Now that we are in a position to generate faces, let's go ahead and perform some vector arithmetic on top of the generated images.
For this exercise, we will perform the transition of face generation from one face to another.
Now that we are in a position to generate faces, let's go ahead and perform some vector arithmetic on top of the generated images.
For this exercise, we will perform the transition of face generation from one face to another.
We'll continue from the image generation model that we built in the Face generation using a Deep Convolutional GAN section.
Let's say we want to see the transition of one generated face image into another generated face image. This process is enabled by slowly varying the vector from the first vector (the vector of the first generated image) to the second vector (the vector of the second generated image). You can essentially think of each...