<?xml version="1.0" encoding="UTF-8"?>
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ali="http://www.niso.org/schemas/ali/1.0/" article-type="research-article" dtd-version="1.2" xml:lang="en">
  <front>
    <journal-meta>
      <journal-id journal-id-type="issn">1561-5405</journal-id>
	    <journal-id journal-id-type="doi">10.24151/1561-5405</journal-id>	  
      <journal-id journal-id-type="publisher-id">Proceedings of Universities. Electronics</journal-id>
      <journal-title-group>
        <journal-title xml:lang="en">Scientifical and technical journal "Proceedings of Universities. Electronics"</journal-title>
        <trans-title-group xml:lang="ru">
          <trans-title>Научно-технический журнал «Известия высших учебных заведений. Электроника»</trans-title>
        </trans-title-group>        
      </journal-title-group>      
      <issn publication-format="print">1561-5405</issn>
      <issn publication-format="online">2587-9960</issn>
      <publisher>
        <publisher-name xml:lang="en">National Research University of Electronic Technology</publisher-name>
        <publisher-name xml:lang="ru">Национальный исследовательский университет "Московский институт электронной техники"</publisher-name>
      </publisher>
    </journal-meta>
    <article-meta>                                    
      
    <article-id pub-id-type="doi">10.24151/1561-5405-2026-31-4-487-494</article-id><article-id pub-id-type="risc">YIBZTC</article-id><article-id pub-id-type="udk">004.932.4</article-id><article-categories><subj-group><subject>Информационно-коммуникационные технологии</subject></subj-group></article-categories><title-group><article-title xml:lang="en">Assessment method for biometric quality of facial images</article-title><trans-title-group xml:lang="ru"><trans-title>Метод оценки биометрического качества изображений лиц</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author"><string-name xml:lang="ru">Бордюжа Виктор</string-name><name-alternatives><name xml:lang="ru"><surname>Бордюжа</surname><given-names>Виктор</given-names></name><name xml:lang="en"><surname>Bordiuzha</surname><given-names>Viktor</given-names></name></name-alternatives><string-name xml:lang="en">Viktor Bordiuzha</string-name><xref ref-type="aff" rid="AFF-1"/></contrib><contrib contrib-type="author"><string-name xml:lang="ru">Умняшкин Сергей Владимирович</string-name><name-alternatives><name xml:lang="ru"><surname>Умняшкин</surname><given-names>Сергей Владимирович</given-names></name><name xml:lang="en"><surname>Umnyashkin</surname><given-names>Sergey V.</given-names></name></name-alternatives><string-name xml:lang="en">Sergey V. Umnyashkin</string-name><xref ref-type="aff" rid="AFF-1"/></contrib><aff id="AFF-1" xml:lang="ru">National Research University of Electronic Technology, Russia, 124498, Moscow, Zelenograd, Shokin sq., 1</aff></contrib-group><pub-date iso-8601-date="2026-08-18" date-type="pub" publication-format="electronic"><day>18</day><month>08</month><year>2026</year></pub-date><volume>Том. 31 №4</volume><fpage>487</fpage><lpage>494</lpage><self-uri>http://ivuz-e.ru/en/issues/Том 31 №4/metod_otsenki_biometricheskogo_kachestva_izobrazheniy_lits/</self-uri><self-uri content-type="pdf">http://ivuz-e.ru#</self-uri><abstract xml:lang="en"><p>In automatic face recognition the quality of the input image directly affects the accuracy of face identification and verification. In this work the authors presented a novel method for non-reference biometric evaluation of face image quality based on the analysis of cosine similarity between feature vectors extracted by a neural network recognition model. Unlike general image quality assessment methods like BRISQUE, the proposed approach utilizes both the distance to the class center and the degree of proximity to the most similar external classes, which allows more accurate ranking of images according to their suitability for biometric tasks. By using the AUC metric at different FMR levels, the experiments results on five common datasets (LFW, CALFW, AgeDB-30, CFP-FP, CPLFW) have demonstrated the advantage of the proposed method over modern approaches (CR-FIQA, SDD-FIQA, TOPIQ, etc.), thus it a promising technique for integration into practical biometric systems.