If you have a massive polygonal monster (err literally) to display that's a lot of vertices to process, on top of texture mapping, normal mapping, light mapping, per pixel shading etc etc yawn.
This vertex grouping reduces the number of vertices to schedule, speeding up the scheduling process.
And the Laplacian term in Equation 29 constraints uncontrolled vertices to their original shape.
Consequently, the number of vertices to schedule is larger than in the original graph.
So, we aim to distribute the vertices to different voxels as much as possible.
The typical number of vertices to schedule in PREESM is between one hundred and several thousands.
In each iteration, we identify a set of valid vertices to be explored according Observation 1.
However, this may lead to a wrong solution that assigns all vertices to +1 or −1.
Therefore, we consider there are two or three fixed vertices to find the restricted Steiner tree problem.
Path-based algorithms usually perform traversals in different directions to build a large ranking of vertices to recommend, whereas random walk-based algorithms build an initial subgraph and perform several iterations on those vertices to compute the final ranking.
In this part, we proposed a local-mean diffusion operator, which is used to assign the reconstructed residual from the sampled vertices to their adjacent unsampled vertices, to achieve the local-mean diffusion operation.
