×

A hybrid inference system for improved curvature estimation in the level-set method using machine learning. (English) Zbl 07536781

Summary: We present a novel hybrid strategy based on machine learning to improve curvature estimation in the level-set method. The proposed inference system couples enhanced neural networks with standard numerical schemes to compute curvature more accurately. The core of our hybrid framework is a switching mechanism that relies on well established numerical techniques to gauge curvature. If the curvature magnitude is larger than a resolution-dependent threshold, it uses a neural network to yield a better approximation. Our networks are multilayer perceptrons fitted to synthetic data sets composed of sinusoidal- and circular-interface samples at various configurations. To reduce data set size and training complexity, we leverage the problem’s characteristic symmetry and build our models on just half of the curvature spectrum. These savings lead to a powerful inference system able to outperform any of its numerical or neural component alone. Experiments with stationary, smooth interfaces show that our hybrid solver is notably superior to conventional numerical methods in coarse grids and along steep interface regions. Compared to prior research, we have observed outstanding gains in precision after training the regression model with data pairs from more than a single interface type and transforming data with specialized input preprocessing. In particular, our findings confirm that machine learning is a promising venue for reducing or removing mass loss in the level-set method.

MSC:

65Mxx Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems
68Txx Artificial intelligence
68-XX Computer science
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] Friedman, A., Variational Principles of Free-Boundary Problems (2010), Dover Publications
[2] Popinet, S., Numerical models of surface tension, Annu. Rev. Fluid Mech., 50, 1, 49-75 (January 2018) · Zbl 1384.76016
[3] Sussman, M.; Smereka, P.; Osher, S., A level set approach for computing solutions to incompressible two-phase flow, J. Comput. Phys., 114, 1, 146-159 (September 1994) · Zbl 0808.76077
[4] Sussman, M.; Fatemi, E.; Smereka, P.; Osher, S., An improved level set method for incompressible two-phase flows, Comput. Fluids, 27, 5-6, 663-680 (June 1998) · Zbl 0967.76078
[5] Gibou, F.; Chen, L.; Nguyen, D.; Banerjee, S., A level set based sharp interface method for the multiphase incompressible Navier-Stokes equations with phase change, J. Comput. Phys., 222, 2, 536-555 (March 2007) · Zbl 1158.76403
[6] Theillard, M.; Gibou, F.; Saintillan, D., Sharp numerical simulation of incompressible two-phase flows, J. Comput. Phys. (2019) · Zbl 1452.76135
[7] Losasso, F.; Gibou, F.; Fedkiw, R., Simulating water and smoke with an octree data structure, ACM Trans. Graph., 23, 3, 457-462 (August 2004)
[8] Losasso, F.; Shinar, T.; Selle, A.; Fedkiw, R., Multiple interacting liquids, ACM Trans. Graph., 25, 3, 812-819 (July 2006)
[9] Gibou, F.; Hyde, D.; Fedkiw, R., Sharp interface approaches and deep learning techniques for multiphase flows, J. Comput. Phys., 380, 442-463 (2019) · Zbl 1451.76131
[10] Chan, T.; Vese, L., Active contour without edges, IEEE Trans. Image Process., 10, 266-277 (2001) · Zbl 1039.68779
[11] Gibou, F.; Fedkiw, R., A fast hybrid k-means level set algorithm for segmentation, (Proceedings of the 4th Annual Hawaii International Conference on Statistics and Mathematics. Proceedings of the 4th Annual Hawaii International Conference on Statistics and Mathematics, Honolulu, US (2005)), 281-291
