skinning is a notion in graphics which solves Cloth Fitting Task which adds a prior S\left(\theta\right) which is not learned to make the learning of D\left(\theta\right) easier
\begin{equation} \phi = f_{}\left(\theta\right) = S \left(\theta\right) + D \left(\theta\right) \end{equation}
so then, when we make a model, keeping around S\left(\theta\right) which is learned by physical dynamics, and constraining a displacement map D\left(\theta\right) within each square S\left(\theta\right). We can then rasterize the diffs into RGB space into a grid, then we can run a CNN.