Merging Realms: Classical and Quantum-Enhanced Deep Learning Techniques for CIFAR-10 Classification

Classical and quantum-enhanced deep learning techniques are revolutionizing areas of artificial intelligence. This leads to substantial innovations in various domains, including the image processing. The CIFAR-10 dataset is a widely used and benchmarked dataset for assessing the performance of deep learning models. In this proposed method, the CIFAR-10 dataset is used for image classification in an effort to improve model performance. The modern and strong models are employed, such as CNN, InceptionV3, DenseNet, and a Hybrid Quantum-DenseNet. The precision, recall, accuracy, loss, and F1-score metrics are used in this analysis to assess these models. From the analysis, it is observed that CNN and Hybrid Quantum-DenseNet attained 77%-78% accuracy. The InceptionV3 reached the maximum accuracy of 90.15%. The performance of the developed hybrid model was exceptionally well across a range of parameters. DenseNet achieved an accuracy of up to 84%.