this study, a hybrid quantum-classical approach was proposed for utilizing quantum computers’ advantages in image classification. This hybrid technique uses a variational quantum circuit (VQC) on the quantum computer side. To overcome the qubit restriction in the VQC, multiple amplitude encoding was used as the data encoding method. In the classical computer part of the proposed approach, the preprocessing of the image, the convolution operation, and the optimization of the parameters of the single-qubit rotation gates in the VQC were performed. The proposed approach was trained and tested on two different data sets. The accuracy rates acquired in 2-layers and 4-layers VQCs within the data sets were evaluated in the test results. The proposed approach was evaluated against studies that were similar in the literature. Compared to similar studies, it was observed that it is more successful in terms of the number of parameters used and quantum cost. As a result, the effectiveness of the proposed approach was verified.
Quantum machine learning on real noisy intermediate-scale quantum (NISQ) hardware has remained largely confined to binary or few-class tasks, limited by the cost of on-hardware training and the underuse of large devices at inference. We present a unified framework that classifies ten-class MNIST end-to-end on a 127-qubit IBM Eagle processor, with three central contributions. First, a two-phase protocol decouples a gradient-based classical optimization of the encoder and readout from a gradient-free optimization of the quantum parameters, removing the parameter-shift gradient cost that makes on-hardware training impractical. Second, we introduce Quantum Multi-Programming to a trained quantum classifier for the first time, packing multiple circuit copies onto one device to deliver parallel inference at no mean-accuracy cost while cutting quantum-processor job submissions proportionally. Third, a controlled comparison shows that on-hardware fine-tuning yields no measurable accuracy gain, motivating a practical NISQ workflow: train on a classical simulator and reserve the hardware for inference only. Benchmarked against a matched-capacity classical network, the quantum module shows no per-parameter accuracy advantage at this scale; we therefore frame the work as a feasibility-and-workflow demonstration for multi-class quantum image classification on current hardware.
Brain tumor image classification is an essential pre-processing task for tumor diagnosis and treatment planning. In medical imaging, quantum computing analyse medical imaging data, offering potential speedups and new representations compared to classical methods. The brain images are transformed through resizing and normalization to a uniform 32 × 32 × 3 resolution. Subsequently, these processed images are converted into vectors to enable encoding into quantum states. The feature set is reduced using Principal Component Analysis (PCA) to ensure compatibility with the restricted qubit count in the quantum circuits. Through amplitude encoding, these vectors are mapped to quantum states, allowing for a compact and efficient data representation within the quantum domain. A parameterized Variational Quantum Circuit(VQC), consisting of entangling gates and trainable rotations, processes the encoded quantum states to learn features that can effectively discriminate between classes. The loss is calculated from the measured expectation values of Pauli-Z observables and optimized using classical gradient descent, where the objective function is the mean squared error. The trained quantum model is designed for multi-class classification, specifically distinguishing between meningioma, glioma tumors, pituitary tumors, and the absence of a tumor. The outcome of this proposed work demonstrates the VQC’s advantage over classical baselines, including Random Forest, and Support Vector Machine (SVM) and Multi-Layer Perceptron (MLP). The VQC achieved the highest accuracy of 95% and the best robustness (Specificity 0.983). Crucially, the VQC accomplished this using only 24 trainable parameters, showcasing a significant increase in parameter efficiency compared to the classical models. The novelty of this work is the quantitative demonstration that a custom-designed, resource-efficient quantum model can outperform established classical deep learning methods in both performance and efficiency for complex medical image classification. The VQC provides a 3.5-fold gain in parameter efficiency, achieving the highest accuracy of 95% using only 24 trainable parameters. This efficiency is enabled by the novel application of amplitude encoding to the medical images, which compresses 16 features into just 4 qubits. The VQC demonstrates robustness vital for medical diagnostics, achieving a Specificity of 0.983. Its shallow architecture is also NISQ-friendly and remains stable by avoiding the Barren Plateau issue (Gradient Norm: 0.11), validating the path toward quantum advantage in diagnostic tools. The results of this approach highlight the potential amplitude encoding in quantum machine learning, as a feasible and effective method for classifying real-world medical images. This work also opens avenues for future hybrid quantum-classical diagnostic systems that could utilize richer quantum state encodings for significant gains in computational efficiency and accuracy.
