Boundary Correspondence for Iso-Geometric Analysis Based on Deep Learning

Abstract

One of the key problems in isogeometric analysis(IGA) is domain parameterization, i.e., constructing a map between a parametric domain and a computational domain. As a preliminary step of domain parameterization, the mapping between the boundaries of the parametric domain and the computational domain should be established. The boundary correspondence strongly affects the quality of domain parameterization and thus subsequent numerical analysis. Currently, boundary correspondence is generally determined manually and only one approach based on optimal mass transport discusses automatic generation of boundary correspondence. In this article, we propose a deep neural network based approach to generate boundary correspondence for 2D simply connected computational domains. Given the boundary polygon of a planar computational domain, the main problem is to pick four corner vertices on the input boundary in order to subdivide the boundary into four segments which correspond to the four sides of the parametric domain. We synthesize a dataset with corner correspondence and train a fully convolutional network to predict the likelihood of each boundary vertex to be one of the corner vertices, and thus to locate four corner vertices with locally maximum likelihood. We evaluate our method on two types of datasets: MPEG-7 dataset and CAD model dataset. The experiment results demonstrate that our algorithm is faster by several orders of magnitude, and at the same time achieves smaller average angular distortion, more uniform area distortion and higher success rate, compared to the traditional optimization-based method. Furthermore, our neural network exhibits good generalization ability on new datasets.

Keywords

Isogeometric analysis / Domain parameterization / Boundary correspondence / Deep neural network

References

[1] Azulay A, Weiss Y. Why do deep convolutional networks generalize so poorly to small image transformations?. J. Mach. Learn. Res.. 2019, 20 184 1-25

[2] Blidia A, Mourrain B, Xu G. Geometrically smooth spline bases for data fitting and simulation. Computer Aided Geometric Des.. 2020, 78 87

[3] Bouritsas, G., Bokhnyak, S., Ploumpis, S., Bronstein, M., Zafeiriou, S.: Neural 3d morphable models: Spiral convolutional networks for 3d shape representation learning and generation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 7213–7222 (2019)

[4] Buchegger F, Jüttler B. Planar multi-patch domain parameterization via patch adjacency graphs. Computer-Aided Des.. 2017, 82 2-12

[5] Chen L, Xu G, Wang S, Shi Z, Huang J. Constructing volumetric parameterization based on directed graph simplification of $\mathcalligra{l}$1 polycube structure from complex shapes. Computer Methods Appl. Mech. Eng.. 2019, 351 422-440

[6] Cohen E, Martin T, Kirby RM, Lyche T, Riesenfeld RF. Analysis-aware modeling: understanding quality considerations in modeling for isogeometric analysis. Computer Methods Appl. Mech. Eng.. 2010, 199 5 334-356

[7] Dai, A., Ruizhongtai Qi, C., Nießner, M.: Shape completion using 3d-encoder-predictor cnns and shape synthesis. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5868–5877 (2017)

[8] Defferrard, M., Bresson, X., Vandergheynst, P.: Convolutional neural networks on graphs with fast localized spectral filtering. In: Proceedings of the 30th International Conference on Neural Information Processing Systems, pp. 3844–3852. Curran Associates Inc. (2016)

[9] Farin G, Hansford D. Discrete coons patches. Computer Aided Geometric Des.. 1999, 16 7 691-700

[10] Fey, M., Lenssen, J.E., Weichert, F., Müller, H.: Splinecnn: Fast geometric deep learning with continuous b-spline kernels. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 869–877 (2018).

[11] Gao, Z., Yan, J., Zhai, G., Zhang, J., Yang, Y., Yang, X.: Learning Local Neighboring Structure for Robust 3D Shape Representation (2020)

[12] Gong, S., Chen, L., Bronstein, M., Zafeiriou, S.: Spiralnet++: A fast and highly efficient mesh convolution operator. 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW), 4141–4148 (2019).

[13] Gravesen, J., Evgrafov, A., Nguyen, M., NØrtoft, P.: Planar parametrization in isogeometric analysis. In: Mathematical Methods for Curves and Surfaces, pp. 189–212. Springer (2014)

[14] Hanocka R, Hertz A, Fish N, Giryes R, Fleishman S, Cohen-Or D. MeshCNN: a network with an edge. ACM Trans. Graph.. 2019, 38 4 90-19012

[15] He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770–778. IEEE, Las Vegas, NV, USA (2016).

