In the previous chapter, we covered the major concepts surrounding transfer learning. The key idea was that leveraging state-of-the-art, pretrained deep learning models in a wide variety of tasks yields superior results compared to building your own deep learning models and architectures from scratch. In this chapter, we will gain a more hands-on perspective of actually building deep learning models using transfer learning and applying them to a real-world problem. We will build various deep learning models with and without transfer learning. We will analyze their architecture and also compare and contrast their performance. We will be covering the following major aspects in this chapter:
- The need for transfer learning
- Building Convolutional Neural Network (CNN) models from scratch:
- Building a basic CNN model
- Improving our CNN model...