Deep Learning for Vision
A 16-module path through deep learning for vision, in study order from foundations to the frontier.
Modules
- CNNs for Images Revisited
- Image Classification with Deep Networks
- ImageNet and Benchmark Datasets
- Data Augmentation for Vision
- Transfer Learning for Vision
- Object Detection Fundamentals
- Anchor Boxes
- Non-Maximum Suppression
- The R-CNN Family
- Single-Shot Detectors (YOLO/SSD)
- Fully Convolutional Networks
- Semantic Segmentation
- U-Net
- Instance Segmentation with Mask R-CNN
- Vision Transformers
- Self-Supervised Vision