in the like where you sample shape parameter, you add to the mean shape. I am not quite sure if this is right. Since If the goal is to create a new shape then we should add mean shape to the multiplication of beta parameters to PCA vectors. Am I missing something?
Hello,
in the like where you sample shape parameter, you add to the mean shape. I am not quite sure if this is right. Since If the goal is to create a new shape then we should add mean shape to the multiplication of beta parameters to PCA vectors. Am I missing something?
https://github.com/akashsengupta1997/STRAPS-3DHumanShapePose/blob/0983321e631d0bbd2586601e06abe714b53c6790/augmentation/smpl_augmentation.py#L13