Randomized trial evaluates binary image classification efficacy in quantum-classical models, suggesting implications for quantum integration.
Key Points
This research aims to evaluate the impact of qubit numbers on the performance of hybrid quantum-classical convolutional neural networks in binary image classification.
Developed a hybrid quantum-classical convolutional neural network model (HQ-CNN).
Tested the model on MNIST and EMNIST datasets for binary image classification.
Analyzed the challenges of integrating quantum computing into machine learning.
Increasing the number of qubits correlates with improved image classification efficacy.
An increased number of qubits does not guarantee overall model performance enhancement.