Curator's Take
This comprehensive benchmarking study tackles one of quantum machine learning's most pressing questions by directly comparing classical and quantum approaches across multiple practical dimensions including accuracy, runtime, and memory requirements. The finding that quantum support vector machines consistently outperform their classical counterparts on MNIST data, while quantum and classical convolutional neural networks achieve comparable accuracy, provides valuable empirical evidence for where quantum advantages might actually emerge in near-term applications. Particularly noteworthy is the identification of specific operating points where quantum methods offer the best accuracy-to-runtime tradeoffs, giving researchers concrete guidance on when quantum machine learning might be worth the additional complexity. This type of rigorous, multidimensional comparison is exactly what the field needs as it moves beyond theoretical promises toward practical quantum advantage demonstrations.
— Mark Eatherly
Summary
The rapid growth of computer vision and increasingly complex image recognition tasks has exposed fundamental computational limitations of classical machine learning models, motivating the exploration of quantum computing as an emerging new paradigm. This paper presents a comprehensive benchmarking study of classical and quantum machine learning models for image recognition on the MNIST handwritten digit dataset, evaluating both traditional models, a Classical Support Vector Machine (CSVM) and a Quantum Support Vector Machine (QSVM), and deep neural network models, a Classical Convolutional Neural Network (CCNN) and a Quantum Convolutional Neural Network (QCNN), across four performance dimensions: classification accuracy, computational runtime, parameter count, and memory requirements. Experiments are conducted as functions of both feature dimensionality and sample size, and across CPU and GPU execution environments, providing a controlled, multidimensional comparison to address gaps in prior work. For the SVM-based models, QSVM consistently outperforms CSVM in accuracy, reaching $\sim$ 0.90 versus $\sim$ 0.85 at 1,000 samples, with a higher computational cost. A feature count of 10 qubits and a sample size in the range of 200 -- 500 emerge as practical operating points that balance accuracy and runtime. For the neural network models, CCNN and QCNN achieve comparable classification accuracy, both exceeding 0.96 at 64 features and 60,000 samples, yet QCNN offers substantially superior parameter and memory efficiency, requiring $\sim$ 94\% fewer parameters and $\sim$ 75\% less memory than CCNN at higher feature counts, while incurring higher runtime. Across both model families, quantum models consistently outperform classical models by greater margins in accuracy as feature dimensionality or sample size increases.