Deformable 3D Shape Matching with Topological Noise
This dataset consists of a collection of 3D shapes undergoing within-class deformations that include in topological changes. The changes simulate coalescence of spatially close surface regions – a scenario that frequently occurs when dealing with real data under suboptimal acquisition conditions. The dataset is based on the fat kid from the KIDS dataset with additional poses. All shapes are given in OFF format and exist in two different resolutions (~10k and ~60-80k).
Ground-truth matches to the null shape are given for all shapes, but due to the changes in topology not all vertices have a ground-truth match. A map indicating which vertices do not have a match is provided. We also provide a ground-truth symmetry correspondence for all shapes.
If you use the dataset, please cite the following paper:
|Conference and Workshop Papers|
|SHREC’16: Matching of Deformable Shapes with Topological Noise , In Proc. of Eurographics Workshop on 3D Object Retrieval (3DOR), 2016. (to appear) [bib] [pdf]|
- All shapes were created with DAZ Studio and the method from Campen M., Kobbelt L.: Exact and robust (self-)intersections for polygonal meshes. Comput. Graph. Forum 29, 2 (2010).
- A track of the SHREC'16 Contest was based on this dataset, see the contest homepage for code to compare your results to participants of the contest.