[12] Theillard, M.; Gibou, F.; Pollock, T., A sharp computational method for the simulation of the solidification of binary alloys, J. Sci. Comput., 63, 330-354 (2015) · Zbl 1426.76572
[13] Papac, J.; Helgadottir, A.; Ratsch, C.; Gibou, F., A level set approach for diffusion and Stefan-type problems with Robin boundary conditions on quadtree/octree adaptive Cartesian grids, J. Comput. Phys., 233, 241 (2013)
[14] Papac, J.; Gibou, F.; Ratsch, C., Efficient symmetric discretization for the Poisson, heat and Stefan-type problems with Robin boundary conditions, J. Comput. Phys., 229, 3, 875-889 (2010) · Zbl 1182.65140
[15] Chen, H.; Min, C.; Gibou, F., A numerical scheme for the Stefan problem on adaptive Cartesian grids with supralinear convergence rate, J. Comput. Phys., 228, 16, 5803-5818 (September 2009) · Zbl 1176.80059
[16] Mirzadeh, M.; Gibou, F., A conservative discretization of the Poisson-Nernst-Planck equations on adaptive Cartesian grids, J. Comput. Phys., 274, 633-653 (2014) · Zbl 1351.82082
[17] Alias, M. A.; Buenzli, P. R., A level-set method for the evolution of cells and tissue during curvature-controlled growth, Int. J. Numer. Methods Biomed. Eng., 36, 1, Article e3279 pp. (January 2020)
[18] Osher, S.; Sethian, J. A., Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations, J. Comput. Phys., 79, 1, 12-49 (November 1988) · Zbl 0659.65132
[19] Sethian, J. A., Level Set Methods and Fast Marching Methods, Cambridge Monogr. Appl. Comput. Math. (1999), Cambridge University Press: Cambridge University Press Cambridge, UK · Zbl 0929.65066
[20] Gibou, F.; Fedkiw, R.; Osher, S., A review of level-set methods and some recent applications, J. Comput. Phys., 353, 82-109 (January 2018) · Zbl 1380.65196
[21] Hirt, C. W.; Nichols, B. D., Volume of fluid (VOF) method for the dynamics of free boundaries, J. Comput. Phys., 39, 201-225 (1981) · Zbl 0462.76020
[22] (Fasano, A.; Primicerio, M., Free Boundary Problems: Theory and Applications (1983), Pitman: Pitman Boston) · Zbl 0511.35088
[23] Langer, J. S., Models of Pattern Formation in First-Order Phase Transitions (1986), World Scientific, Number 165-186
[24] Larios-Cárdenas, L.Á.; Gibou, F., A deep learning approach for the computation of curvature in the level-set method, SIAM J. Sci. Comput., 43, 3, A1754-A1779 (January 2021) · Zbl 1512.65293
[25] du Chéné, A.; Min, C.; Gibou, F., Second-order accurate computation of curvatures in a level set framework using novel high-order reinitialization schemes, J. Sci. Comput., 35, 114-131 (June 2008) · Zbl 1203.65043
[26] Zhang, Q.; Fogelson, A., Fourth-order interface tracking in two dimensions via an improved polygonal area mapping method, SIAM J. Sci. Comput., 36, 5, A2369-A2400 (October 2014) · Zbl 1426.76584
[27] Zhang, Q.; Fogelson, A., MARS: an analytic framework of interface tracking via mapping and adjusting regular semialgebraic sets, SIAM J. Numer. Anal., 54, 2, 530-560 (March 2016) · Zbl 1332.76066
[28] Zhang, Q., HFES: a height function method with explicit input and signed output for high-order estimations of curvature and unit vectors of planar curves, SIAM J. Numer. Anal., 55, 2, 1024-1056 (April 2017) · Zbl 1362.65028
[29] Osher, S.; Fedkiw, R., Level Set Methods and Dynamic Implicit Surfaces, Appl. Math. Sci., vol. 153 (2002), Springer: Springer Cham
