{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T04:43:32Z","timestamp":1774413812549,"version":"3.50.1"},"reference-count":39,"publisher":"Springer Science and Business Media LLC","issue":"8","license":[{"start":{"date-parts":[[2025,4,1]],"date-time":"2025-04-01T00:00:00Z","timestamp":1743465600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,5,17]],"date-time":"2025-05-17T00:00:00Z","timestamp":1747440000000},"content-version":"vor","delay-in-days":46,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/100017142","name":"Gruppo Nazionale per il Calcolo Scientifico","doi-asserted-by":"publisher","award":["CUP E53C23001670001"],"award-info":[{"award-number":["CUP E53C23001670001"]}],"id":[{"id":"10.13039\/100017142","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003407","name":"Ministero dell\u2019Istruzione, dell\u2019Universit\u00e0 e della Ricerca","doi-asserted-by":"publisher","award":["2022N3ZNAX (CUP E53D23007700006)"],"award-info":[{"award-number":["2022N3ZNAX (CUP E53D23007700006)"]}],"id":[{"id":"10.13039\/501100003407","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003407","name":"Ministero dell\u2019Istruzione, dell\u2019Universit\u00e0 e della Ricerca","doi-asserted-by":"publisher","award":["P2022J9SNP (CUP E53D23017980001)"],"award-info":[{"award-number":["P2022J9SNP (CUP E53D23017980001)"]}],"id":[{"id":"10.13039\/501100003407","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100003407","name":"Ministero dell\u2019Istruzione, dell\u2019Universit\u00e0 e della Ricerca","doi-asserted-by":"publisher","award":["20225STXSB (CUP E53D23005480006)"],"award-info":[{"award-number":["20225STXSB (CUP E53D23005480006)"]}],"id":[{"id":"10.13039\/501100003407","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Soft Comput"],"published-print":{"date-parts":[[2025,4]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>The so-called Deep Image Prior approach is an unsupervised deep learning methodology which has gained great interest in recent years due to its effectiveness in tackling imaging problems. However, a well known drawback of Deep Image Prior is the need for a proper early stopping technique to prevent undesired corruptions in the reconstructed images. Determining the optimal number of iterations depends on both the specific application and the features of the images being processed. As a consequence, several numerical trials are typically required to decide when to stop the Deep Image Prior procedure, resulting in significant computational costs and time requirements. This paper aims to introduce two early stopping techniques for Deep Image Prior, based on different approaches and different perspectives. The first approach relies on the neural architecture search (NAS) strategy. The aim is to equip the neural network used in Deep Image Prior with hyperparameter configurations that produce high-quality reconstructed images that are (i) comparable to those obtained with the optimally stopped, originally configured Deep Image Prior, and (ii) achieved with significantly fewer iterations. The second proposed early stopping strategy is based on a modified version of the BRISQUE metric, a no-reference image quality measure, and it aims to track the behaviour of the PSNR curve, obtained by applying Deep Image Prior, without knowing the ground truth image. While the NAS-based early stopping technique is particularly suited in those situations where the computational time is limited, this latter one is also relevant when a larger number of iterations is allowed. Several numerical experiments on different denoising applications show a promising performance of Deep Image Prior combined with the suggested early stopping procedures.<\/jats:p>","DOI":"10.1007\/s00500-025-10642-8","type":"journal-article","created":{"date-parts":[[2025,5,16]],"date-time":"2025-05-16T22:23:12Z","timestamp":1747434192000},"page":"4153-4174","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["Early stopping strategies in Deep Image Prior"],"prefix":"10.1007","volume":"29","author":[{"given":"Alessandro","family":"Benfenati","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6378-3063","authenticated-orcid":false,"given":"Ambra","family":"Catozzi","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Giorgia","family":"Franchini","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Federica","family":"Porta","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2025,5,17]]},"reference":[{"key":"10642_CR1","unstructured":"Alshubaily I (2021) Efficient neural architecture search with performance prediction. arXiv:2108.01854"},{"key":"10642_CR2","doi-asserted-by":"crossref","unstructured":"Arican ME, Kara O, Bredell G, Konukoglu E (2022) ISNAN-DIP: image-specific neural architecture search for deep image prior. