{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T19:50:49Z","timestamp":1792785049897,"version":"4.5.0"},"reference-count":28,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,7,9]],"date-time":"2026-07-09T00:00:00Z","timestamp":1783555200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"vor","delay-in-days":92,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Budapest University of Technology and Economics","id":[{"id":"https:\/\/ror.org\/02w42ss30","id-type":"ROR","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    The effectiveness of deep learning models is strongly influenced by the quality of training data. Traditional training approaches assume that all samples contribute equally to the learning process, leading to uniform data sampling. However, this assumption overlooks the substantial variation in informational content across samples. This paper introduces a novel and computationally lightweight data prioritization methodology for object detection that dynamically adjusts the sampling probability of training data according to its relevance during learning.The proposed Relative Detection Error (RDE) is a new temporal instability metric defined in a joint classification\u2013regression setting, where classification corresponds to object category prediction and regression corresponds to bounding-box parameter estimation. By explicitly quantifying prediction variability over time, RDE identifies samples with higher learning value and guides an exploration-regularized stochastic sampling policy. As a result, the method improves both classification and localization accuracy while imposing minimal computational overhead and integrating seamlessly into standard training pipelines. The approach is validated using YOLO architectures on diverse datasets, demonstrating consistent improvements and strong generalization across models and domains. Experimental results show that prioritizing high-value samples yields higher F1 scores and mean Average Precision, alongside more efficient and stable convergence. The project source code is available at:\n                    <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"https:\/\/kp-labs-bme.github.io\/Object-Detection-Prioritization\/\" ext-link-type=\"uri\">https:\/\/kp-labs-bme.github.io\/Object-Detection-Prioritization\/<\/jats:ext-link>\n                  <\/jats:p>","DOI":"10.1038\/s41598-026-60420-1","type":"journal-article","created":{"date-parts":[[2026,7,9]],"date-time":"2026-07-09T11:07:12Z","timestamp":1783595232000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Uncertainty aware stochastic sampling for efficient object detection"],"prefix":"10.1038","volume":"16","author":[{"given":"Csan\u00e1d Levente","family":"Balogh","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bence","family":"Pap","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"B\u00e1lint","family":"K\u0151v\u00e1ri","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1487-9672","authenticated-orcid":false,"given":"Tam\u00e1s","family":"B\u00e9csi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,7,9]]},"reference":[{"key":"60420_CR1","doi-asserted-by":"crossref","unstructured":"Purkrabek, M. & Matas, J. Detection, pose estimation and segmentation for multiple bodies: Closing the virtuous circle (2024). arXiv:2412.01562.","DOI":"10.1109\/ICCV51701.2025.00842"},{"key":"60420_CR2","doi-asserted-by":"publisher","first-page":"1680","DOI":"10.3390\/make5040083","volume":"5","author":"J Terven","year":"2023","unstructured":"Terven, J., C\u00f3rdova-Esparza, D.-M. & Romero-Gonz\u00e1lez, J.-A. A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-NAS. Mach. Learn. Knowl. Extr. 5, 1680\u20131716. https:\/\/doi.org\/10.3390\/make5040083 (2023).","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"60420_CR3","doi-asserted-by":"publisher","first-page":"57815","DOI":"10.1109\/ACCESS.2024.3386826","volume":"12","author":"MG Ragab","year":"2024","unstructured":"Ragab, M. G. et al. A comprehensive systematic review of yolo for medical object detection (2018 to 2023). IEEE Access 12, 57815\u201357836. https:\/\/doi.org\/10.1109\/ACCESS.2024.3386826 (2024).","journal-title":"IEEE Access"},{"key":"60420_CR4","doi-asserted-by":"publisher","first-page":"945","DOI":"10.1007\/s10462-021-09997-9","volume":"55","author":"B Singh","year":"2022","unstructured":"Singh, B., Kumar, R. & Singh, V. P. Reinforcement learning in robotic applications: a comprehensive survey. Artif. Intell. Rev. 55, 945\u2013990. https:\/\/doi.org\/10.1007\/s10462-021-09997-9 (2022).","journal-title":"Artif. Intell. Rev."