Summary

This chapter introduced style transfer, which is a popular task nowadays that can be performed using CNNs. It consists of taking both a content image and a style image as inputs and returning a newly created image as output that keeps the content of one of the images and the style of the other. It is typically used to give images an artistic look by combining random regular images with those of the paintings of great artists.

Although style transfer is performed using CNNs, the process of creating the target image is not achieved by training the network conventionally. This chapter explained how to use pre-trained networks to consider the output of some relevant layers that are especially good at identifying certain features.

This chapter explained each of the steps required to develop code that's capable of performing the task of style transfer, where the first step consisted of loading and displaying the inputs. As we mentioned earlier, there are two inputs...