DCN_AlexNet
This model implements a pre-trained deep convolutional neural network
(CNN) known as AlexNet for image classification (Krizhevsky et al.,
2010). It is designed to approximate the version of AlexNet available
from the Berkeley Artificial Intelligence Research Lab (Caffe
website). Some units within this CNN have been found to be
similar to V4 neurons in terms of their tuning for contour curvature and
translation invariance for a set of simple closed shapes (Pospisil et
al.).