We started off saying that capsule tries to detect the face in the image. So, we need to convert into probabilities to get the probability of the existence of a face in the image.

Apart from calculating probabilities, we also need to preserve the direction of the vectors, so we use an activation function called the squash function. It is given as follows:

Now, (also referred to as the activity vector) gives us the probability of the existence of a face in a given image.