[30] Aggarwal, C. C., Neural Networks and Deep Learning - A Textbook (2018), Springer: Springer Cham · Zbl 1402.68001
[31] Mehta, P.; Bukov, M.; Wang, C.; Day, A. G.R.; Richardson, C.; Fisher, C. K.; Schwabd, D. J., A high-bias, low-variance introduction to machine learning for physicists, Phys. Rep., 810, 1-124 (May 2019)
[32] Gerón, A., Hands-on Machine Learning with Scikit-Learn, Keras & TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems (November 2019), O’Reilly
[33] Ngo, T. A.; Lu, Z.; Carneiro, G., Combining deep learning and level set for the automated segmentation of the left ventricle of the heart from cardiac cine magnetic resonance, Med. Image Anal., 35, 159-171 (January 2017)
[34] Le, T. H.N.; Quach, K. G.; Luu, K.; Duong, C. N.; Savvides, M., Reformulating level sets as deep recurrent neural network approach to semantic segmentation, IEEE Trans. Image Process., 27, 5, 2393-2407 (May 2018)
[35] Krizhevsky, A.; Sutskever, I.; Hinton, G., Imagenet classification with deep convolutional neural networks, (NIPS Conference (2012)), 1097-1105
[36] He, K.; Zhang, X.; Ren, S.; Sun, J., Deep residual learning for image recognition, (CVPR (2016)), 770-778
[37] Vincent, P.; Larochelle, H.; Lajoie, I.; Bengio, Y.; Manzagol, P.-A., Stacked denoising autoencoders: learning useful representations in a deep network with a local denoising criterion, J. Mach. Learn. Res., 11, 3371-3408 (December 2010) · Zbl 1242.68256
[38] Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y., Generative adversarial nets, (Ghahramani, Z.; Welling, M.; Cortes, C.; Lawrence, N. D.; Weinberger, K. Q., Proc. NeurIPS 27 (2014), Curran Associates, Inc.), 2672-2680
[39] Mikolov, T.; Sutskever, I.; Chen, K.; Corrado, G. S.; Dean, J., Distributed representations of words and phrases and their compositionality, (Burges, C. J.C.; Bottou, L.; Welling, M.; Ghahramani, Z.; Weinberger, K. Q., NeurIPS 26 (2013), Curran Associates, Inc.), 3111-3119
[40] Bojanowski, P.; Grave, E.; Joulin, A.; Mikolov, T., Enriching word vectors with subword information, Trans. Assoc. Comput. Linguist., 5, 135-146 (2017)
[41] Peters, M. E.; Neumann, M.; Iyyer, M.; Gardner, M.; Clark, C.; Lee, K.; Zettlemoyer, L., Deep contextualized word representations, (Proc. of NAACL (2018))
[42] Arora, S.; Liang, Y.; Ma, T., A simple but tough-to-beat baseline for sentence embeddings, (Proc. ICLR (April 2017))
[43] Cho, K.; Merrienboer, B. V.; Gülçehre, Ç.; Bahdanau, D.; Bougares, F.; Schwenk, H.; Bengio, Y., Learning phrase representations using RNN encoder-decoder for statistical machine translation, (Proc. EMNLP (2014))
[44] Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, Ł.; Polosukhin, I., Attention is all you need, (Guyon, I.; Luxburg, U. V.; Bengio, S.; Wallach, H.; Fergus, R.; Vishwanathan, S.; Garnett, R., NeurIPS 30 (2017), Curran Associates, Inc.), 5998-6008
[45] Yadav, N.; Yadav, A.; Kumar, M., An Introduction to Neural Network Methods for Differential Equations, SpringerBriefs in Computational Intelligence (2015), Springer: Springer Netherlands · Zbl 1328.92006
[46] Lagaris, I. E.; Likas, A.; Fotiadis, D. I., Artificial neural networks for solving ordinary and partial differential equations, IEEE Trans. Neural Netw., 9, 5, 987-1000 (1998)
[47] Lagaris, I. E.; Likas, A. C.; Papageorgiou, D. G., Neural-network methods for boundary value problems with irregular boundaries, IEEE Trans. Neural Netw., 11, 5, 1041-1049 (September 2000)
[48] Raissi, M.; Perdikaris, P.; Karniadakis, G. E., Physics informed deep learning (part I): data-driven solution of nonlinear partial differential equations (November 2017)
