Hi ,
I have two question:
1. What are the advantages of using Local Contrast Normalization and Local Response Normalization ?
2. Would you explain further the meaning of the sentences in bold ?
Besides , for LRN , I found this on the paper of AlexNet , that LRN it : creates competition for big activities amongst neuron outputs computed using different kernels . For LCN , I found on INTERNET that it enforces a sort of local competition between adjacent features in a feature map
and between features at the same spatial location in different maps .
Thanks