This work is based on Zhifei Zhang's paper Derivation of Backpropagation in Convolutional Neural Network (see the paper for more details).
The file cnn.sac
implements the various network components to construct a CNN.
In zhang.sac
we implementat a CNN that evaluates the MNIST/EMNIST datasets. We read in MNIST/EMNIST
datasets using functions defined in mnist.sac
.
To compile, simply call make
which will create binaries for x64 sequential and multi-threaded.