📑   Deep Learning. Practice Project 8_1: Exercises with Adversarial Regularization

🌀   Homepage     🌀   Project List     🌀   GitHub Pages     🌀   Previous     🌀   Main     🌀   Instagram Posts     🌀   Pinterest Posts
In this project, we'll classify images from the dataset Art Pictograms.
The content is about 3500 images (64x64x3) with drawings stored in the file of Hierarchical Data Format (HDF5).
In the original dataset, photo files have the extension PNG and they are labeled by file prefixes.
We'll preprocess the images, represent their examples, then train neural networks.
We are going to apply Keras: a Python Deep Learning Library.

✒️   Code Modules, Helpful Functions, & Settings





✒️   Data Loading & Processing


💡 The cells below do not work with Python 3.9.

Unfortunately, we can train with guarantee only one neural network at ones.

Please restart the page and load the data again for training each of them.

✒️   Models with Adversarial Regularization => MLP




✒️   Models with Adversarial Regularization => CNN