Region-based Correspondence Between 3D Shapes via Spatially Smooth Biclustering

Matteo Denitto (1) , Simone Melzi (1) , Manuele Bicego (1) , Umberto Castellani (1) , Alessandro Farinelli (1) , Mário a T Figueiredo (2) , Yanir Kleiman (3) , Maks Ovsjanikov (3)

1 UNIVR | DI - Department of Computer Science [Verona]
2 IST - Instituto Superior Técnico, Universidade Técnica de Lisboa
3 LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau]

Region-based correspondence (RBC) is a highly relevant and non-trivial computer vision problem. Given two 3D shapes, RBC seeks segments/regions on these shapes that can be reliably put in correspondence. The problem thus consists both in finding the regions and determining the correspondences between them. This problem statement is similar to that of " biclustering " , implying that RBC can be cast as a biclustering problem. Here, we exploit this implication by tackling RBC via a novel biclustering approach, called S 4 B (spatially smooth spike and slab biclustering), which: (i) casts the problem in a probabilistic low-rank matrix fac-torization perspective; (ii) uses a spike and slab prior to induce sparsity; (iii) is enriched with a spatial smoothness prior, based on geodesic distances, encouraging nearby vertices to belong to the same bicluster. This type of spatial prior cannot be used in classical biclustering techniques. We test the proposed approach on the FAUST dataset, out-performing both state-of-the-art RBC techniques and classical biclustering methods.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Matteo Denitto, Simone Melzi, Manuele Bicego, Umberto Castellani, Alessandro Farinelli, et al.. Region-based Correspondence Between 3D Shapes via Spatially Smooth Biclustering. International Conference on Computer Vision, Oct 2017, Venice, Italy. ⟨hal-01741765⟩