{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T21:18:08Z","timestamp":1792790288067,"version":"4.5.0"},"reference-count":87,"publisher":"Society for Industrial & Applied Mathematics (SIAM)","issue":"1","funder":[{"name":"European Community","award":["DEDALE 665044"],"award-info":[{"award-number":["DEDALE 665044"]}]},{"name":"European Community","award":["NORIA 724175"],"award-info":[{"award-number":["NORIA 724175"]}]},{"DOI":"10.13039\/501100001665","name":"Agence Nationale de la Recherche","doi-asserted-by":"publisher","award":["ROOT"],"award-info":[{"award-number":["ROOT"]}],"id":[{"id":"10.13039\/501100001665","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002830","name":"Centre National d\u0092Etudes Spatiales","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100002830","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["SIAM J. Imaging Sci."],"published-print":{"date-parts":[[2018,1]]},"abstract":"<jats:p>This paper introduces a new nonlinear dictionary learning method for histograms in the probability simplex. The method leverages optimal transport theory, in the sense that our aim is to reconstruct histograms using so-called displacement interpolations (a.k.a. Wasserstein barycenters) between dictionary atoms; such atoms are themselves synthetic histograms in the probability simplex. Our method simultaneously estimates such atoms and, for each datapoint, the vector of weights that can optimally reconstruct it as an optimal transport barycenter of such atoms. Our method is computationally tractable thanks to the addition of an entropic regularization to the usual optimal transportation problem, leading to an approximation scheme that is efficient, parallel, and simple to differentiate. Both atoms and weights are learned using a gradient-based descent method. Gradients are obtained by automatic differentiation of the generalized Sinkhorn iterations that yield barycenters with entropic smoothing. Because of its formulation relying on Wasserstein barycenters instead of the usual matrix product between dictionary and codes, our method allows for nonlinear relationships between atoms and the reconstruction of input data. We illustrate its application in several different image processing settings.<\/jats:p>","DOI":"10.1137\/17m1140431","type":"journal-article","created":{"date-parts":[[2018,3,1]],"date-time":"2018-03-01T19:45:34Z","timestamp":1519933534000},"page":"643-678","source":"Crossref","is-referenced-by-count":100,"title":["Wasserstein Dictionary Learning: Optimal Transport-Based Unsupervised Nonlinear Dictionary Learning"],"prefix":"10.1137","volume":"11","author":[{"given":"Morgan A.","family":"Schmitz","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Matthieu","family":"Heitz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nicolas","family":"Bonneel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fred","family":"Ngol\u00e8","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David","family":"Coeurjolly","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marco","family":"Cuturi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gabriel","family":"Peyr\u00e9","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jean-Luc","family":"Starck","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"351","published-online":{"date-parts":[[2018,3,1]]},"reference":[{"key":"atypb1","doi-asserted-by":"publisher","DOI":"10.1137\/100805741"},{"key":"atypb2","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2006.881199"},{"key":"atypb3","unstructured":"N. Aifanti, C. Papachristou, and A. Delopoulos,\n                      The MUG facial expression database\n                      , in Proceedings of the 11th IEEE International Conference on Image Analysis for Multimedia Interactive Services (WIAMIS), 2010, pp. 1-4."},{"key":"atypb4","unstructured":"J. Altschuler, J. Weed, and P. Rigollet,\n                      Near-Linear Time Approximation Algorithms for Optimal Transport via Sinkhorn Iteration\n                      , preprint,https:\/\/arxiv.org\/abs\/1705.09634, 2017."},{"key":"atypb5","unstructured":"M. Arjovsky, S. Chintala, and L. Bottou,\n                      Wasserstein GAN\n                      , preprint,https:\/\/arxiv.org\/abs\/1701.07875, 2017."