{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,28]],"date-time":"2026-01-28T04:31:46Z","timestamp":1769574706268,"version":"3.49.0"},"reference-count":24,"publisher":"Springer Science and Business Media LLC","issue":"3","license":[{"start":{"date-parts":[[2019,10,3]],"date-time":"2019-10-03T00:00:00Z","timestamp":1570060800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2019,10,3]],"date-time":"2019-10-03T00:00:00Z","timestamp":1570060800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100002428","name":"Austrian Science Fund","doi-asserted-by":"publisher","award":["P 30747-N32"],"award-info":[{"award-number":["P 30747-N32"]}],"id":[{"id":"10.13039\/501100002428","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["J Math Imaging Vis"],"published-print":{"date-parts":[[2020,4]]},"abstract":"<jats:title>Abstract<\/jats:title>\n<jats:p>Deep learning and (deep) neural networks are emerging tools to address inverse problems and image reconstruction tasks. Despite outstanding performance, the mathematical analysis for solving inverse problems by neural networks is mostly missing. In this paper, we introduce and rigorously analyze families of deep regularizing neural networks (RegNets) of the form <jats:inline-formula><jats:alternatives><jats:tex-math>$$\\mathbf {B}_\\alpha + \\mathbf {N}_{\\theta (\\alpha )} \\mathbf {B}_\\alpha $$<\/jats:tex-math><mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\"><mml:mrow><mml:msub><mml:mi>B<\/mml:mi><mml:mi>\u03b1<\/mml:mi><\/mml:msub><mml:mo>+<\/mml:mo><mml:msub><mml:mi>N<\/mml:mi><mml:mrow><mml:mi>\u03b8<\/mml:mi><mml:mo>(<\/mml:mo><mml:mi>\u03b1<\/mml:mi><mml:mo>)<\/mml:mo><\/mml:mrow><\/mml:msub><mml:msub><mml:mi>B<\/mml:mi><mml:mi>\u03b1<\/mml:mi><\/mml:msub><\/mml:mrow><\/mml:math><\/jats:alternatives><\/jats:inline-formula>, where <jats:inline-formula><jats:alternatives><jats:tex-math>$$\\mathbf {B}_\\alpha $$<\/jats:tex-math><mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\"><mml:msub><mml:mi>B<\/mml:mi><mml:mi>\u03b1<\/mml:mi><\/mml:msub><\/mml:math><\/jats:alternatives><\/jats:inline-formula> is a classical regularization and the network <jats:inline-formula><jats:alternatives><jats:tex-math>$$\\mathbf {N}_{\\theta (\\alpha )} \\mathbf {B}_\\alpha $$<\/jats:tex-math><mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\"><mml:mrow><mml:msub><mml:mi>N<\/mml:mi><mml:mrow><mml:mi>\u03b8<\/mml:mi><mml:mo>(<\/mml:mo><mml:mi>\u03b1<\/mml:mi><mml:mo>)<\/mml:mo><\/mml:mrow><\/mml:msub><mml:msub><mml:mi>B<\/mml:mi><mml:mi>\u03b1<\/mml:mi><\/mml:msub><\/mml:mrow><\/mml:math><\/jats:alternatives><\/jats:inline-formula> is trained to recover the missing part <jats:inline-formula><jats:alternatives><jats:tex-math>$${\\text {Id}}_X - \\mathbf {B}_\\alpha $$<\/jats:tex-math><mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\"><mml:mrow><mml:msub><mml:mtext>Id<\/mml:mtext><mml:mi>X<\/mml:mi><\/mml:msub><mml:mo>-<\/mml:mo><mml:msub><mml:mi>B<\/mml:mi><mml:mi>\u03b1<\/mml:mi><\/mml:msub><\/mml:mrow><\/mml:math><\/jats:alternatives><\/jats:inline-formula> not found by the classical regularization. We show that these regularizing networks yield a convergent regularization method for solving inverse problems. Additionally, we derive convergence rates (quantitative error estimates) assuming a sufficient decay of the associated distance function. We demonstrate that our results recover existing convergence and convergence rates results for filter-based regularization methods as well as the recently introduced null space network as special cases. Numerical results are presented for a tomographic sparse data problem, which clearly demonstrate that the proposed RegNets improve classical regularization as well as the null space network.<\/jats:p>","DOI":"10.1007\/s10851-019-00911-1","type":"journal-article","created":{"date-parts":[[2019,10,3]],"date-time":"2019-10-03T16:02:46Z","timestamp":1570118566000},"page":"445-455","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":11,"title":["Big in Japan: Regularizing Networks for Solving Inverse Problems"],"prefix":"10.1007","volume":"62","author":[{"given":"Johannes","family":"Schwab","sequence":"first","affiliation":[]},{"given":"Stephan","family":"Antholzer","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5715-0331","authenticated-orcid":false,"given":"Markus","family":"Haltmeier","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2019,10,3]]},"reference":[{"key":"911_CR1","unstructured":"Adler, J., Lunz, S.: Banach Wasserstein GAN. In: Bengio, S., Wallach, H., Larochelle, H., Grauman, K., Cesa-Bianchi, N., Garnett, R. (eds.) Advances in Neural Information Processing Systems, pp. 6754\u20136763 (2018)"},{"issue":"12","key":"911_CR2","doi-asserted-by":"publisher","first-page":"124007","DOI":"10.1088\/1361-6420\/aa9581","volume":"33","author":"J Adler","year":"2017","unstructured":"Adler, J., \u00d6ktem, O.: Solving ill-posed inverse problems using iterative deep neural networks. Inverse Probl. 