This will take any image at any resolution and recreate it. This is not a redux, this takes the image, reads it and describes the image. Then generates a whole new image with key elements of that description.
The text field as been added in case you want to add elements or change something such as hair color. This is optional, if you don't want to change anything leave it "empty" (Leave the word there)
For example, if you take the example image on the APP and enter "She has red hair" it will re-imagine the image but give her red hair. This can also work with environments that are not present, for example "There is a snow capped mountain in the background" it'll add a snow capped mountain, but because there are trees and mist in the original image, it'll have troubles that. Only because it's an element in the original image. It may have troubles with some elements that are too predominant or blend two elements. Sometimes hair color will be mutli-colored if two hair colors are present.
