NilDa is a small C++ implementation of different deep learning algorithms for machine learning applications.
The code leverage the Eigen library for efficient algebra operations and openCV for image/video handling. The code started as a small personal project and its objective is not to compete with other famous codes in the artificial intelligence world but, rather, it wants to offer a less intimidating experience for people that want to know how machine learning algorithms work under the hood.
Dense
2D Convolution with padding
MaxPooling
BatchNormalization
Dropout
ReLU
Sigmoid
Softmax
Tanh
Stochastic and mini-batch gradient descent with momentum
AdaGrad
RSMProp
Adam