CIFAR-10 Dataset Image Classification Using CNN

Image classification is an essential technology that is widely used in all aspects of human life. This work combined multiple image feature sources using deep learning algorithms to identify photos from the publicly available CIFAR10 image dataset. The same architecture, which consists of convolution modification and max-pooling of layers connected with several completely linked layers, is used to build most conventional convolutional neural networks (CNNs). This system consists of two steps: Pre-processing and Classification. The modified bilateral filtering method was employed in this study’s preprocessing steps. This study used convolutional neural network (CNN) architectures as suggested by the VGG-19 model to classify the CIFAR10 image dataset. The accuracy, precision, recall, and F1-score of the results are evaluated after simulating the suggested model. The testing results demonstrate that the proposed method reach a recall value of 96.4%, F1-score of 94.1%, accuracy of 97.3%, and maximum precision of 96.1%.