From the course: TensorFlow 2.0: Working with Images
Join today to access over 23,200 courses taught by industry experts.
What is transfer learning?
From the course: TensorFlow 2.0: Working with Images
What is transfer learning?
- [Instructor] Transfer learning is made up of two components, pretraining and fine tuning. Pretraining involves training a model from scratch. So this means the model weights are randomly initialized. The model is of no use at this point. The model is then trained on thousands of images and becomes useful for computer vision tasks such as image classification, but you can use it for a wide variety of domain. So you could use it for text or video or audio. And as you can imagine, we need both a lot of data and a lot of compute power. For the ImageNet Challenge, computer vision models had to distinguish between 1000 different categories of images. This means that these deep learning networks learn a whole load of features such as edges and corners and textures of images in the process. There are many models that performed very well on this dataset, including VGG16, VGG19, Inception version two and three, and ResNet50…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.