Caffe L2 Normalization Using Existing Layers

Question

I am trying to perform L2 normalization in Caffe for a layer. The idea is to use these L2-normalized fc7 functions in contrast loss, for example http://www.cs.cornell.edu/~kb/publications/SIG15ProductNet.pdf .

I could find some links where people posted code for the L2 normalization level there. However, I was wondering if it is possible to do this using the normalization level of the local Caffe response or, possibly, any other.

I have a final vector fc 1x2048 (2048 channels of 1x1 size). Can anyone advise me on this?