Two-Layer Sparse Compression of Dense-Weight Blend Skinning

Binh Huy Le and Zhigang Deng

ACM Transactions on Graphics 32(4) - Proceedings of ACM SIGGRAPH 2013

Our introduced compression model blends master bone transformations and caches them as virtual bone transformations (left most). This model can compress the Linear Blend Skinning (LBS) model with dense weights and generate a fast and compact skinning model with parallel implementations on multi-core CPUs or GPUs. In this example, the performance of our compressed model is 20% faster than that of the de facto standard LBS model with 4 bones per vertex, and 5 times faster than that of dense-weight LBS model.

Abstract: Weighted linear interpolation has been widely used in many skinning techniques including linear blend skinning, dual quaternion blend skinning, and cage based deformation. To speed up performance, these skinning models typically employ a sparseness constraint, in which each 3D model vertex has a small fixed number of non-zero weights. However, the sparseness constraint also imposes certain limitations to skinning models and their various applications. This paper introduces an efficient two-layer sparse compression technique to substantially reduce the computational cost of a dense-weight skinning model, with insignificant loss of its visual quality. It can directly work on dense skinning weights or use example-based skinning decomposition to further improve its accuracy. Experiments and comparisons demonstrate that the introduced sparse compression model can significantly outperform state of the art weight reduction algorithms, as well as skinning decomposition algorithms with a sparseness constraint.

Download: [paper] [video]

Video on Youtube

Image

Representative image

Bibtex

@article{BinhLe:TOG:2013,
author = {Le, Binh and Deng, Zhigang},
title = {Two-layer Sparse Compression of Dense-Weight Blend Skinning},
journal = {ACM Trans. Graph.},
issue_date = {July 2013},
volume = {32},
number = {4},
month = jul,
year = {2013},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {skinning, linear blend skinning, skinning from examples, sparse coding, dictionary learning},
note = "to appear"
}

@inproceedings{BinhLe:S:2013,
author = {Le, Binh and Deng, Zhigang},
title = {Two-layer Sparse Compression of Dense-Weight Blend Skinning},
booktitle = {Proceedings of the 2013 SIGGRAPH Conference},
series = {SIGGRAPH '13},
year = {2013},
location = {Anaheim, CA, USA},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {skinning, linear blend skinning, skinning from examples, sparse coding, dictionary learning},
note = "to appear"
}