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 1960\u20131968","DOI":"10.1109\/CVPR52688.2022.00200"},{"issue":"11","key":"10642_CR3","doi-asserted-by":"publisher","first-page":"0276972","DOI":"10.1371\/journal.pone.0276972","volume":"17","author":"A Benfenati","year":"2022","unstructured":"Benfenati A, Bolzi D, Causin P, Oberti R (2022) A deep learning generative model approach for image synthesis of plant leaves. PLoS ONE 17(11):0276972","journal-title":"PLoS ONE"},{"key":"10642_CR4","doi-asserted-by":"publisher","first-page":"054003","DOI":"10.1088\/1361-6420\/acc2e0","volume":"39","author":"A Benfenati","year":"2023","unstructured":"Benfenati A, Catozzi A, Ruggiero V (2023a) Neural blind deconvolution with Poisson data. Inverse Probl 39:054003","journal-title":"Inverse Probl"},{"key":"10642_CR5","doi-asserted-by":"crossref","unstructured":"Benfenati A, Catozzi A, Franchini G, Porta F (2023b) Piece-wise constant image segmentation with a deep image prior approach. In: International conference on scale space and variational methods in computer vision. Springer, pp 352\u2013362","DOI":"10.1007\/978-3-031-31975-4_27"},{"key":"10642_CR6","doi-asserted-by":"publisher","DOI":"10.1088\/2053-2563\/aae109","volume-title":"Inverse imaging with Poisson data: from cells to galaxies","author":"M Bertero","year":"2018","unstructured":"Bertero M, Boccacci P, Ruggiero V (2018) Inverse imaging with Poisson data: from cells to galaxies. IOP Publishing"},{"key":"10642_CR7","doi-asserted-by":"crossref","unstructured":"Burrows L, Chen K, Torella F (2021) A deep image prior learning algorithm for joint selective segmentation and registration. In: International conference on scale space and variational methods in computer vision. Springer, pp 411\u2013422","DOI":"10.1007\/978-3-030-75549-2_33"},{"key":"10642_CR8","doi-asserted-by":"publisher","first-page":"6923","DOI":"10.1007\/s00500-023-09576-w","volume":"28","author":"S Charfi","year":"2024","unstructured":"Charfi S, Ansari ME, Koutti L, Ellahyani A, Eljaafari I (2024) Modified residual attention network for abnormalities segmentation and detection in WCE images. Soft Comput 28:6923\u20136936","journal-title":"Soft Comput"},{"key":"10642_CR9","doi-asserted-by":"crossref","unstructured":"Chen Y-C, Gao C, Robb E, Huang J-B (2020) NAS-DIP: learning deep image prior with neural architecture search. In: Computer Vision\u2014ECCV 2020: 16th European Conference, Glasgow, UK, August 23\u201328, 2020, Proceedings, Part XVIII, vol 16. Springer, pp 442\u2013459","DOI":"10.1007\/978-3-030-58523-5_26"},{"key":"10642_CR10","doi-asserted-by":"crossref","unstructured":"Cheng Z, Gadelha M, Maji S, Sheldon D (2019) A Bayesian perspective on the deep image prior. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","DOI":"10.1109\/CVPR.2019.00559"},{"issue":"55","key":"10642_CR11","first-page":"1","volume":"20","author":"T Elsken","year":"2019","unstructured":"Elsken T, Metzen JH, Hutter F (2019) Neural architecture search: a survey. J Mach Learn Res 20(55):1\u201321","journal-title":"J Mach Learn Res"},{"issue":"1","key":"10642_CR12","doi-asserted-by":"publisher","first-page":"1","DOI":"10.3934\/mine.2023012","volume":"5","author":"G Franchini","year":"2023","unstructured":"Franchini G, Ruggiero V, Porta F, Zanni L (2023) Neural architecture search via standard machine learning methodologies. Math Eng 5(1):1\u201321","journal-title":"Math Eng"},{"key":"10642_CR13","doi-asserted-by":"publisher","first-page":"188","DOI":"10.1016\/j.cag.2021.05.013","volume":"98","author":"K Ho","year":"2021","unstructured":"Ho K, Gilbert A, Jin H, Collomosse J (2021) Neural architecture search for deep image prior. Comput Graph 98:188\u2013196","journal-title":"Comput Graph"},{"key":"10642_CR14","doi-asserted-by":"crossref","unstructured":"Hor\u00e9 A, Ziou D (2010) Image quality metrics: PSNR vs. SSIM. In: 2010 20th international conference on pattern recognition, pp 2366\u20132369","DOI":"10.1109\/ICPR.2010.579"},{"issue":"10","key":"10642_CR15","doi-asserted-by":"publisher","first-page":"11472","DOI":"10.1109\/TPAMI.2023.3283979","volume":"45","author":"D