</p></abstract><trans-abstract xml:lang="ru"><p>В задачах автоматического распознавания лиц качество входного изображения напрямую влияет на точность идентификации и верификации личности. Возникающие дефекты снижают точность идентификации, особенно при работе с ограниченными базами эталонных образцов. В работе представлен новый метод безэталонной биометрической оценки качества изображений лиц, основанный на анализе косинусного сходства между векторами признаков, извлекаемыми нейросетевой моделью распознавания. В отличие от методов общей оценки качества изображений, например BRISQUE, предложенный метод учитывает как расстояние до центра класса, так и степень близости к наиболее похожим внешним классам, что позволяет более точно ранжировать изображения по степени их пригодности для биометрических задач. При использовании метрики AUC &amp;#40;Area Under the Curve&amp;#41; на разных уровнях FMR результаты экспериментов на пяти распространенных наборах данных &amp;#40;LFW, CALFW, AgeDB-30, CFP-FP, CPLFW&amp;#41; демонстрируют преимущество предложенного метода по сравнению с современными подходами &amp;#40;CR-FIQA, SDD-FIQA, TOPIQ и др.&amp;#41;, что делает его перспективным решением для интеграции в практические биометрические системы.</p></trans-abstract><kwd-group xml:lang="ru"><kwd>biometry</kwd><kwd>face quality assessment</kwd><kwd>face recognition</kwd><kwd>image quality assessment</kwd><kwd>neural networks</kwd></kwd-group><kwd-group xml:lang="en"><kwd>biometry</kwd><kwd>face quality assessment</kwd><kwd>face recognition</kwd><kwd>image quality assessment</kwd><kwd>neural networks</kwd></kwd-group><funding-group/></article-meta>
  </front>
  <body/>
  <back>
    <ref-list><ref id="B1"><label>1.</label><mixed-citation xml:lang="ru">Mittal A., Moorthy A. K., Bovik A. C. No-reference image quality assessment in the spatial domain. IEEE Trans. Image Process. 2012;21(12):4695–4708. https://doi.org/10.1109/TIP.2012.2214050</mixed-citation></ref><ref id="B2"><label>2.</label><mixed-citation xml:lang="ru">Mittal A., Soundararajan R., Bovik A. C. Making a “completely blind” image quality analyzer. IEEE Signal Process. Lett. 2013;20(3):209–212. https://doi.org/10.1109/LSP.2012.2227726</mixed-citation></ref><ref id="B3"><label>3.</label><mixed-citation xml:lang="ru">Бордюжа В., Умняшкин С. В. Ансамблирование метрик качества изображений для повышения корреляции с экспертными оценками. In: Цифровая обработка сигналов и ее применение DSPA–2024: докл. XXVI Междунар. конф. (Москва, 27–29 марта 2024 г.). М.: Рос. науч.-техн. о-во радиотехники, электроники и связи им. А. С. Попова; 2024, с. 187–192. EDN: BTDCNT.</mixed-citation></ref><ref id="B4"><label>5.</label><mixed-citation xml:lang="ru">Bordiuzha V., Umnyashkin S. V. Ensembling image quality metrics to enhance correlation with expert evaluations. In: Tsifrovaya obrabotka signalov i yeye primeneniye DSPA–2024: proc. of 26th International conf. (Moscow, March 27–29, 2024). Moscow: Ros. nauch.-tekhn. o-vo radiotekhniki, elektroniki i svyazi im. A. S. Popova Publ.; 2024, pp. 187–192. (In Russ).</mixed-citation></ref><ref id="B5"><label>4.</label><mixed-citation xml:lang="ru">Бордюжа В., Умняшкин С. В. Ансамблирование безэталонных мер качества изображений для повышения корреляции с экспертными оценками. Изв. вузов. Электроника. 2025;30(2):229–239. https://doi.org/10.24151/1561-5405-2025-30-2-229-239. EDN: EBIJVB.</mixed-citation></ref><ref id="B6"><label>7.</label><mixed-citation xml:lang="ru">Bordiuzha V., Umnyashkin S. V. The ensemble of reference-free image quality measures to improve correlation with expert evaluations. Izv. vuzov. Elektronika = Proc. Univ. Electronics. 2025;30(2):229–239. https://doi.org/10.24151/1561-5405-2025-30-2-229-239</mixed-citation></ref><ref id="B7"><label>5.</label><mixed-citation xml:lang="ru">Yi D., Lei Z., Liao S., Li S. Z. Learning face representation from scratch. arXiv.org.</mixed-citation></ref><ref id="B8"><label>28.</label><mixed-citation xml:lang="ru">11.2014. Available at: https://arxiv.org/pdf/1411.7923 (accessed: 26.06.2025).</mixed-citation></ref><ref id="B9"><label>6.</label><mixed-citation xml:lang="ru">Boutros F., Fang M., Klemt M., Fu B., Damer N. CR-FIQA: Face image quality assessment by learning sample relative classifiability. In: 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver: IEEE; 2023, pp. 5836–5845. https://doi.org/10.1109/CVPR52729.2023.00565</mixed-citation></ref><ref id="B10"><label>7.</label><mixed-citation xml:lang="ru">Ou F.-Z., Chen X., Zhang R., Huang Y., Li S., Li J. et al. SDD-FIQA: Unsupervised face image quality assessment with similarity distribution distance. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Nashville, TN: IEEE; 2021, pp. 7666–7675. https://doi.org/10.1109/CVPR46437.2021.00758</mixed-citation></ref><ref id="B11"><label>8.