Unet MNIST Denoising: Use MNIST data set and train a UNet autoencoder for denoising images with 80% Gaussian noise.
VGGNet CIFAR10: Use CIFAR-10 data set and train a VGGNet-like network a for 10 classes. Also, Plot learning rates and classification performance, and confusion Matrix for training and testing datasets.
Lenet MNIST: Use the MNIST data set and train a Lenet-like network a for 10 classes.
Lenet MNIST with Zero padding: Use the MNIST dataset with zero padding and train a Lenet-like network for 10 classes.