"DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition."

Jeff Donahue et al. (2013)
a service of  Schloss Dagstuhl - Leibniz Center for Informatics