</label><mixed-citation xml:lang="ru">He K., Zhang X., Ren S., Sun J. Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV: IEEE; 2016, pp. 770–778. https://doi.org/10.1109/CVPR.2016.90</mixed-citation></ref><ref id="B12"><label>9.</label><mixed-citation xml:lang="ru">Deng J., Guo J., Xue N., Zafeiriou S. ArcFace: Additive angular margin loss for deep face recognition. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, CA: IEEE; 2019, pp. 4685–4694. https://doi.org/10.1109/CVPR.2019.00482</mixed-citation></ref><ref id="B13"><label>10.</label><mixed-citation xml:lang="ru">Deng J., Guo J., Ververas E., Kotsia I., Zafeiriou S. RetinaFace: Single-shot multi-level face localisation in the wild. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA: IEEE; 2020, pp. 5202–5211. https://doi.org/10.1109/CVPR42600.2020.00525</mixed-citation></ref><ref id="B14"><label>11.</label><mixed-citation xml:lang="ru">Huang G. B., Mattar M., Berg T., Learned-Miller E. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. Workshop on Faces in ‘Real-Life’ Images: Detection, Alignment, and Recognition (Marseille, France, Oct. 2008).</mixed-citation></ref><ref id="B15"><label>16.</label><mixed-citation xml:lang="ru">09.2008. Available at: https://inria.hal.science/inria-00321923v1 (accessed: 04.05.2026).</mixed-citation></ref><ref id="B16"><label>12.</label><mixed-citation xml:lang="ru">Zheng T., Deng W., Hu J. Cross-Age LFW: A database for studying cross-age face recognition in unconstrained environments. arXiv.org.</mixed-citation></ref><ref id="B17"><label>2017.</label><mixed-citation xml:lang="ru">Available at: https://arxiv.org/abs/1708.08197 (accessed: 16.06.2025). https://doi.org/10.48550/arXiv.1708.08197</mixed-citation></ref><ref id="B18"><label>13.</label><mixed-citation xml:lang="ru">Moschoglou S., Papaioannou A., Sagonas C., Deng J., Kotsia I., Zafeiriou S. AgeDB: The first manually collected, in-the-wild age database. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Honolulu, HI: IEEE; 2017, pp. 1997–2005. https://doi.org/10.1109/CVPRW.2017.250</mixed-citation></ref><ref id="B19"><label>14.</label><mixed-citation xml:lang="ru">Sengupta S., Chen J.-C., Castillo C., Patel V. M., Chellappa R., Jacobs D. W. Frontal to profile face verification in the wild. In: 2016 IEEE Winter Conference on Applications of Computer Vision (WACV). Lake Placid, NY: IEEE; 2016, pp. 1–9. https://doi.org/10.1109/WACV.2016.7477558</mixed-citation></ref><ref id="B20"><label>15.</label><mixed-citation xml:lang="ru">Zheng T., Deng W. Cross-Pose LFW: A database for studying cross-pose face recognition in unconstrained environments. Weihong Deng. 2018. Available at: http://www.whdeng.cn/CPLFW/Cross-Pose-LFW.pdf (accessed: 06.05.2026).</mixed-citation></ref><ref id="B21"><label>16.</label><mixed-citation xml:lang="ru">Chen C., Mo J., Hou J., Wu H., Liao L., Sun W. et al. TOPIQ: A top-down approach from semantics to distortions for image quality assessment. arXiv.org. 2023. Available at: https://arxiv.org/abs/2308.03060 (accessed: 15.06.2024). https://doi.org/10.48550/arXiv.2308.03060</mixed-citation></ref><ref id="B22"><label>17.</label><mixed-citation xml:lang="ru">Hernandez-Ortega J., Galbally J., Fierrez J., Haraksim R., Beslay L. FaceQnet: Quality assessment for face recognition based on deep learning. In: 2019 International Conference on Biometrics (ICB). Crete: IEEE; 2019, pp. 1–8. https://doi.org/10.1109/ICB45273.2019.8987255</mixed-citation></ref><ref id="B23"><label>18.</label><mixed-citation xml:lang="ru">Huang Y., Wang Y., Tai Y., Liu X., Shen P., Li S. CurricularFace: Adaptive curriculum learning loss for deep face recognition. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA: IEEE; 2020, pp. 5900–5909. https://doi.org/10.1109/CVPR42600.2020.00594</mixed-citation></ref><ref id="B24"><label>19.</label><mixed-citation xml:lang="ru">Meng Q., Zhao S., Huang Z., Zhou F. MagFace: A universal representation for face recognition and quality assessment. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Nashville, TN: IEEE; 2021, pp. 14220–14229. https://doi.org/10.1109/CVPR46437.2021.01400</mixed-citation></ref><ref id="B25"><label>20.</label><mixed-citation xml:lang="ru">Grother P., Tabassi E. Performance of biometric quality measures. IEEE Trans. Pattern Anal. Mach. Intell.</mixed-citation></ref><ref id="B26"><label>27.</label><mixed-citation xml:lang="ru">2007;29(4):531–543. https://doi.org/10.1109/TPAMI.2007.1019</mixed-citation></ref></ref-list>    
  </back>
</article>
