Modification of a neural network algorithm for highly reliable iris-based biometric authentication
DOI: 10.21293/1818-0442-2026-29-2-148-159
Abstract: Relevance. The growing number of systems processing confidential data calls for authentication mechanisms that remain secure when the stored reference data is compromised. Purpose of the study. To modify a neural network algorithm for highly reliable iris-based biometric authentication, enabling the cryptographic key recovery with an exact bit-wise match. Methods. A convolutional neural network is used to extract the iris region, followed by crop-based normalization. Joint training of a convolutional model for feature extraction and a multilayer perceptron is performed using a combined loss function comprising metric terms and two terms acting directly on the bit-wise output of the model. The operating point is selected according to an admissibility criterion, and Hamming‐neighborhood search is employed at the verification stage. Scientific novelty. Unlike known schemes, exact key recovery is achieved without auxiliary data or external error correction, allowing only the hash of the key to be stored. Results. For 100 enrolled users, the genuine acceptance rate among the instances admitted for operation was 0.875 with a zero false acceptance probability over 8000 impostor presentations. Searching within a radius of two bits reduces the type I error probability to 0.107 without changing the type II error probability. Rejection on both attempts under exact key match was observed for 1 out of 100 users. It is shown that the margin secured on the threshold‐selection group transfers to the test set only in part. Practical significance. The algorithm is applicable in systems where the storage of biometric reference data and cryptographic keys in plaintext is unacceptable.
Keywords: biometric authentication, iris recognition, image segmentation, neural networks, cryptographic key, Triplet Loss, Center Loss, biometric cryptography
For citation:
Shtuka S. E., Borshevnikov A. E. Modification of a neural network algorithm for highly reliable iris-based biometric authentication. Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki, 2026, vol. 29, no. 2, pp. 148–159. DOI: 10.21293/1818-0442-2026-29-2-148-159
Authors and copyright holders:
- Shtuka S. E. , Far Eastern Federal University (FEFU)
- Borshevnikov A. E. , Industrial Cryptosystems JSC (ProCSy JSC)
- 1. Shelupanov A., Evsyutin O., Konev A., Kostyuchenko E., Kruchinin D., Nikiforov D. Information security methods – modern research directions. Symmetry, 2019, vol. 11, no. 2, art. 150.
- 2. Nguyen K., Proença H., Alonso-Fernandez F. Deep learning for iris recognition: a survey. ACM Computing Surveys, 2024, vol. 56, no. 9, pp. 1–35.
- 3. GOST R 52633.0-2006. Zashchita informatsii. Tekhnika zashchity informatsii. Trebovaniya k sredstvam vysokonadezhnoy biometricheskoy autentifikatsii [Information protection. Information protection technique. Requirements for means of high-reliability biometric authentication]. Moscow, Standartinform Publ., 2020 (in Russ.).
- 4. Müller D., Soto-Rey I., Kramer F. Towards a guideline for evaluation metrics in medical image segmentation. BMC Research Notes, 2022, vol. 15, no. 1, art. 210.
- 5. Mayya A., Saii M. Iris and palmprint decision fusion to enhance human recognition. International Journal of Computer Science Trends and Technology, 2017, vol. 5, no. 5, pp. 42–46.
- 6. Semyonov M.S., Myasnikov E.V. Sravnenie metodov segmentatsii izobrazheniya raduzhnoy obolochki glaza [Comparison of iris image segmentation methods]. Informatsionnye tehnologii i nanotekhnologii (ITNT-2018) [Information Technologies and Nanotechnologies]. Proc. of the IV Int. Conf., Samara, 24–27 April 2018, Samara, Samara National Research University named after Academician S.P. Korolev, 2018, pp. 814–821 (in Russ.).
- 7. Daugman J.G. High confidence visual recognition of persons by a test of statistical independence. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1993, vol. 15, no. 11, pp. 1148–1161.