33(12), 124007 (2017)","journal-title":"Inverse Probl."},{"key":"911_CR3","unstructured":"Adler, J., \u00d6ktem, O.: Deep Bayesian inversion (2018). \narXiv:1811.05910"},{"key":"911_CR4","unstructured":"Adler, J., Ringh, A., \u00d6ktem, O., Karlsson, J.: Learning to solve inverse problems using Wasserstein loss (2017). \narXiv:1710.10898"},{"issue":"7","key":"911_CR5","doi-asserted-by":"publisher","first-page":"987","DOI":"10.1080\/17415977.2018.1518444","volume":"27","author":"S Antholzer","year":"2019","unstructured":"Antholzer, S., Haltmeier, M., Schwab, J.: Deep learning for photoacoustic tomography from sparse data. Inverse Problems Sci. Eng. 27(7), 987\u20131005 (2019). \nhttps:\/\/doi.org\/10.1080\/17415977.2018.1518444","journal-title":"Inverse Problems Sci. Eng."},{"issue":"6","key":"911_CR6","doi-asserted-by":"publisher","first-page":"064002","DOI":"10.1088\/1361-6420\/ab10ca","volume":"35","author":"TA Bubba","year":"2019","unstructured":"Bubba, T.A., Kutyniok, G., Lassas, M., M\u00e4rz, M., Samek, W., Siltanen, S., Srinivasan, V.: Learning the invisible: a hybrid deep learning-shearlet framework for limited angle computed tomography. Inverse Problems 35(6), 064002 (2019)","journal-title":"Inverse Problems"},{"key":"911_CR7","doi-asserted-by":"publisher","DOI":"10.1007\/978-94-009-1740-8","volume-title":"Regularization of Inverse Problems","author":"HW Engl","year":"1996","unstructured":"Engl, H.W., Hanke, M., Neubauer, A.: Regularization of Inverse Problems, vol. 375. Springer, Berlin (1996)"},{"issue":"6","key":"911_CR8","doi-asserted-by":"publisher","first-page":"1440","DOI":"10.1109\/TMI.2018.2832656","volume":"37","author":"H Gupta","year":"2018","unstructured":"Gupta, H., Jin, K.H., Nguyen, H.Q., McCann, M.T., Unser, M.: CNN-based projected gradient descent for consistent CT image reconstruction. IEEE Trans. Med. Imaging 37(6), 1440\u20131453 (2018)","journal-title":"IEEE Trans. Med. Imaging"},{"issue":"6","key":"911_CR9","doi-asserted-by":"publisher","first-page":"1418","DOI":"10.1109\/TMI.2018.2823768","volume":"37","author":"Y Han","year":"2018","unstructured":"Han, Y., Ye, J.C.: Framing U-Net via deep convolutional framelets: application to sparse-view CT. IEEE Trans. Med. Imaging 37(6), 1418\u20131429 (2018)","journal-title":"IEEE Trans. Med. Imaging"},{"key":"911_CR10","volume-title":"Introduction to Spectral Theory in Hilbert Space. North-Holland Series in Applied Mathematics and Mechanics","author":"G Helmberg","year":"1969","unstructured":"Helmberg, G.: Introduction to Spectral Theory in Hilbert Space. North-Holland Series in Applied Mathematics and Mechanics, vol. 6. North-Holland Publishing Co., Amsterdam (1969)"},{"issue":"3","key":"911_CR11","doi-asserted-by":"publisher","first-page":"805","DOI":"10.1088\/0266-5611\/21\/3\/002","volume":"21","author":"B Hofmann","year":"2005","unstructured":"Hofmann, B., Yamamoto, M.: Convergence rates for Tikhonov regularization based on range inclusions. Inverse Problems 21(3), 805 (2005)","journal-title":"Inverse Problems"},{"key":"911_CR12","doi-asserted-by":"crossref","unstructured":"Kobler, E., Klatzer, T., Hammernik, K., Pock, T.: Variational networks: connecting variational methods and deep learning. In: German Conference on Pattern Recognition, pp. 281\u2013293. Springer (2017)","DOI":"10.1007\/978-3-319-66709-6_23"},{"key":"911_CR13","doi-asserted-by":"crossref","unstructured":"Kofler, A., Haltmeier, M., Kolbitsch, C., Kachelrie\u00df, M., Dewey, M.: A U-Nets cascade for sparse view computed tomography. In: International Workshop on Machine Learning for Medical Image Reconstruction, pp. 91\u201399. Springer (2018)","DOI":"10.1007\/978-3-030-00129-2_11"},{"issue":"10","key":"911_CR14","doi-asserted-by":"publisher","first-page":"1834","DOI":"10.1364\/JOSAA.7.001834","volume":"7","author":"RM Lewitt","year":"1990","unstructured":"Lewitt, R.M.: Multidimensional digital image representations using generalized Kaiser\u2013Bessel