Question about supervision for training on ClothesSeq data.

[GostInShell]

Hi, thanks for releasing your code!

If i get the following codes right. In the first stage, the weights for both SMPL and clothed human meshes are predicted.
My question is how the gt weight for the clothed human (skin_bp) is calculated.

neuralgif/models/train_shape.py

Lines 138 to 139 in 5e60256

	weight_pred = self.model_wgt(pts, body_enc_feat, pose_in)
	weight_smpl = self.model_wgt(smpl_vert, body_enc_feat_smpl, pose_in)

neuralgif/models/train_shape.py

Lines 150 to 151 in 5e60256

	w1 = self.loss_l1(weight_smpl, gt_skin)
	w2 = self.loss_l1(weight_pred, skin_bp)

Moreover, here the clothed human in the canonical pose (can_pts_gt) is also used for supervision.

neuralgif/models/train_shape.py

Lines 148 to 149 in 5e60256

	diff_can_bp = can_pt_bp - can_pts_gt
	diff_can_bp = torch.sqrt(torch.sum(diff_can_bp*diff_can_bp, dim=2)).mean()

I can understand these variables for a SMPL mesh can be obtaiend. But how to obtain these parameters for a clothed human mesh?