},{"key":"atypb6","doi-asserted-by":"publisher","DOI":"10.1016\/j.spl.2006.02.001"},{"key":"atypb7","doi-asserted-by":"publisher","DOI":"10.1137\/S0040585X97981664"},{"key":"atypb8","doi-asserted-by":"publisher","DOI":"10.1137\/141000439"},{"key":"atypb9","unstructured":"E. Bernton, P. E. Jacob, M. Gerber, and C. P. Robert,\n                      Inference in Generative Models Using the Wasserstein Distance\n                      , preprint,https:\/\/arxiv.org\/abs\/1701.05146, 2017."},{"key":"atypb10","doi-asserted-by":"publisher","DOI":"10.1007\/BF02186476"},{"key":"atypb11","unstructured":"J. Bigot, R. Gouet, T. Klein, and A. Lo\u0301pez,\n                      Geodesic PCA in the Wasserstein Space\n                      , preprint,https:\/\/arxiv.org\/abs\/1307.7721, 2013."},{"key":"atypb12","first-page":"71","author":"Blei D.","year":"2009","journal-title":"London"},{"key":"atypb13","doi-asserted-by":"publisher","DOI":"10.3150\/13-BEJ585"},{"key":"atypb14","doi-asserted-by":"publisher","DOI":"10.1145\/2897824.2925918"},{"key":"atypb15","doi-asserted-by":"publisher","DOI":"10.1007\/s10851-014-0506-3"},{"key":"atypb16","doi-asserted-by":"publisher","DOI":"10.1051\/m2an\/2015033"},{"key":"atypb17","unstructured":"L. Chizat, G. Peyre\u0301, B. Schmitzer, and F.X. Vialard,\n                      Scaling Algorithms for Unbalanced Transport Problems\n                      , preprint,https:\/\/arxiv.org\/abs\/1607.05816, 2016."},{"key":"atypb18","first-page":"2292","author":"Cuturi M.","year":"2013","journal-title":"MA"},{"key":"atypb19","unstructured":"M. Cuturi and A. Doucet,\n                      Fast computation of Wasserstein barycenters\n                      , in Proceedings of the 31st International Conference on Machine Learning, 2014, pp. 685-693."},{"key":"atypb20","doi-asserted-by":"publisher","DOI":"10.1137\/15M1032600"},{"key":"atypb21","doi-asserted-by":"publisher","DOI":"10.1137\/050645506"},{"key":"atypb22","doi-asserted-by":"publisher","DOI":"10.1214\/aoms\/1177731829"},{"key":"atypb23","unstructured":"S. Erlander and N. F. Stewart,\n                      The Gravity Model in Transportation Analysis: Theory and Extensions\n                      , Vol. 3, VSP, Utrecht, The Netherlands, 1990."},{"key":"atypb24","doi-asserted-by":"publisher","DOI":"10.1109\/TMI.2004.831793"},{"key":"atypb25","doi-asserted-by":"publisher","DOI":"10.1016\/0024-3795(89)90490-4"},{"key":"atypb26","first-page":"215","volume":"10","author":"Fre\u0301chet M.","year":"1948","journal-title":"Ann. Inst. H. Poincare\u0301"},{"key":"atypb27","first-page":"2053","author":"Frogner C.","year":"2015","journal-title":"MA"},{"key":"atypb28","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2014.2318132"},{"key":"atypb29","unstructured":"A. Genevay, G. Peyre\u0301, and M. Cuturi,\n                      Learning Generative Models with Sinkhorn Divergences\n                      , preprint,https:\/\/arxiv.org\/abs\/1706.00292, 2017."},{"key":"atypb30","doi-asserted-by":"crossref","unstructured":"A. Griewank and A. Walther,\n                      Evaluating Derivatives: Principles and Techniques of Algorithmic Differentiation\n                      , 2nd ed., SIAM, Philadelphia, 2008,https:\/\/doi.org\/10.1137\/1.9780898717761.","DOI":"10.1137\/1.9780898717761"},{"key":"atypb31","doi-asserted-by":"publisher","DOI":"10.1023\/B:VISI.0000036836.66311.97"},{"key":"atypb32","first-page":"3926","author":"Harandi M.","year":"2015","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"atypb33","first-page":"3120","author":"Harandi M.","year":"2013","journal-title":"Proceedings of the IEEE International Conference on Computer Vision"},{"key":"atypb34","doi-asserted-by":"crossref","unstructured":"M. T. Harandi, C. Sanderson, R. Hartley, and B. C. Lovell,\n                      Sparse coding and dictionary learning for symmetric positive definite matrices: A kernel approach\n                      , in Computer Vision-ECCV 2012, Springer, Berlin, Heidelberg, 2012, pp. 216-229,https:\/\/doi.org\/10.1007\/978-3-642-33709-3_16.","DOI":"10.1007\/978-3-642-33709-3_16"},{"key":"atypb35","doi-asserted-by":"publisher","DOI":"10.1126\/science.1127647"},{"key":"atypb36","first-page":"1480","author":"Ho J.","year":"2013","journal-title":"Proceedings of the International Conference on Machine Learning"},{"key":"atypb37","unstructured":"A. Hyva\u0308rinen, J. Karhunen, and E. Oja,\n                      Independent Component Analysis\n                      , Wiley Ser. Adapt. Learn. Syst. Signal Process. Commun. Control 46, John Wiley & Sons, New York, 2004."