[16] Hinz J, Möller M, Vuik C. Elliptic grid generation techniques in the framework of isogeometric analysis applications. Computer Aided Geometric Des.. 2018, 65 48-75

[17] Hu K, Zhang YJ. Centroidal Voronoi tessellation based polycube construction for adaptive all-hexahedral mesh generation. Computer Methods Appl. Mech. Eng.. 2016, 305 405-421

[18] Hu K, Zhang YJ, Liao T. Surface segmentation for polycube construction based on generalized centroidal Voronoi tessellation. Computer Methods Appl. Mech. Eng.. 2017, 316 280-296

[19] Hughes TJR, Cottrell JA, Bazilevs Y. Isogeometric analysis: Cad, finite elements, nurbs, exact geometry and mesh refinement. Computer Methods Appl. Mech. Eng.. 2005, 194 39 4135-4195

[20] Ji Y, Yu Y-Y, Wang M-Y, Zhu C-G. Constructing high-quality planar NURBS parameterization for isogeometric analysis by adjustment control points and weights. J. Comput. Appl. Math.. 2021, 396

[21] Kapl M, Sangalli G, Takacs T. Dimension and basis construction for analysis-suitable G(1) two-patch parameterizations. Computer Aided Geometric Des.. 2017, 52–53 75-89

[22] Kapl M, Buchegger F, Bercovier M, Juttler B. Isogeometric analysis with geometrically continuous functions on planar multi-patch geometries. Computer Methods Appl. Mech. Eng.. 2017, 316 209-234

[23] Kapl M, Sangalli G, Takacs T. Construction of analysis-suitable G(1) planar multi-patch parameterizations. Computer Aided Des.. 2018, 97 41-55

[24] Kapl M, Sangalli G, Takacs T. An isogeometric C-1 subspace on unstructured multi-patch planar domains. Computer Aided Geometric Des.. 2019, 69 55-75

[25] Litany, O., Remez, T., Rodolá, E., Bronstein, A., Bronstein, M.: Deep functional maps: Structured prediction for dense shape correspondence. In: 2017 IEEE International Conference on Computer Vision (ICCV), pp. 5660–5668 (2017).

[26] Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, van der Laak JAWM, van Ginneken B, Sánchez CI. A survey on deep learning in medical image analysis. Med. Image Anal.. 2017, 42 60-88

[27] Liu L, Zhang Y, Hughes T, Scott M, Sederberg T. Volumetric T-spline construction using Boolean operations. Eng. Computers. 2014, 30 4 425-439

[28] Liu L, Zhang Y, Liu Y, Wang W. Feature-preserving T-mesh construction using skeleton-based polycubes. Computer-Aided Des.. 2015, 58 162-172

[29] Liu H, Yang Y, Liu Y, Fu X-M. Simultaneous interior and boundary optimization of volumetric domain parameterizations for iga. Computer Aided Geometric Des.. 2020, 79

[30] Liu L, Ouyang W, Wang X, Fieguth P, Chen J, Liu X, Pietikäinen M. Deep learning for generic object detection: a survey. Int. J. Computer Vision. 2020, 128 87

[31] Long, J., Shelhamer, E., Darrell, T.: Fully convolutional networks for semantic segmentation. In: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3431–3440 (2015).

[32] Mantzaflaris A, Jüttler B, Khoromskij BN, Langer U. Low rank tensor methods in Galerkin-based isogeometric analysis. Computer Methods Appl. Mech. Eng.. 2017, 316 1062-1085

[33] Minaee S, Boykov YY, Porikli F, Plaza AJ, Kehtarnavaz N, Terzopoulos D. Image segmentation using deep learning: a survey. IEEE Trans. Pattern Anal. Mach. Intell.. 2021, 1 1

[34] Minaee S, Kalchbrenner N, Cambria E, Nikzad N, Chenaghlu M, Gao J. Deep learning-based text classification: a comprehensive review. ACM Comput. Surv.. 2021, 54 3

[35] Monti, F., Boscaini, D., Masci, J., Rodolá, E., Svoboda, J., Bronstein, M.M.: Geometric deep learning on graphs and manifolds using mixture model cnns. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5425–5434 (2017).

[36] Nguyen, T., Jüttler, B.: Parameterization of contractible domains using sequences of harmonic maps. In: Curves and Surfaces. Lecture Notes in Computer Science, pp. 501–514. Springer, Berlin, Heidelberg (2010).

[37] Nian X, Chen F. Planar domain parameterization for isogeometric analysis based on teichmüller mapping. Computer Methods Appl. Mech. Eng.. 2016, 311 41-55

[38] Pan M, Chen F, Tong W. Low-rank parameterization of planar domains for isogeometric analysis. Computer Aided Geometric Des.. 2018, 63 1-16

[39] Pauley, M., Nguyen, D.-M., Mayer, D., Špeh, J., Weeger, O., Jüttler, B.: The Isogeometric Segmentation Pipeline. In: Jüttler, B., Simeon, B. (eds.) Isogeometric Analysis and Applications 2014. Lecture Notes in Computational Science and Engineering, pp. 51–72. Springer International Publishing, Cham (2015).

