Here I show how to improve the dimention of image. I used Autoencoder from TF with keras. It has 2 part, Encoder & Decoder. The encoder will take an image as input and output a compressed reepresentatino. Then the decoder will take the input which is outputted by the Encoder and return a reconstructed image as output.
The dataset, trained model weights can befound in the data folder. All source code is available in the jupyter notebook. Some theory is also included with the notebook ,which can also be found at images folder.
Dont forget to change the path of the project if you want to run it. Check the Part 5, 6 & 7 where the path is specified to the dataset and the saved model weight folder.