You can see why I changed the code and some experiments I did in this page:
https://github.com/B-C-WANG/AI.Experiment/tree/master/AI.Exp.LayerNeuralStylize
We often use one style image to calculate its output from beginning to relu1-1,relu2-1,relu3-1,relu4-1 and relu5-1 in VGG model. Here I split them, means that you can feed with 1 to 5 images to individually calculate its relu1-1, relu2-1, relu3-1, relu4-1 and relu5-1.
You can see why I changed the code and some experiments I did in this page: https://github.com/B-C-WANG/AI.Experiment/tree/master/AI.Exp.LayerNeuralStylize We often use one style image to calculate its output from beginning to relu1-1,relu2-1,relu3-1,relu4-1 and relu5-1 in VGG model. Here I split them, means that you can feed with 1 to 5 images to individually calculate its relu1-1, relu2-1, relu3-1, relu4-1 and relu5-1.