In this post you’ll learn how to train on large scale image datasets with Keras. We’ll leverage python generators to load and preprocess images in batches.
We’ll use the IMDB-WIKI dataset as an example.
Loading the Dataset in Batches
Let’s load the image dataset in batches of 100 images.
Keras Image Generators
Keras image generators allow you to preprocess batches of images in real-time.
Training Code ⤧ Next post Predicting Vehicle Motion and Direction using Optical Flow ⤧ Previous post Running Tensorflow with Nvidia Docker