- 8. Chigrinskiy V.V., Efimov Yu.S., Matveev I.A. Bystryy algoritm poiska granits zrachka i raduzhnoy obolochki glaza [A fast algorithm for locating pupil and iris boundaries]. Mashinnoe obuchenie i analiz dannykh, 2016, vol. 2, no. 2, pp. 159–172 (in Russ.).
- 9. Ganeeva Yu.Kh., Myasnikov E.V. Segmentatsiya raduzhnoy obolochki glaza na izobrazhenii s pomoshchyu svertochnoy neyronnoy seti arkhitektury U-Net [Iris segmentation in an image using a U-Net convolutional neural network]. Informatsionnye tekhnologii i nanotekhnologii (ITNT-2020). [Information Technologies and Nanotechnologies] Proc. of the VI Int. Conf., Samara, 26–29 May 2020, Samara, Samara National Research University named after Academician S.P. Korolev, 2020, vol. 2, pp. 57–63 (in Russ.).
- 10. Ronneberger O., Fischer P., Brox T. U-Net: convolutional networks for biomedical image segmentation. Proc. of the 18th Int. Conf.: Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, Munich, Germany, Cham, Springer, 2015, pp. 234–241.
- 11. Lozej J., Meden B., Struc V., Peer P. End-to-end iris segmentation using U-Net. 2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI), San Carlos, Costa Rica, IEEE, 2018, pp. 1–6. DOI: 10.1109/IWOBI.2018.8464213.
- 12. Zhang W., Lu X., Gu Y., Liu Y., Meng X., Li J. A robust iris segmentation scheme based on improved U-Net. IEEE Access, 2019, vol. 7, pp. 85082–85089.
- 13. Wu X., Zhao L. Study on iris segmentation algorithm based on dense U-Net. IEEE Access, 2019, vol. 7, pp. 123959–123968.
- 14. Li Y.H., Putri W.R., Aslam M.S., Chang C.C. Robust iris segmentation algorithm in non-cooperative environments using interleaved residual U-Net. Sensors, 2021, vol. 21, no. 4, p. 1434.
- 15. Miron C., Pasarica A., Manta V., Timofte R. Efficient and robust eye images iris segmentation using a lightweight U-Net convolutional network. Multimedia Tools and Applications, 2022, vol. 81, no. 11, pp. 14961–14977.
- 16. Farmanifard P., Ross A. Iris-SAM: iris segmentation using a foundation model. Pattern Recognition and Artificial Intelligence. Proc. of the 4th Int. Conf. ICPRAI 2024, Jeju Island, South Korea, Singapore, Springer Nature Singapore, 2024, pp. 394–409.
- 17. Wei J., Gao X., Wang Y., He R., Sun Z. Uncertainty-aware bilateral transformer for accurate and reliable iris segmentation. IEEE Transactions on Information Forensics and Security, 2025, vol. 20, pp. 5919–5933.
- 18. Jiang J., Zhang Q., Wang C. SAM-Iris: a SAM-based iris segmentation algorithm. Electronics, 2025, vol. 14, no. 2, p. 246.
- 19. Liu Y., Shen W., Wu D., Shao J. IrisST-Net for iris segmentation and contour parameters extraction. Applied Intelligence, 2023, vol. 53, no. 9, pp. 11267–11281.
- 20. Ganeeva Y.K., Myasnikov E.V. Identifying persons from iris images using neural networks for image segmentation and feature extraction. Computer Optics, 2022, vol. 46, no. 2, pp. 308–316.
- 21. Shtuka S.E., Borshevnikov A.E. Razrabotka neyrosetevogo algoritma vysokonadezhnoy biometricheskoy autentifikatsii na osnove raduzhnoy obolochki glaza [Development of a neural network algorithm for high-reliability biometric authentication based on the iris]. Regionalnaya informatika i informatsionnaya bezopasnost. [Regional Informatics and Information Security] Proc. of the XIV St. Petersburg Interregional Conf., St. Petersburg, 29–31 October 2025, St. Petersburg, SPOISU Publ., 2025, pp. 204–206 (in Russ).