},{"key":"60420_CR5","doi-asserted-by":"publisher","first-page":"94","DOI":"10.3311\/PPtr.37089","volume":"53","author":"A Mih\u00e1ly","year":"2025","unstructured":"Mih\u00e1ly, A. et al. Linear parameter varying and reinforcement learning approaches for trajectory tracking controller of autonomous vehicles. Periodica Polytechnica Transp. Eng. 53, 94\u2013102. https:\/\/doi.org\/10.3311\/PPtr.37089 (2025).","journal-title":"Periodica Polytechnica Transp. Eng."},{"key":"60420_CR6","doi-asserted-by":"publisher","first-page":"529","DOI":"10.1038\/nature14236","volume":"518","author":"V Mnih","year":"2015","unstructured":"Mnih, V. et al. Human-level control through deep reinforcement learning. Nature 518, 529\u2013533. https:\/\/doi.org\/10.1038\/nature14236 (2015).","journal-title":"Nature"},{"key":"60420_CR7","doi-asserted-by":"crossref","unstructured":"Caicedo, J.\u00a0C. & Lazebnik, S. Active object localization with deep reinforcement learning. In Proc. of the IEEE International Conference on Computer Vision (ICCV) (2015).","DOI":"10.1109\/ICCV.2015.286"},{"key":"60420_CR8","doi-asserted-by":"publisher","first-page":"6390","DOI":"10.48550\/arXiv.2105.02340","volume":"34","author":"D Dablain","year":"2023","unstructured":"Dablain, D., Krawczyk, B. & Chawla, N. V. Deepsmote: Fusing deep learning and smote for imbalanced data. IEEE Transac. Neural Netw. Learn. Syst. 34, 6390\u20136404. https:\/\/doi.org\/10.48550\/arXiv.2105.02340 (2023).","journal-title":"IEEE Transac. Neural Netw. Learn. Syst."},{"key":"60420_CR9","doi-asserted-by":"publisher","first-page":"791","DOI":"10.1007\/s00778-022-00775-9","volume":"32","author":"SE Whang","year":"2023","unstructured":"Whang, S. E. et al. Data collection and quality challenges in deep learning: A data-centric ai perspective. VLDB J. 32, 791\u2013813. https:\/\/doi.org\/10.1007\/s00778-022-00775-9 (2023).","journal-title":"VLDB J."},{"key":"60420_CR10","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1007\/s10462-024-10813-3","volume":"57","author":"CL Balogh","year":"2024","unstructured":"Balogh, C. L., Pelenczei, B., K\u0151v\u00e1ri, B. & B\u00e9csi, T. Strategic data navigation: Information value-based sample selection. Artif. Intell. Rev. 57, 187. https:\/\/doi.org\/10.1007\/s10462-024-10813-3 (2024).","journal-title":"Artif. Intell. Rev."},{"key":"60420_CR11","doi-asserted-by":"publisher","first-page":"64","DOI":"10.3390\/axioms14010064","volume":"14","author":"U Daraz","year":"2025","unstructured":"Daraz, U., Agustiana, D., Wu, J. & Emam, W. Twofold auxiliary information under two-phase sampling: An improved family of double-transformed variance estimators. Axioms 14, 64. https:\/\/doi.org\/10.3390\/axioms14010064 (2025).","journal-title":"Axioms"},{"key":"60420_CR12","doi-asserted-by":"publisher","DOI":"10.1016\/j.heliyon.2024.e33402","volume":"10","author":"U Daraz","year":"2024","unstructured":"Daraz, U., Wu, J., Alomair, M. A. & Aldoghan, L. A. New classes of difference cum-ratio-type exponential estimators for a finite population variance in stratified random sampling. Heliyon 10, e33402. https:\/\/doi.org\/10.1016\/j.heliyon.2024.e33402 (2024).","journal-title":"Heliyon"},{"key":"60420_CR13","doi-asserted-by":"publisher","first-page":"957","DOI":"10.3390\/sym16080957","volume":"16","author":"U Daraz","year":"2024","unstructured":"Daraz, U., Alomair, M. A. & Albalawi, O. Variance estimation under some transformation for both symmetric and asymmetric data. Symmetry 16, 957. https:\/\/doi.org\/10.3390\/sym16080957 (2024).","journal-title":"Symmetry"},{"key":"60420_CR14","doi-asserted-by":"publisher","first-page":"2839","DOI":"10.3390\/math12182839","volume":"12","author":"MA Alomair","year":"2024","unstructured":"Alomair, M. A. & Daraz, U. Dual transformation of auxiliary variables by using outliers in stratified random sampling. Mathematics 12, 2839. https:\/\/doi.org\/10.3390\/math12182839 (2024).","journal-title":"Mathematics"},{"key":"60420_CR15","doi-asserted-by":"crossref","unstructured":"Freund, Y. & Schapire, R.\u00a0E. A desicion-theoretic generalization of on-line learning and an application to boosting. In Computational Learning Theory, (Vit\u00e1nyi, P. ed) 23\u201337 (Springer Berlin Heidelberg, Berlin, Heidelberg, 1995).","DOI":"10.1007\/3-540-59119-2_166"},{"key":"60420_CR16","volume-title":"Explaining AdaBoost, 37\u201352","author":"RE Schapire","year":"2013","unstructured":"Schapire, R. E. Explaining AdaBoost, 37\u201352 (Springer, Berlin Heidelberg, Berlin, Heidelberg, 2013)."