},{"key":"atypb38","unstructured":"Z. Irace and H. Batatia,\n                      Motion-based interpolation to estimate spatially variant PSF in positron emission tomography\n                      , in Proceedings of the 21st European IEEE Signal Processing Conference (EUSIPCO), 2013, pp. 1-5."},{"key":"atypb39","doi-asserted-by":"publisher","DOI":"10.1002\/nav.3800020109"},{"key":"atypb40","unstructured":"R. Laureijs, J. Amiaux, S. Arduini, J.L. Augueres, J. Brinchmann, R. Cole, M. Cropper, C. Dabin, L. Duvet, A. Ealet, et al.\n                      Euclid Definition Study Report\n                      , preprint,https:\/\/arxiv.org\/abs\/1110.3193, 2011."},{"key":"atypb41","doi-asserted-by":"publisher","DOI":"10.1038\/44565"},{"key":"atypb42","first-page":"801","author":"Lee H.","year":"2007","journal-title":"MA"},{"key":"atypb43","doi-asserted-by":"publisher","DOI":"10.3934\/dcds.2014.34.1533"},{"key":"atypb44","first-page":"1601","author":"Li P.","year":"2013","journal-title":"Proceedings of the IEEE International Conference on Computer Vision"},{"key":"atypb45","unstructured":"H. Liu, J. Qin, H. Cheng, and F. Sun,\n                      Robust kernel dictionary learning using a whole sequence convergent algorithm\n                      , in Proceedings of the 24th International Conference on Artificial Intelligence (IJCAI), 2015, pp. 3678-3684."},{"key":"atypb46","first-page":"19","volume":"11","author":"Mairal J.","year":"2010","journal-title":"J. Mach. Learn. Res."},{"key":"atypb47","doi-asserted-by":"crossref","unstructured":"S. Mallat,\n                      A Wavelet Tour of Signal Processing\n                      , Academic Press, New York, 1999,https:\/\/doi.org\/10.1016\/b978-0-12-374370-1.00002-1.","DOI":"10.1016\/B978-012466606-1\/50008-8"},{"key":"atypb48","doi-asserted-by":"publisher","DOI":"10.1006\/aima.1997.1634"},{"key":"atypb49","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-8659.2011.02032.x"},{"key":"atypb50","first-page":"666","volume":"1781","author":"Monge G.","journal-title":"Paris"},{"key":"atypb51","first-page":"3711","author":"Montavon G.","year":"2016","journal-title":"MA"},{"key":"atypb52","doi-asserted-by":"publisher","DOI":"10.1145\/2049662.2049669"},{"key":"atypb53","doi-asserted-by":"crossref","unstructured":"Y. Nesterov,\n                      Introductory Lectures on Convex Optimization: A Basic Course\n                      , Appl. Optim. 87, Springer Science & Business Media, New York, 2013,https:\/\/doi.org\/10.1007\/978-1-4419-8853-9.","DOI":"10.1007\/978-1-4419-8853-9"},{"key":"atypb54","unstructured":"F. Ngole\u0300 and J.L. Starck,\n                      PSF Field Learning Based on Optimal Transport Distances\n                      , preprint,https:\/\/arxiv.org\/abs\/1703.06066, 2017."},{"key":"atypb55","doi-asserted-by":"publisher","DOI":"10.1051\/0004-6361\/201424167"},{"key":"atypb56","unstructured":"N. Papadakis,\n                      Optimal Transport for Image Processing\n                      , habilitation a\u0300 diriger des recherches, Universite\u0301 de Bordeaux, Bordeaux, France, 2015,https:\/\/hal.archives-ouvertes.fr\/tel-01246096."},{"key":"atypb57","doi-asserted-by":"publisher","DOI":"10.1080\/14786440109462720"},{"key":"atypb58","doi-asserted-by":"crossref","unstructured":"J. Pennington, R. Socher, and C. D. Manning,\n                      GloVe: Global vectors for word representation\n                      , in Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), Vol. 14, 2014, pp. 1532-1543,https:\/\/doi.org\/10.3115\/v1\/d14-1162.","DOI":"10.3115\/v1\/D14-1162"},{"key":"atypb59","unstructured":"G. Peyre\u0301, L. Chizat, F.X. Vialard, and J. Solomon,\n                      Quantum Optimal Transport for Tensor Field Processing\n                      , preprint,https:\/\/arxiv.org\/abs\/1612.08731, 2016."},{"key":"atypb60","first-page":"1434","author":"Pitie\u0301 F.","year":"2005","journal-title":"D.C."},{"key":"atypb61","doi-asserted-by":"publisher","DOI":"10.1016\/0041-5553(64)90137-5"},{"key":"atypb62","first-page":"308","author":"Quan Y.","year":"2016","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"atypb63","first-page":"435","author":"Rabin J.","year":"2011","journal-title":"Heidelberg"},{"key":"atypb64","unstructured":"S. Rachev and L. Ru\u0308schendorf,\n                      Mass Transportation Problems: Theory\n                      , Vol. 1, Springer-Verlag, New York, 1998."