AutoEncoder part of the project is almost done.
The file which is main.py encodes mnist dataset
and produces an encoded numpy array. After that if
you want to decode the numpy array then you can..
| Real | Encoded | Predicted | Decoded |
|---|---|---|---|
The NNs which helped them to be produced are trained 100 epoch.
Predicted one is predicted from encoded image by another
neural network which is trained by using predicted images (encoded)
from first autoencoder neural network and test images.
Decoded one is basically predicted by using the autoencoder neural network which is the main neural network to encode and then decode the images but even though it does both job, we need just the first one: Encoding.