LearningNeuralNetworks

Code for my experiments learning deep learning and neuronal networks

Here is a showcase of all most important projects, click the images to show the code:

Surnames origins

A RRN using PyTorch (+ a DIY LSTM module) that can predict the origin of a surname:

Show me the code

Style transfer

A neural network that can transfer style to an image, this uses convolutional neuronal networks and is an Udacity's PyTorch exercise.

Show me the code

Sign Languages

A neural network that can recognize with ~95% accuracy a letter written in sign language from a photo, created using TensorFlow.

Show me the code

CIFAR10

A RRN using PyTorch (using convolutional layers) that classifies images suing the CIFAR10 dataset.

Show me the code

MNIST using a MLP

This is a classic example in deep learning, this is a net that recognize digits:

Show me the code

 

XOR network

This is a classic exercise, is a simple net that can predict the XOR gate, for this I created myself an nn, and implemented backpropagation for scratch:

Show me the code