In this 1.5-hour long project-based course, you will learn how to apply image data augmentation in Keras. We are going to focus on using the ImageDataGenerator class from Keras’ image preprocessing package, and will take a look at a variety of options available in this class for data augmentation and data normalization.
Since this is a practical, project-based course, you will need to prior experience with Python programming, convolutional neural networks, and Keras with a TensorFlow backend.
Data augmentation is a technique used to create more examples, artificially, from an existing dataset. This is useful if your dataset is small and you want to increase the number of examples. Data augmentation can often solve over-fitting so that your model generalizes well after training. For images, a variety of augmentation can be applied to increase the number of examples.
Note: This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.
Since this is a practical, project-based course, you will need to prior experience with Python programming, convolutional neural networks, and Keras with a TensorFlow backend.
Data augmentation is a technique used to create more examples, artificially, from an existing dataset. This is useful if your dataset is small and you want to increase the number of examples. Data augmentation can often solve over-fitting so that your model generalizes well after training. For images, a variety of augmentation can be applied to increase the number of examples.
Note: This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.
