Decaf: A deep convolutional activation feature for generic visual recognition. (2014)

by J Donahue, Y Jia, O Vinyals, J Hoffman, N Zhang, E Tzeng, T Darrell
Venue:In International Conference on Machine Learning (ICML),