This Project is a deep-dive into the use of Neural Networks as Image Compression tools.
The File entitled 'First_attempt.ipynb' is an Autoencoder that is capable of compressing large groups of images with very little space for all of them. The Latent Representations are tiny and can be saved to reconstruct the full image with high accuracy. The layers should be adjusted for the encoder/decoder as your image size changes. The current layers are good for 64x64 images, as the image becomes larger, we should add more layers.