Problem In research paper?

[thakkarvishal226]

Hello i read the research paper of DCTON

and i confuse in some part like in MPN network we genrate ( M1_cloth and M1_skin ) that means we genrate mask of input image cloth but the image of DCTON netwrok show M1_cloth as output Cloth Mask

if anyone clear this doubt will help me alot.

and also can anyone explain why We send C2 ( targeted Cloth ) in MPN we just generate C1_mask from D ( human Representation ) so why we send C2 In MPN Network