Computer Vision (CV) is the field of Machine Learning in which visual data is processed (images, videos) and labeled.
Classifying dog breeds is CV
Models that attempt to solve CV problems usually present the following architecture:
Typical CV model architecture
A series of convolutions and poolings are done one the input image to detect features, edges, shapes and to reduce the dimensionality. This greatly simplified object is then fed into a fully connected layer that attempts to classify it.