Autoencoders MNIST Project

This is a repository demonstrating a vanilla Autoencoder using MNIST dataset.

Autoencoders are used for creating a compressed representation of the given data while retaining the original quality of the data.

Here are some examples generated using the model given in code:

Network Type	Fully Connected Layers	Convolution Neural Networks
Original
Reconstructed

Here are the Loss graphs for Fully connected Layers:

Here are the Loss graphs for Convolutional Neural Networks: