CNN internals, data augmentation, transfer learning with ResNet, MobileNet, and EfficientNet on Imagenette, and object detection with YOLO and the two-stage R-CNN family on the Malaysia road-sign dataset.
IntermediateAbout 8 hours8 lessons
What you will learn
How a CNN sees, filters and feature maps
Grad-CAM, seeing what a CNN looks at
Data augmentation that actually helps
Fine-tune ResNet18 on Imagenette
MobileNet vs EfficientNet, speed vs accuracy
YOLO in practice, boxes and anchors
Region-based detectors, the R-CNN family
Curriculum
1
CNN internals
3 lessonsFree preview
How a CNN sees, filters and feature mapsPreview10m
Grad-CAM, seeing what a CNN looks atPreview8m
Data augmentation that actually helpsPreview8m
2
Transfer learning
2 lessons
Fine-tune ResNet18 on Imagenette12m
MobileNet vs EfficientNet, speed vs accuracy9m
3
Object detection and checkpoint
3 lessons
YOLO in practice, boxes and anchors10m
Region-based detectors, the R-CNN family1m
Checkpoint: train a YOLO detector on Malaysia road signs20m
Hands-on checkpoint
Every course ends with a Colab task you complete yourself. Our Gemini-assisted review gives feedback and points you to what to fix, it coaches you, it does not do it for you. This course uses the checkpoint dataset.