Course 2
Deep Learning for Advanced Image Classification
This course focuses on using deep learning techniques for advanced image classification tasks. It covers more sophisticated architectures such as deeper CNNs and transfer learning to improve performance on complex image datasets.
Week 1
Recap of CNNs and Deep Learning Basics
- Understanding the structure of deeper CNNs.
- CNN architectures: LeNet, AlexNet, and VGG.
Week 2
Transfer Learning for Image Classification
- Introduction to pre-trained models and fine-tuning.
- Using models like ResNet, Inception, and MobileNet for image classification.
Week 3
Advanced CNN Architectures for Image Recognition
- Deep architectures and their implementation: ResNet, DenseNet, etc.
- Regularization techniques: dropout, batch normalization.
Week 4
Object Detection with Deep Learning
- Understanding object detection techniques: YOLO, SSD, and Faster R-CNN.
- Applying object detection to real-world images (e.g., street scenes, medical images).
Week 5
Image Classification with Multi-Label and Multi-Class Problems
- Techniques for handling multi-label and multi-class classification problems.
- Evaluating multi-label models and threshold tuning.
Week 6
Final Project: Building a Real-World Image Classifier
Build a model for classifying complex images using deep learning and transfer learning techniques.