Skip to content

Cut VGG19

class Cut_VGG19

Class object that fetches keras' VGG19 model trained on the imagenet dataset and declares as output layers. Used as feature extractor for the perceptual loss function.

  • layers_to_extract: list of layers to be declared as output layers.

  • patch_size: integer, defines the size of the input (patch_size x patch_size).

  • loss_model: multi-output vgg architecture with as output layers.


def __init__(patch_size, layers_to_extract)