},{"key":"60420_CR17","doi-asserted-by":"publisher","first-page":"358","DOI":"10.1016\/j.ins.2021.03.042","volume":"563","author":"W Wang","year":"2021","unstructured":"Wang, W. & Sun, D. The improved adaboost algorithms for imbalanced data classification. Inf. Sci. 563, 358\u2013374. https:\/\/doi.org\/10.1016\/j.ins.2021.03.042 (2021).","journal-title":"Inf. Sci."},{"key":"60420_CR18","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1007\/s10994-021-06003-9","volume":"111","author":"VL Nguyen","year":"2022","unstructured":"Nguyen, V. L., Shaker, M. H. & H\u00fcllermeier, E. How to measure uncertainty in uncertainty sampling for active learning. Mach. Learn. 111, 89\u2013122. https:\/\/doi.org\/10.1007\/s10994-021-06003-9 (2022).","journal-title":"Mach. Learn."},{"key":"60420_CR19","doi-asserted-by":"publisher","first-page":"113","DOI":"10.1007\/s11263-014-0781-x","volume":"113","author":"Y Yang","year":"2015","unstructured":"Yang, Y. et al. Multi-class active learning by uncertainty sampling with diversity maximization. Int. J. Comput. Vision 113, 113\u2013127. https:\/\/doi.org\/10.1007\/s11263-014-0781-x (2015).","journal-title":"Int. J. Comput. Vision"},{"key":"60420_CR20","doi-asserted-by":"publisher","first-page":"6453","DOI":"10.1007\/s10994-024-06567-2","volume":"113","author":"A Hoarau","year":"2024","unstructured":"Hoarau, A. et al. Evidential uncertainty sampling strategies for active learning. Mach. Learn. 113, 6453\u20136474. https:\/\/doi.org\/10.1007\/s10994-024-06567-2 (2024).","journal-title":"Mach. Learn."},{"key":"60420_CR21","doi-asserted-by":"publisher","unstructured":"Beluch, W.\u00a0H., Genewein, T., Nurnberger, A. & Kohler, J.\u00a0M. The power of ensembles for active learning in image classification. In 2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition, https:\/\/doi.org\/10.1109\/CVPR.2018.00976 (2018).","DOI":"10.1109\/CVPR.2018.00976"},{"key":"60420_CR22","doi-asserted-by":"crossref","unstructured":"Yoo, D. & Kweon, I.\u00a0S. Learning loss for active learning. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2019).","DOI":"10.1109\/CVPR.2019.00018"},{"key":"60420_CR23","unstructured":"Schaul, T., Quan, J., Antonoglou, I. & Silver, D. Prioritized experience replay (2016). arXiv:1511.05952."},{"key":"60420_CR24","unstructured":"Horgan, D. Distributed prioritized experience replay (2018) arXiv:1803.00933."},{"key":"60420_CR25","unstructured":"Brittain, M., Bertram, J., Yang, X. & Wei, P. Prioritized sequence experience replay (2020). arXiv:1905.12726."},{"key":"60420_CR26","doi-asserted-by":"publisher","unstructured":"K?v\u00e1ri, B., Pelenczei, B. & B\u00e9csi, T.,. Enhanced experience prioritization: A novel upper confidence bound approach. IEEE Access 11, 138488\u2013138501. https:\/\/doi.org\/10.1109\/ACCESS.2023.3339248 (2023).","DOI":"10.1109\/ACCESS.2023.3339248"},{"key":"60420_CR27","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1007\/BF00115009","volume":"3","author":"R Sutton","year":"1988","unstructured":"Sutton, R. Learning to predict by the method of temporal differences. Mach. Learn. 3, 9\u201344. https:\/\/doi.org\/10.1007\/BF00115009 (1988).","journal-title":"Mach. Learn."},{"key":"60420_CR28","doi-asserted-by":"crossref","unstructured":"Geiger, A., Lenz, P. & Urtasun, R. Are we ready for autonomous driving? the kitti vision benchmark suite. In Conference on Computer Vision and Pattern Recognition (CVPR) (2012).","DOI":"10.1109\/CVPR.2012.6248074"}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-60420-1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-60420-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-60420-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,8]],"date-time":"2026-10-08T22:53:46Z","timestamp":1791500026000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-026-60420-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,9]]},"references-count":28,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["60420"],"URL":"https:\/\/doi.org\/10.1038\/s41598-026-60420-1","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,9]]},"assertion":[{"value":"12 December 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"29 June 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"9 July 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare no competing interests.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"31556"}}