Advancements in Image Classification: From Machine Learning to Deep Learning

Abstract

Image classification, as an essential task within the realm of computer vision, has evolved from traditional machine learning methods to deep learning techniques. This paper systematically reviews the growth of image classification technology, beginning with the introduction of commonly used datasets such as CIFAR-10, ImageNet, and MNIST, and exploring their impact on algorithm development. Subsequently, the paper provides an in-depth analysis of image classification methods based on machine learning, including traditional algorithms such as Support Vector Machine (SVM), Random Forest, and Decision Tree. These methods achieve image classification through two stages: feature extraction and classification, but they encounter limitations when confronted with large-scale datasets and complicated tasks. Convolutional Neural Networks (CNNs) have gradually replaced traditional methods in image classification due to the rise of deep learning, resulting in improved accuracy and robustness. The paper also focuses on discussing classic deep learning models such as AlexNet, VGGNet, ResNet and ViT, analyzing their strengths and weaknesses. By comparing the performance of different methods, this paper aims to provide references for researchers in the realm of image classification, promoting further development in this area.