},{"key":"atypb65","unstructured":"A. Rolet, M. Cuturi, and G. Peyre\u0301,\n                      Fast dictionary learning with a smoothed Wasserstein loss\n                      , in Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, 2016, pp. 630-638."},{"key":"atypb66","first-page":"1","volume":"40","author":"Rubinstein R.","year":"2008","journal-title":"Cs Technion"},{"key":"atypb67","doi-asserted-by":"publisher","DOI":"10.1023\/A:1026543900054"},{"key":"atypb68","unstructured":"G. Salton and M. J. McGill,\n                      Introduction to Modern Information Retrieval\n                      , McGraw-Hill, New York, 1986."},{"key":"atypb69","first-page":"1873","author":"Sandler R.","year":"2009","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"atypb70","doi-asserted-by":"crossref","unstructured":"M. A. Schmitz, M. Heitz, N. Bonneel, F. Ngole\u0300, D. Coeurjolly, M. Cuturi, G. Peyre\u0301, and J.L. Starck,\n                      Optimal Transport-Based Dictionary Learning and Its Application to Euclid-Like Point Spread Function Representation\n                      , Proc. SPIE 10394, International Society for Optics and Photonics, Bellingham, WA, 2017,https:\/\/doi.org\/10.1117\/12.2270641.","DOI":"10.1117\/12.2270641"},{"key":"atypb71","unstructured":"B. Schmitzer,\n                      Stabilized Sparse Scaling Algorithms for Entropy Regularized Transport Problems\n                      , preprint,https:\/\/arxiv.org\/abs\/1610.06519, 2016."},{"key":"atypb72","doi-asserted-by":"crossref","unstructured":"B. Scho\u0308lkopf, A. Smola, and K.R. Mu\u0308ller,\n                      Kernel principal component analysis\n                      , in Proceedings of the 7th International Conference on Artificial Neural Networks (ICANN), 1997, pp. 583-588.","DOI":"10.1007\/BFb0020217"},{"key":"atypb73","unstructured":"E. Schro\u0308dinger,\n                      U\u0308ber die umkehrung der naturgesetze\n                      , Verlag Akademie der wissenschaften in kommission bei Walter de Gruyter u. Company, Berlin, 1931."},{"key":"atypb74","first-page":"3312","author":"Seguy V.","year":"2015","journal-title":"MA"},{"key":"atypb75","first-page":"1","author":"Shirdhonkar S.","year":"2008","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"atypb76","doi-asserted-by":"publisher","DOI":"10.2307\/2314570"},{"key":"atypb77","doi-asserted-by":"publisher","DOI":"10.1145\/2766963"},{"key":"atypb78","unstructured":"J. Solomon, R. Rustamov, L. Guibas, and A. Butscher,\n                      Wasserstein propagation for semi-supervised learning\n                      , in Proceedings of the 31st International Conference on Machine Learning, 2014, pp. 306-314."},{"key":"atypb79","doi-asserted-by":"publisher","DOI":"10.1007\/BF02249265"},{"key":"atypb80","unstructured":"Theano Development Team,\n                      Theano: A Python Framework for Fast Computation of Mathematical Expressions\n                      , preprint,https:\/\/arxiv.org\/abs\/1605.02688, 2016."},{"key":"atypb81","doi-asserted-by":"publisher","DOI":"10.1162\/jocn.1991.3.1.71"},{"key":"atypb82","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2013.2282078"},{"key":"atypb83","doi-asserted-by":"crossref","unstructured":"C. Villani,\n                      Topics in Optimal Transportation\n                      , Grad. Stud. Math. 58, AMS, Providence, RI, 2003,https:\/\/doi.org\/10.1090\/gsm\/058.","DOI":"10.1090\/gsm\/058"},{"key":"atypb84","unstructured":"C. Villani,\n                      Optimal Transport: Old and New\n                      , Grundlehren Math. Wiss. 338, Springer, Berlin, Heidelberg, 2008."},{"key":"atypb85","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-012-0566-z"},{"key":"atypb86","doi-asserted-by":"publisher","DOI":"10.1109\/TSP.2017.2659647"},{"key":"atypb87","doi-asserted-by":"publisher","DOI":"10.1007\/BF01128757"}],"container-title":["SIAM Journal on Imaging Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/epubs.siam.org\/doi\/pdf\/10.1137\/17M1140431","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T12:45:14Z","timestamp":1787316314000},"score":1,"resource":{"primary":{"URL":"https:\/\/epubs.siam.org\/doi\/10.1137\/17M1140431"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,1]]},"references-count":87,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2018,1]]}},"alternative-id":["10.1137\/17M1140431"],"URL":"https:\/\/doi.org\/10.1137\/17m1140431","relation":{},"ISSN":["1936-4954"],"issn-type":[{"value":"1936-4954","type":"electronic"}],"subject":[],"published":{"date-parts":[[2018,1]]}}}