Chapter 11 — Convolutional Neural Networks

(Roadmap — content to be written)

This chapter covers:

• LeNet → AlexNet → VGG → ResNet progression

• Pooling and spatial invariance

• How depth builds hierarchical features

• Why CNNs outperform SIFT at the normalisation ceiling

Depends on: Chapters 9–10 Builds toward: Chapter 12 (training), Chapter 14 (ViT)