Huo","year":"2023","unstructured":"Huo D, Masoumzadeh A, Kushol R, Yang Y-H (2023) Blind image deconvolution using variational deep image prior. IEEE Trans Pattern Anal Mach Intell 45(10):11472\u201311483","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"10642_CR16","doi-asserted-by":"crossref","unstructured":"Kotera J, \u0160roubek F, \u0160m\u00eddl V (2021) Improving neural blind deconvolution. In: 2021 IEEE International Conference on Image Processing (ICIP). IEEE, pp 1954\u20131958","DOI":"10.1109\/ICIP42928.2021.9506502"},{"key":"10642_CR17","doi-asserted-by":"crossref","unstructured":"Lasmar NE, Stitou Y, Berthoumieu Y (2009) Multiscale skewed heavy tailed model for texture analysis. In: Proc. IEEE Int. Conf. Image Process., pp 2281\u20132284","DOI":"10.1109\/ICIP.2009.5414404"},{"issue":"10","key":"10642_CR18","doi-asserted-by":"publisher","first-page":"105003","DOI":"10.1088\/1361-6560\/ab8105","volume":"65","author":"Y-C Lin","year":"2020","unstructured":"Lin Y-C, Huang H-M (2020) Denoising of multi b-value diffusion-weighted MR images using deep image prior. Phys Med Biol 65(10):105003","journal-title":"Phys Med Biol"},{"key":"10642_CR19","unstructured":"Li H, Xu Z, Taylor G, Studer C, Goldstein T (2018) Visualizing the loss landscape of neural nets. In: Advances in neural information processing systems, vol 31"},{"key":"10642_CR20","doi-asserted-by":"crossref","unstructured":"Li T, Zhuang Z, Liang H, Peng L, Wang H, Sun J (2021) Self-validation: early stopping for single-instance deep generative priors. In: 32nd British Machine Vision Conference 2021, BMVC 2021, Online, November 22\u201325, 2021. BMVA Press, p 108","DOI":"10.5244\/C.35.432"},{"key":"10642_CR21","doi-asserted-by":"crossref","unstructured":"Martin D, Fowlkes C, Tal D, Malik J (2001) A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics. In: Proc. 8th Int\u2019l Conf. Computer Vision, vol 2, pp 416\u2013423","DOI":"10.1109\/ICCV.2001.937655"},{"key":"10642_CR22","doi-asserted-by":"publisher","first-page":"743","DOI":"10.1007\/s10915-017-0460-5","volume":"74","author":"J-J Mei","year":"2018","unstructured":"Mei J-J, Dong Y, Huang T-Z, Yin W (2018) Cauchy noise removal by nonconvex ADMM with convergence guarantees. J Sci Comput 74:743\u2013766","journal-title":"J Sci Comput"},{"issue":"12","key":"10642_CR23","doi-asserted-by":"publisher","first-page":"4695","DOI":"10.1109\/TIP.2012.2214050","volume":"21","author":"A Mittal","year":"2012","unstructured":"Mittal A, Moorthy AK, Bovik AC (2012) No-reference image quality assessment in the spatial domain. IEEE Trans Image Process 21(12):4695\u20134708","journal-title":"IEEE Trans Image Process"},{"key":"10642_CR24","doi-asserted-by":"crossref","unstructured":"Nuttall AH (1983) Accurate efficient evaluation of cumulative or exceedance probability distributions directly from characteristic functions. Naval Underwater Systems Center, New London, CT, Tech. Rep. ADA133703","DOI":"10.21236\/ADA133703"},{"issue":"16","key":"10642_CR25","doi-asserted-by":"publisher","first-page":"11895","DOI":"10.1007\/s00500-023-08557-3","volume":"27","author":"N Renuka","year":"2023","unstructured":"Renuka N (2023) Semantic segmentation-based skin cancer detection. Soft Comput 27(16):11895\u201311903","journal-title":"Soft Comput"},{"key":"10642_CR26","doi-asserted-by":"crossref","unstructured":"Ronneberger O, Fischer P, Brox T (2015) U-net: convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-assisted intervention\u2014MICCAI 2015: 18th International Conference, Munich, Germany, October 5\u20139, 2015, Proceedings, Part III, vol 18. Springer, pp 234\u2013241","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"10642_CR27","doi-asserted-by":"crossref","unstructured":"Ruderman DL (1994) The statistics of natural images. Netw Comput Neural Syst 5(4):517\u2013548","DOI":"10.1088\/0954-898X\/5\/4\/006"},{"issue":"3","key":"10642_CR28","doi-asserted-by":"publisher","first-page":"1894","DOI":"10.1137\/140997816","volume":"8","author":"F Sciacchitano","year":"2015","unstructured":"Sciacchitano F, Dong Y, Zeng T (2015) Variational approach for restoring blurred images with Cauchy noise. SIAM J Imaging Sci 8(3):1894\u20131922","journal-title":"SIAM J Imaging Sci"},{"issue":"1","key":"10642_CR29","doi-asserted-by":"publisher","first-page":"52","DOI":"10.1109\/76.350779","volume":"5","author":"K Sharifi","year":"1995","unstructured":"Sharifi K, Leon-Garcia A (1995) Estimation of shape parameter for