window functions. JOSA A 7(10), 1834\u20131846 (1990)","journal-title":"JOSA A"},{"key":"911_CR15","unstructured":"Li, H., Schwab, J., Antholzer, S., Haltmeier, M.: NETT: solving inverse problems with deep neural networks (2018). \narXiv:1803.00092"},{"key":"911_CR16","unstructured":"Mardani, M., Gong, E., Cheng, J.Y., Vasanawala, S., Zaharchuk, G., Alley, M., Thakur, N., Han, S., Dally, W., Pauly, J.M., et\u00a0al.: Deep generative adversarial networks for compressed sensing automates MRI (2017). \narXiv:1706.00051"},{"key":"911_CR17","unstructured":"Mardani, M., Monajemi, H., Papyan, V., Vasanawala, S., Donoho, D., Pauly, J.: Recurrent generative adversarial networks for proximal learning and automated compressive image recovery (2017). \narXiv:1711.10046"},{"key":"911_CR18","doi-asserted-by":"crossref","unstructured":"Ronneberge, O., Fischer, P., Brox, T.: U-net: Convolutional networks for biomedical image segmentation. In: International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 234\u2013241 (2015)","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"911_CR19","volume-title":"Variational Methods in Imaging. Applied Mathematical Sciences","author":"O Scherzer","year":"2009","unstructured":"Scherzer, O., Grasmair, M., Grossauer, H., Haltmeier, M., Lenzen, F.: Variational Methods in Imaging. Applied Mathematical Sciences, vol. 167. Springer, New York (2009)"},{"key":"911_CR20","doi-asserted-by":"crossref","unstructured":"Schlemper, J., Caballero, J., Hajnal, J.V., Price, A., Rueckert, D.: A deep cascade of convolutional neural networks for MR image reconstruction. In: International Conference on Information Processing in Medical Imaging, pp. 647\u2013658. Springer (2017)","DOI":"10.1007\/978-3-319-59050-9_51"},{"issue":"2","key":"911_CR21","doi-asserted-by":"publisher","first-page":"025008","DOI":"10.1088\/1361-6420\/aaf14a","volume":"35","author":"J Schwab","year":"2019","unstructured":"Schwab, J., Antholzer, S., Haltmeier, M.: Deep null space learning for inverse problems: convergence analysis and rates. Inverse Probl. 35(2), 025008 (2019)","journal-title":"Inverse Probl."},{"key":"911_CR22","doi-asserted-by":"crossref","unstructured":"Schwab, J., Antholzer, S., Nuster, R., Paltauf, G., Haltmeier, M.: Deep learning of truncated singular values for limited view photoacoustic tomography. In: Photons Plus Ultrasound: Imaging and Sensing 2019, vol. 10878, p. 1087836 (2019)","DOI":"10.1117\/12.2508418"},{"key":"911_CR23","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4612-6027-1","volume-title":"Linear Operators in Hilbert Spaces. Graduate Texts in Mathematics","author":"J Weidmann","year":"1980","unstructured":"Weidmann, J.: Linear Operators in Hilbert Spaces. Graduate Texts in Mathematics, vol. 68. Springer, New York (1980)"},{"issue":"2","key":"911_CR24","doi-asserted-by":"publisher","first-page":"991","DOI":"10.1137\/17M1141771","volume":"11","author":"JC Ye","year":"2018","unstructured":"Ye, J.C., Han, Y., Cha, E.: Deep convolutional framelets: a general deep learning framework for inverse problems. SIAM J. Imaging Sci. 11(2), 991\u20131048 (2018)","journal-title":"SIAM J. Imaging Sci."}],"container-title":["Journal of Mathematical Imaging and Vision"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/link.springer.com\/content\/pdf\/10.1007\/s10851-019-00911-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"http:\/\/link.springer.com\/article\/10.1007\/s10851-019-00911-1\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"http:\/\/link.springer.com\/content\/pdf\/10.1007\/s10851-019-00911-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2020,10,1]],"date-time":"2020-10-01T23:35:01Z","timestamp":1601595301000},"score":1,"resource":{"primary":{"URL":"http:\/\/link.springer.com\/10.1007\/s10851-019-00911-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2019,10,3]]},"references-count":24,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2020,4]]}},"alternative-id":["911"],"URL":"https:\/\/doi.org\/10.1007\/s10851-019-00911-1","relation":{},"ISSN":["0924-9907","1573-7683"],"issn-type":[{"value":"0924-9907","type":"print"},{"value":"1573-7683","type":"electronic"}],"subject":[],"published":{"date-parts":[[2019,10,3]]},"assertion":[{"value":"29 November 2018","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 September 2019","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"3 October 2019","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}]}}