What you'll learn

Machine learning practitioners often use CNNs as black boxes, limiting their ability to diagnose issues, select appropriate architectures, or explain model behavior to stakeholders. In this course, Convolutional Neural Networks (CNNs), you’ll gain the ability to understand how CNNs process spatial data and make informed architectural decisions for computer vision tasks. First, you’ll explore the fundamental building blocks of CNNs, including convolutional layers, pooling operations, and how these components address the limitations of fully connected networks. Next, you’ll discover how CNNs learn hierarchical feature representations through training, from low-level edges to high-level object recognition. Finally, you’ll learn how landmark architectures like LeNet, AlexNet, VGG, and ResNet have evolved to solve increasingly complex vision problems. When you’re finished with this course, you’ll have the skills and knowledge of convolutional neural networks needed to evaluate, select, and reason about CNN architectures for real-world image processing applications.