[49] Raissi, M., Deep hidden physics models: deep learning of nonlinear partial differential equations, J. Mach. Learn. Res., 19, 25, 1-24 (2018) · Zbl 1439.68021
[50] Ray, D.; Hesthaven, J. S., An artificial neural network as a troubled-cell indicator, J. Comput. Phys., 367, 166-191 (April 2018) · Zbl 1415.65229
[51] Morgan, N. R.; Tokareva, S.; Liu, X.; Morgan, A. D., A machine learning approach for detecting shocks with high-order hydrodynamic methods, (AIAA SciTech Forum (January 2020))
[52] Després, B.; Jourdren, H., Machine learning design of volume of fluid schemes for compressible flows, J. Comput. Phys., 408, 1, Article 109275 pp. (May 2020) · Zbl 07505611
[53] Qi, Y.; Lu, J.; Scardovelli, R.; Zaleski, S.; Tryggvason, G., Computing curvature for volume of fluid methods using machine learning, J. Comput. Phys., 377, 155-161 (2019)
[54] Patel, H. V.; Panda, A.; Kuipers, J. A.M.; Peters, E. A.J. F., Computing interface curvature from volume fractions: a machine learning approach, Comput. Fluids, 193, Article 104263 pp. (October 2019) · Zbl 1458.76081
[55] Ataei, M.; Bussmann, M.; Shaayegan, V.; Costa, F.; Han, S.; Park, C. B., NPLIC: a machine learning approach to piecewise linear interface construction (January 2021) · Zbl 1521.76739
[56] Min, C., On reinitializing level set functions, J. Comput. Phys., 229, 8, 2764-2772 (April 2010) · Zbl 1188.65122
[57] Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V.; Vanderplas, J.; Passos, A.; Cournapeau, D.; Brucher, M.; Perrot, M.; Duchesnay, E., Scikit-learn: machine learning in Python, J. Mach. Learn. Res., 12, 2825-2830 (2011) · Zbl 1280.68189
[58] Strain, J., Tree methods for moving interfaces, J. Comput. Phys., 151, 2, 616-648 (May 1999) · Zbl 0942.76061
[59] Min, C.; Gibou, F., A second order accurate level set method on non-graded adaptive Cartesian grids, J. Comput. Phys., 225, 1, 300-321 (July 2007) · Zbl 1122.65077
[60] Mirzadeh, M.; Guittet, A.; Burstedde, C.; Gibou, F., Parallel level-set methods on adaptive tree-based grids, J. Comput. Phys., 322, 345-364 (October 2016) · Zbl 1352.65253
[61] de Berg, M.; van Kreveld, M.; Overmars, M.; Schwarzkopf, O., Computational Geometry - Algorithms and Applications (2000), Springer: Springer Cham · Zbl 0939.68134
[62] Parent, R., Computer Animation: Algorithms and Techniques (2008), Morgan Kaufmann
[63] Jr., F. S. Hill, Computer Graphics Using OpenGL (2001), Prentice-Hall Inc.
[64] Abadi, M.; Agarwal, A.; Barham, P.; Brevdo, E.; Chen, Z.; Citro, C.; Corrado, G. S.; Davis, A.; Dean, J.; Devin, M.; Ghemawat, S.; Goodfellow, I.; Harp, A.; Irving, G.; Isard, M.; Jozefowicz, R.; Jia, Y.; Kaiser, L.; Kudlur, M.; Levenberg, J.; Mané, D.; Schuster, M.; Monga, R.; Moore, S.; Murray, D.; Olah, C.; Shlens, J.; Steiner, B.; Sutskever, I.; Talwar, K.; Tucker, P.; Vanhoucke, V.; Vasudevan, V.; Viégas, F.; Vinyals, O.; Warden, P.; Wattenberg, M.; Wicke, M.; Yu, Y.; Zheng, X., TensorFlow: large-scale machine learning on heterogeneous systems (2015)
[65] Chollet, F., Keras (2015)
[66] LeCun, Y. A.; Bottou, L.; Orr, G. B.; Müller, K.-R., Efficient BackProp, (Lecture Notes in Comput. Sci., vol. 7700 (2012), Springer Berlin Heidelberg: Springer Berlin Heidelberg Berlin, Heidelberg), 9-48
[67] Parker, D. S., Exploring the Matrix - Adventures in Modeling with Matlab, UCLA Course Reader Solutions (January 2016)
