poplasugar.blogg.se - Keras data augmentation tutorial

Keras data augmentation tutorial how to#
Keras data augmentation tutorial generator#

Let’s try to process the previously seen picture of the building in Crete with this generator. The Keras library offers a set of helpful tools for generating data. using realistic game environments such as GTA. As a curiosity: technologies related to autonomous vehicles are also trained on artificially generated data sets, e.g. This is a field where the possibilities are really great. In practice, a collection of tens of thousands of images can become a collection with millions of elements. Change its colors or make other more or less subtle changes that will give the model tons of new data. What can we actually do with the image we want to artificially process? In theory, we have infinitely many solutions: we can slightly rotate the image, in any direction, at any angle. For the algorithm, looking at the object from a different perspective is valuable information that allows to better generalize the training process. It will still be the same vehicle and it will definitely be a car for our brain. Standing in front of the car, we can look at it centrally or slightly from the side. Moreover, such “minor modifications” occur in the real world. Why? Because the images are susceptible to minor modifications, which will be a new data for the algorithm, although they will still be basically the same for the human eye. This is a particularly valuable technique for image analysis. one that cannot be significantly enlarged or enlarged is very expensive, we can reach for the so-called data augmentation.

And how well will our model do on the set of artificially generated (augmented) data?Īs I mentioned in the previous part of the tutorial, if we are dealing with a closed data set, i.e.

Keras data augmentation tutorial how to#

How to artificially generate new data for the CIFAR-10 set?.

Keras data augmentation tutorial generator#

How to use the data generator from the Keras library?.

In the fourth part of the tutorial you will learn: the artificial generation of large amounts of new data. I will use the so-called data augmentation, i.e. I leave the first two solutions for you 😉 to experiment with, and in this part of the tutorial I want to feed our network with more data. To go further we would have to change the architecture of our network, experiment with hyperparameters or get more data. In the classification of the CIFAR-10 dataset we achieved 81% on the test set. We looked at the convolution operation, the convolutional network architecture, and the problem of overfitting. In the previous three parts of the tutorial, we learned about convolutional networks in detail.