02 — Convolutional Networks

Convolutional operations, pooling, and deep CNN architectures for spatial data. Includes applications in object detection, face recognition, and neural style transfer.

Notes

• Convolution, Padding, Stride, and Pooling — discrete convolution, cross-correlation, padding modes, max pooling, average pooling
• CNN Architectures — LeNet, AlexNet, VGG, Inception, ResNet residual connections
• Object Detection — sliding window, YOLO, anchor boxes, region-based CNNs
• Face Recognition — one-shot learning, Siamese networks, triplet loss
• Neural Style Transfer — content and style cost functions, Gram matrix

Links

← 01 — MLP and Representation Learning → 03 — Sequence Models

• Deep Learning Theory