[68] Glorot, X.; Bordes, A.; Bengio, Y., Deep sparse rectifier neural networks, (Gordon, G.; Dunson, D.; Dudík, M., Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, Fort Lauderdale, FL, USA. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, Fort Lauderdale, FL, USA, Proceedings of Machine Learning Research, vol. 15 (April 2011)), 315-323
[69] Goodfellow, I.; Bengio, Y.; Courville, A., Deep Learning (2016), MIT Press · Zbl 1373.68009
[70] Kingma, D. P.; Ba, J. L., Adam: a method for stochastic optimization, (ICLR (2015)), 1-13
[71] Glorot, X.; Bengio, Y., Understanding the difficulty of training deep feedforward neural networks, (Teh, Y. W.; Titterington, M., Proceedings of the Thirteen International Conference on Artificial Intelligence and Statistics. Proceedings of the Thirteen International Conference on Artificial Intelligence and Statistics, Chia Laguna Resort, Sardinia, Italy. Proceedings of the Thirteen International Conference on Artificial Intelligence and Statistics. Proceedings of the Thirteen International Conference on Artificial Intelligence and Statistics, Chia Laguna Resort, Sardinia, Italy, Proceedings of Machine Learning Research, vol. 9 (May 2010)), 249-256
[72] Karnakov, P.; Litvinov, S.; Koumoutsakos, P., A hybrid particle volume-of-fluid method for curvature estimation in multiphase flows, Int. J. Multiph. Flow, 125, Article 103209 pp. (January 2020)
[73] Burstedde, C.; Wilcox, L. C.; Ghattas, O., p4est: scalable algorithms for parallel adaptive mesh refinement on forests of octrees, SIAM J. Sci. Comput., 33, 3, 1103-1133 (May 2011) · Zbl 1230.65106
[74] Popinet, S., An accurate adaptive solver for surface-tension-driven interfacial flows, J. Comput. Phys., 228, 16, 5838-5866 (September 2009) · Zbl 1280.76020
[75] Cummins, S. J.; Francois, M. M.; Kothe, D. B., Estimating curvature from volume fractions, Comput. Struct., 83, 6-7, 425-434 (February 2005)
[76] Renardy, Y.; Renardy, M., PROST: a parabolic reconstruction of surface tension for the volume-of-fluid method, J. Comput. Phys., 183, 2, 400-421 (December 2002) · Zbl 1057.76569
[77] Peng, D.; Merriman, B.; Osher, S.; Zhao, H.; Kang, M., A PDE-based fast local level set method, J. Comput. Phys., 155, 2, 410-438 (November 1999) · Zbl 0964.76069
[78] Nielsen, M. B.; Museth, K., Dynamic tubular grid: an efficient data structure and algorithms for high resolution level sets, J. Sci. Comput., 26, 3, 261-299 (January 2006) · Zbl 1096.65020
[79] Brun, E.; Guittet, A.; Gibou, F., A local level-set method using a hash table data structure, J. Comput. Phys., 231, 6, 2528-2536 (March 2012) · Zbl 1242.65156
[80] Ervik, Å.; Lervåg, K. Y.; Munkejord, S. T., A robust method for calculating interface curvature and normal vectors using an extracted local level set, J. Comput. Phys., 257, Part A, 259-277 (January 2014) · Zbl 1349.76457
[81] Macklin, P.; Lowengrub, J., Evolving interfaces via gradients of geometry-dependent interior Poisson problems: application to tumor growth, J. Comput. Phys., 203, 1, 191-220 (February 2005) · Zbl 1067.65111
[82] Macklin, P.; Lowengrub, J., An improved geometry-aware curvature discretization for level set methods: application to tumor growth, J. Comput. Phys., 215, 2, 392-401 (July 2006) · Zbl 1089.92024
[83] Płoński, P., keras2cpp (December 2020)
[84] Zhu, M.; Gupta, S., To prune, or not to prune: exploring the efficacy of pruning for model compression, (Proc. ICLR Workshop (2018))
[85] Frankle, J.; Carbin, M., The lottery ticket hypothesis: finding sparse, trainable neural networks (2019)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.