FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoising (TIP, 2018)

MCNet: An Efficient CNN Architecture for Robust Automatic Modulation Classification

Semantic information can help CNNs to get better illuminant estimation -- a proof of concept

Code for paper "Application of Convolutional Neural Networks to Four-Class Motor Imagery Classification Problem"

A convolutional neural network (CNN or ConvNet) is one of the most popular algorithms for deep learning, a type of machine learning in which a model learns to perform classification tasks directly from images, video, text, or sound.

Stanford dl excercise

This toolbox offers convolution neural networks (CNN) using k-fold cross-validation, which are simple and easy to implement.

Neural Network implemented with Matlab.

University Project for "Intelligent Systems" course (MSc Computer Engineering @ University of Pisa). Design and implementation of several artificial intelligences (MLP, RBFN, FIS, CNN, RNN) on a dataset composed on biophysical signals