Autoencoder to denoise images based on the fashion_mnist dataset
The noisy data is randomly applied to the data with a noise factor of 0.2 and then clipped between 0.0 and 1.0.
The encoder comprises of 2 Convolutional layers used to highlight the features and negate the noise. The decoder works in sequence and uses Conv2dTranspose layers to undo the Convolutional layers but without the noise.
Data is from the fashion_mnist dataset on tensorflow datasets
from keras.datasets import fashion_mnist
(x_train, _), (x_test, _) = fashion_mnist.load_data()Install Python 3.8+
pip install tensorflow keras numpy