This thesis presents a systematic empirical evaluation of quantum machine learning performance under noisy intermediate-scale quantum (NISQ) era constraints. Through 670 controlled experiments, it evaluated quantum kernel support vector machines and variational quantum classifiers against classical baselines on MNIST binary and multiclass classification tasks with systematic variation of problem difficulty, feature dimensionality (4, 8 qubits), and training set size (n ∈{100, 250, 400, 500, 2000, 4000}). Statistical rigor was ensured through five random seeds per condition and comprehensive significance testing. During the testing with binary datasets, classical methods (SVM, logistic regression, k-NN, neural networks) achieved 85.9% to 99.6% accuracy with training times under one second across all sample sizes. Quantum kernel methods utilizing ZZ feature maps yielded 70.5% to 96.0% accuracy, requiring 129 to 2,140 seconds training with quadratic sample size scaling. Larger scale quantum experiments (n = 4000 for both architectures, 8-qubit encoding with n = 2000 for quantum kernels) proved computationally impossible on available hardware leading to out of memory errors, highlighting practical NISQ-era scalability constraints. Variational quantum classifiers exhibited persistent performance stagnation, failing to exceed 66.5% accuracy despite increases in training data and feature dimensionality. Performance gaps ranged from 3.3 to 35.2 percentage points. All quantum classical differences achieved statistical significance (p< 0.05, with 13 of 14 comparisons meeting p< 0.01) and exhibited large to very large efect sizes (d> 1.0, with 13 of 14 comparisons exceeding d> 2.0). To further evaluate performance and scalability, a 4 class classification experiment was conducted following an identical experimental protocol, with training set sizes of n ∈{100, 250, 400}. Across 180 runs, classical baselines achieved mean accuracies between 87.2% and 95% across both 4 and 8 dimensional data. Quantum Kernel SVM struggled during runs with 4 dimensions, accuracy varying between 64.6% and 74%. Quantum Kernel SVM accuracy declined 16%-25% when the dimension and qubit count were increased to 8, showing findings of exponential kernel concentration. The Variational Quantum Classifier also performed poorly, with accuracy ranging from 43.4% to 44.6% across all runs, providing further evidence that it failed to leverage additional features or training data. Experiments employed noiseless statevector simulation, establishing that quantum underperformance stems from fundamental algorithmic limitations independent of hardware noise. A further extension to 8 class classification (digits 0–7) confirmed and amplified these findings: classical methods maintained 65–83% accuracy while Quantum Kernel SVM degraded to 34–47% and VQC collapsed to 21–23%, approaching the 12.5% random baseline. Supplementary appendix experiments further examined amplitude encoding on raw 784 dimensional pixel features and an alternative angle encoding scheme on the 8-class task; the latter demonstrated that replacing the ZZ Feature Map with simple per qubit rotations substantially recovered quantum kernel performance, suggesting feature map design as a more consequential variable than circuit depth or qubit count in near-term quantum classification.
Abstract With the significant enhancement of quantum computing capabilities and their expanding applications in various fields, quantum image processing combined with quantum computing has become a key area of research in quantum computing, holding important theoretical value. In the face of the challenges posed by the digital information era, image classification technology demands higher standards from machine learning algorithms in terms of accuracy, computational speed, generalization capability, and complexity. This paper studies a series of quantum data encoding methods and parametric quantum circuit architectures, and proposes the application of Hybrid Hierarchical Encoding (HHE) and Hybrid Quantum Convolutional Neural Network (HQCNN). The new encoding method HHE integrates the benefits of angle encoding and enhances the hybrid amplitude encoding design to achieve optimal classification performance without compromising the depth of quantum circuits.HQCNN demonstrates outstanding classification performance and remarkable control over model complexity by leveraging the data processing capabilities of classical convolution and the complex computational power of quantum convolution. Experimental results show that HHE and HQCNN achieve better classification accuracy on the MNIST and Fashion-MNIST datasets compared to traditional classical CNN models, particularly when the number of parameters is significantly reduced.