This repository contains Deep Learning code to understand DL concepts like CNN, R-CNN, GANs etc.
On the highest levels ('cnn_model.py') there is a general Neural Network architecture. Within this you can use the specific variables, weights, and the layers belonging to the LeNet5, AlexNet of VGGNet-16 architectures.
The LeNet5 architectures have been trained with the MNIST, CIFAR-10 and Oxflower-17 datasets and yield the following accuracies:
The AlexNet architecture haS been trained with the Oxflower-17 dataset for different optimizers:
