Local bending coeffient Failed. #29
Description
@Dolorousrtur
Thank you for sharing this great work with us!
My question is : in order to reproduce Fig.10(B), i.e. different materials for different parts of a garment, do we need to retrain the model ?
When i tried to reproduce Fig.10(B) with the pretrained cvpr model(postcvpr), it seems that change of bending coeffient in local area influence all garment verts movement. NO change in edge bending coeffienct feature also failed to reproduce local influence.
I noticed that during training, all verts and edges assigned to same bending coeffient, it seems strange that local bending coeffient during inference can work.
Could you give me some suggestions ? Thansk a lot !!