testgithubrittttttt/Image-Denoising-with-Autoencoders
This project focuses on building an image denoising system using autoencoders. The autoencoder is trained to remove noise from the MNIST dataset images, which are commonly used for digit recognition tasks. The project includes several advanced features such as data augmentation, model checkpointing, learning rate scheduling, and comprehensive eval.
Jupyter NotebookMIT