MRF as regularization on matching surfaces

The inferred data-model associations from the forests often look noisy (left image) because it does not respect the triangles/edge connectivity. In this project, we mitigate this by proposing a new way to aggregate leaf-node predictions from each tree. As the correspondences are highly interdependent among the local neighborhood, we apply random fields (MRF/CRF) to regularize them, yield the optimal configuration w.r.t. the whole shape (right image). The aggregation is therefore mesh-based rather than vertex-based.

#randomforest #markovrandomfield