generalized gaussian distributions in subband decompositions of video. IEEE Trans Circuits Syst Video Technol 5(1):52\u201356","journal-title":"IEEE Trans Circuits Syst Video Technol"},{"issue":"5","key":"10642_CR30","doi-asserted-by":"publisher","first-page":"95","DOI":"10.1109\/MSP.2020.3004124","volume":"37","author":"R Sun","year":"2020","unstructured":"Sun R, Li D, Liang S, Ding T, Srikant R (2020) The global landscape of neural networks: an overview. IEEE Signal Process Mag 37(5):95\u2013108","journal-title":"IEEE Signal Process Mag"},{"key":"10642_CR31","unstructured":"Top 100 Hubble Telescope Images. https:\/\/www.kaggle.com\/datasets\/redwankarimsony\/top-100-hubble-telescope-images"},{"key":"10642_CR32","unstructured":"Top 100 images ESA-Hubble. https:\/\/esahubble.org\/images\/archive\/top100\/"},{"key":"10642_CR33","doi-asserted-by":"crossref","unstructured":"Torbunov D, Huang Y, Yu H, Huang J, Yoo S, Lin M, Viren B, Ren Y (2023) Uvcgan: Unet vision transformer cycle-consistent gan for unpaired image-to-image translation. In: Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV), pp 702\u2013712","DOI":"10.1109\/WACV56688.2023.00077"},{"key":"10642_CR34","doi-asserted-by":"publisher","first-page":"178","DOI":"10.1016\/j.patrec.2021.01.036","volume":"145","author":"K Trebing","year":"2021","unstructured":"Trebing K, Sta\u0144czyk T, Mehrkanoon S (2021) SmaAt-UNet: precipitation nowcasting using a small attention-UNet architecture. Pattern Recognit Lett 145:178\u2013186","journal-title":"Pattern Recognit Lett"},{"key":"10642_CR35","doi-asserted-by":"crossref","unstructured":"Ulyanov D, Vedaldi A, Lempitsky V (2018) Deep image prior. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 9446\u20139454","DOI":"10.1109\/CVPR.2018.00984"},{"key":"10642_CR36","doi-asserted-by":"publisher","first-page":"1867","DOI":"10.1007\/s11263-020-01303-4","volume":"128","author":"D Ulyanov","year":"2020","unstructured":"Ulyanov D, Vedaldi A, Lempitsky V (2020) Deep Image Prior. Int J Comput Vis 128:1867\u20131888","journal-title":"Int J Comput Vis"},{"key":"10642_CR37","unstructured":"Wang H, Li T, Zhuang Z, Chen T, Liang H, Sun J (2023) Early stopping for deep image prior. Trans Mach Learn Res 2023. https:\/\/www.scopus.com\/record\/display.uri?eid=2-s2.0-86000639497&origin=inward &txGid=69e3899f7221f3143b89f40954ba7e44"},{"issue":"12","key":"10642_CR38","doi-asserted-by":"publisher","first-page":"3337","DOI":"10.1109\/TMI.2021.3084288","volume":"40","author":"J Yoo","year":"2021","unstructured":"Yoo J, Jin KH, Gupta H, Yerly J, Stuber M, Unser M (2021) Time-dependent deep image prior for dynamic MRI. IEEE Trans Med Imaging 40(12):3337\u20133348","journal-title":"IEEE Trans Med Imaging"},{"issue":"4","key":"10642_CR39","doi-asserted-by":"publisher","first-page":"045010","DOI":"10.1088\/0266-5611\/25\/4\/045010","volume":"25","author":"R Zanella","year":"2009","unstructured":"Zanella R, Boccacci P, Zanni L, Bertero M (2009) Efficient gradient projection methods for edge-preserving removal of Poisson noise. Inverse Probl 25(4):045010","journal-title":"Inverse Probl"}],"container-title":["Soft Computing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00500-025-10642-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00500-025-10642-8\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00500-025-10642-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,17]],"date-time":"2025-06-17T06:18:59Z","timestamp":1750141139000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00500-025-10642-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4]]},"references-count":39,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2025,4]]}},"alternative-id":["10642"],"URL":"https:\/\/doi.org\/10.1007\/s00500-025-10642-8","relation":{"has-preprint":[{"id-type":"doi","id":"10.21203\/rs.3.rs-4396753\/v1","asserted-by":"object"}]},"ISSN":["1432-7643","1433-7479"],"issn-type":[{"value":"1432-7643","type":"print"},{"value":"1433-7479","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4]]},"assertion":[{"value":"6 February 2025","order":1,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 May 2025","order":2,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors have no relevant financial or non-financial interests to disclose. The authors have no Conflict of interest to declare that are relevant to the content of this article.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval and consent to participate"}}]}}