Convolutional Neural Networks

Convolutional neural networks (CNNs), also referred to as convolutional networks, are a class of deep feedforward neural networks. In recent years, many advances in deep learning—particularly in computer vision and image processing—have been driven in part by the widespread adoption of CNN-based architectures and the continued development of related training techniques (Li et al. in IEEE transactions on neural networks and learning systems 33:6999–7019, 2021). In this chapter, we introduce CNNs in a structured manner following their architectural components. We also provide several supplementary sections to highlight key concepts and practical considerations. In addition to the theoretical discussion, we include illustrative code examples to help readers better understand and apply convolutional neural networks.