- 22. Demirel H., Anbarjafari G. Iris recognition system using combined histogram statistics. 2008 23rd International Symposium on Computer and Information Sciences (ISCIS), Istanbul, Turkey, IEEE, 2008, pp. 1–4.
- 23. Monro D.M., Rakshit S., Zhang D. DCT-based iris recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, vol. 29, no. 4, pp. 586–595.
- 24. Balashanmugam T., Sengottaiyan K., Kulandairaj M.S., Dang H. An effective model for the iris regional characteristics and classification using deep learning Alex network. IET Image Processing, 2023, vol. 17, no. 1, pp. 227–238.
- 25. Sun J., Zhao S., Yu Y., Wang X., Zhou L. Iris recognition based on local circular Gabor filters and multi-scale convolution feature fusion network. Multimedia Tools and Applications, 2022, vol. 81, no. 23, pp. 33051–33065.
- 26. Li Y., Chen C.L.P., Zhang T. A survey on Siamese network: methodologies, applications, and opportunities. IEEE Transactions on Artificial Intelligence, 2022, vol. 3, no. 6, pp. 994–1014.
- 27. Latif S.A., Sidek K.A., Hashim A.H.A. Hybrid CNN-ViT integration into Siamese networks for robust iris biometric verification. Bulletin of Electrical Engineering and Informatics, 2025, vol. 14, no. 4, pp. 3101–3109.
- 28. GOST R 52633.5-2011. Zashchita informatsii. Tekhnika zashchity informatsii. Avtomaticheskoe obuchenie neyrosetevykh preobrazovateley biometriya-kod dostupa [Information protection. Information protection technique. Automatic training of neural network converters for biometrics-access code]. Moscow, Standartinform Publ., 2018 (in Russ.).
- 29. Zhou X., Kuijper A., Veldhuis R., Busch C. Quantifying privacy and security of biometric fuzzy commitment. 2011 International Joint Conference on Biometrics (IJCB), Washington, DC, USA, IEEE, 2011, pp. 1–8.
- 30. Kausar F. Iris based cancelable biometric cryptosystem for secure healthcare smart card. Egyptian Informatics Journal, 2021, vol. 22, no. 4, pp. 447–453.
- 31. Matveev I., Safonov I. From iris image to embedded code: system of methods. Algorithms, 2023, vol. 16, no. 2, art. 87.
- 32. Lin K.-C., Chen Y.-M. A high-security-level iris cryptosystem based on fuzzy commitment and soft reliability extraction. IEEE Transactions on Dependable and Secure Computing, 2024, vol. 21, no. 4, pp. 1770–1784.
- 33. Shukla A., Demarest L., Fuller B., Ahmad S., Manicke C., Russell A., Chen S. Fuzzy extractors are practical: cryptographic strength key derivation from the iris. Proc. of the 2025 ACM SIGSAC Conf. on Computer and Communications Security (CCS '25), Taipei, Taiwan, New York, ACM, 2025, pp. 3605–3619.
- 34. Schroff F., Kalenichenko D., Philbin J. FaceNet: a unified embedding for face recognition and clustering. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, USA, IEEE, 2015, pp. 815–823.
- 35. Wen Y., Zhang K., Li Z., Qiao Y. A discriminative feature learning approach for deep face recognition. Computer Vision – ECCV 2016. Proc. of the 14th European Conf., Amsterdam, The Netherlands, Cham, Springer, 2016, pp. 499–515.
Last issue
Editorial office address
Executive Secretary of the Editor’s Office
Editor’s Office: 40 Lenina Prospect, Tomsk, 634050, Russia
Phone / Fax: + 7 (3822) 701-582
Executive Secretary of the Editor’s Office
Editor’s Office: 40 Lenina Prospect, Tomsk, 634050, Russia
Phone / Fax: + 7 (3822) 51-21-21 / 51-43-02
Subscription for updates
