ABSTRACT
This chapter focuses on something very interesting called Transfer Learning. Transfer Learning is the process of transferring the knowledge acquired in implementing or learning one task to implement another related task. The first section covers what Transfer Learning is, followed by implementation of most simple VGG 16 model for the scenario, Cat or Dog Using Transfer Learning with VGG 16. The Keras library has a wide range of state-of-the-art models that can be used for customized solutions. The next example shows, Identify Your Relatives’ Faces Using Transfer Learning by Leveraging VGG16 Architecture. The original vgg16 architecture had 1000 output classes, wherein the Transfer Learning counterpart for identifying relatives’ faces will just have 4 to 5 output classes, depending on the number of relatives we want to identify. Next, Transfer Learning strategies are discussed, and the chapter ends by summarizing the key points and has a quiz.
