(a) This problem is long known as color/contrast transfer, and it was solved > 10yrs ago, (b) the results shown in this paper aren't objectively or subjectively

(a) This problem is long known as color/contrast transfer, and it was solved > 10yrs ago, (b) the results shown in this paper aren't objectively or subjectively better/more photo-realistic than Kokaram etc.'s work; and (c) I question whether this task even requires deep learning at all. https://francois.pitie.net/colour/ https://francois.pitie.net/colour/