[40] Pilgerstorfer E, Juttler B. Bounding the influence of domain parameterization and knot spacing on numerical stability in Isogeometric Analysis. Computer Methods Appl. Mech. Eng.. 2014, 268 589-613

[41] Qi CR, Yi L, Su H, Guibas LJ. Guyon I, Luxburg UV, Bengio S, Wallach H, Fergus R, Vishwanathan S, Garnett R. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. Advances in neural information processing systems. 2017 UK: Curran Associates Inc

[42] Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation (2015)

[43] Sahillioğlu Y. Recent advances in shape correspondence. Visual Computer. 2020, 36 871

[44] Scholz F, Jüttler B. Parameterization for polynomial curve approximation via residual deep neural networks. Computer Aided Geometric Des.. 2021, 85

[45] Speleers H, Manni C. Optimizing domain parameterization in isogeometric analysis based on Powell-Sabin splines. J. Comput. Appl. Math.. 2015, 289 68-86

[46] van Kaick O, Zhang H, Hamarneh G, Cohen-Or D. A survey on shape correspondence. Computer Gr. Forum. 2011, 30 6 1681-1707

[47] Veličković, P., Cucurull, G., Casanova, A., Romero, A., Lio, P., Bengio, Y.: Graph Attention Networks (2017)

[48] Wang, W., Zhang, Y., Liu, L., Hughes, T.J.R.: Trivariate solid t-spline construction from boundary triangulations with arbitrary genus topology. Computer-Aided Design 45(2), 351–360 (2013). Solid and Physical Modeling 2012

[49] Wang S. Iga-suitable planar parameterization with patch structure simplification of closed-form polysquare. Computer Methods Appl. Mech. Eng.. 2022, 392

[50] Wang Y, Sun Y, Liu Z, Sarma SE, Bronstein MM, Solomon JM. Dynamic graph CNN for learning on point clouds. ACM Trans. Gr.. 2019, 38 5 146-114612

[51] Wang D, Xu J, Gao F, Wang CCL, Gu R, Lin F, Rabczuk T, Xu G. Iga-reuse-net: a deep-learning-based isogeometric analysis-reuse approach with topology-consistent parameterizationimage 1. Computer Aided Geometric Des.. 2022, 95

[52] Wen, C., Zhang, Y., Li, Z., Fu, Y.: Pixel2mesh++: Multi-view 3d mesh generation via deformation. In: 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pp. 1042–1051 (2019).

[53] Xiao S, Kang H, Fu X-M, Chen F. Computing IGA-suitable planar parameterizations by PolySquare-enhanced domain partition. Computer Aided Geometric Des.. 2018, 62 29-43

[54] Xu G, Mourrain B, Duvigneau R, Galligo A. Parameterization of computational domain in isogeometric analysis: methods and comparison. Computer Methods Appl. Mech. Eng.. 2011, 200 23 2021-2031

[55] Xu G, Mourrain B, Duvigneau R, Galligo A. Constructing analysis-suitable parameterization of computational domain from CAD boundary by variational harmonic method. J. Comput. Phys.. 2013, 252 275-289

[56] Xu, G., Mourrain, B., Galligo, A., Rabczuk, T.: High-quality construction of analysis-suitable trivariate nurbs solids by reparameterization methods. Comput. Mech. (2014).

[57] Xu J, Chen F, Deng J. Two-dimensional domain decomposition based on skeleton computation for parameterization and isogeometric analysis. Computer Methods Appl. Mech. Eng.. 2015, 284 541-555

[58] Xu G, Li M, Mourrain B, Rabczuk T, Xu J, Bordas SPA. Constructing IGA-suitable planar parameterization from complex CAD boundary by domain partition and global/local optimization. Computer Methods Appl. Mech. Eng.. 2018, 328 175-200

[59] Xu G, Li B, Shu L, Chen L, Xu J, Khajah T. Efficient r-adaptive isogeometric analysis with winslow’s mapping and monitor function approach. J. Comput. Appl. Math.. 2019, 351 186-197

[60] Zhang Y, Bazilevs Y, Goswami S, Bajaj CL, Hughes TJR. Patient-specific vascular NURBS modeling for isogeometric analysis of blood flow. Computer Methods Appl. Mech. Eng.. 2007, 196 29 2943-2959

[61] Zheng Y, Chen F. Volumetric Boundary Correspondence for Isogeometric Analysis Based on Unbalanced Optimal Transport. Computer-Aided Des.. 2021, 140

[62] Zheng Y, Pan M, Chen F. Boundary correspondence of planar domains for isogeometric analysis based on optimal mass transport. Computer-Aided Des.. 2019, 114 28-36