a new feature that could help retrieve blur image

Most of us capture pictures that could be of random i.e, accidental which could be a better picture than compared and it could be very hard to make it better to use w.r.to the present AI tools which could only make a rough image which couldn't generate an accurate face. 

Here, for the Retrieved image, if we could give our original face image as a secondary input can give a better image.