Development of biometric identification tools for solving information security problems based on multimodal neural  networks

Anna Poluyan; Sofya Petrenkova; Kseniia Korovina

doi:10.57239/PJLSS-2025-23.2.00218

Author(s)

Anna Poluyan Don State Technical University,Rostov-on-Don,Russia

Sofya Petrenkova Don State Technical University,Rostov-on-Don,Russia

Kseniia Korovina Don State Technical University,Rostov-on-Don,Russia

Abstract

Biometric identification is currently one of the most pressing issues in information security. Biometric identification, unlike traditional methods (passwords, tokens), offers a fundamentally different approach to authentication based on the uniqueness of human physiological and behavioral characteristics. Its implementation is accelerating in the commercial sector (for example, fingerprint payments), at the government level (electronic passports with biometrics) and in consumer electronics (facial recognition in smartphones). According to Juniper Research, by 2026, more than 4 billion devices will use biometric identification. The biometric technology market, according to MarketsandMarkets, will reach $82.9 billion by 2027, reflecting the growing demand for reliable and convenient solutions to ensure the security of biometric data. However, traditional methods are vulnerable to, for example, spoofing, adversarial attacks (fakes and synthesized voices), noise or low data quality. The article proposes measures to solve the problem of biometric identification using artificial intelligence tools, namely multimodal language models, which are currently one of the best methods in the field of machine learning. A comparative analysis is conducted and the advantages of using a multimodal approach compared to unimodal systems are indicated. Data protection measures are proposed and the effectiveness of this approach is assessed.

Publication Details

Page(s) 872-880

DOI 10.57239/PJLSS-2025-23.2.00218

Published Journal: Pakistan Journal of Life and Social Sciences, Volume: 23, Issue: 2, Year: 2025

Request the full text from PASTIC

Keywords

Authentication Biometric identification Biometrics Vulnerabilities Multimodality Unimodality

References

Bal H. .2024 .Deep dive into IDS-IPS architecture: Building a robust network defense. Hostomize, : .

Biggio Demetrio L.,Lagorio B.,Armando G.,A. G. .1810 .Exploring adversarial examples in malware detection. , : .

Eykholt K.,Evtimov I.,Fernandes E.,Li B.,Rahmati A.,Xiao C.,Prakash A.,Kohno T.,Song D. .2018 .Robust physical-world attacks on deep learning visual classification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), : 1625-1634.

George A. ,Geissenbuhler Z.,Nikisins D.,Anjos O.,Marcel A.,S. A. .2019 .Biometric face presentation attack detection with multi-channel convolutional neural network. IEEE Transactions on Information Forensics and Security, 15 : 42-55.

Goodfellow I.,Shlens J.,Szegedy C. .2014 .Explaining and harnessing adversarial examples. , : .

Habr .2023 .Methods of attack on AI. Habr Articles. , : .

Hammad M.,Liu Y.,Wang K. .2018 .Multimodal biometric authentication systems using convolutional neural network based on different level fusion of ECG and fingerprint. IEEE Access, 7 : 26527-26542.

Kaspersky .2019 .attacks-on-artificial-intelligencewhitepaper. Attacks on artificial intelligence. Kaspersky Whitepaper, : .

Keanini T.K .2019 .-of-machine-learning-in-2019. Cisco Blogs, : .

Lin T.Y.,Maji S. .2015 .Bilinear CNN models for fine-grained visual recognition. Proceedings of the IEEE International Conference on Computer Vision, : 1449-1457.

Yu Matveev .2012 .Technology of biometric identification of a person by voice and other modalities [Tekhnologiya biometricheskoy identifikatsii lichnosti po golosu i drugim modal'nosti]. Instrument Engineering, 3 : 46-61.

Neubeck A. and Van Gool L. .2006 .Efficient non-maximum suppression. 18th International Conference on Pattern Recognition (ICPR'06), 3 : 850-855.

Chan Park D.S.,Chiu W.,Zoph C.C.,Cubuk B.,Le E.D.,Q.V. E.D. .1904 .SpecAugment: A simple data augmentation method for automatic speech recognition. , : .

Tseligorova Poluyan E.N.,Galushka V.V.,Kodatsky N.M.,N.M. N.M. .2023 .Modern approaches in biometric recognition]. Modern Science: Actual Problems of Theory and Practice. Series: Natural and Technical Sciences, : 123-127.

Powers D.M.W. .2010 .Evaluation: From precision, recall and F-measure to ROC, informedness, markedness and correlation. , : .

Poluyan Purchina O.A.,Fugarov D.D. A.Y.,Don A.Y. .2023 .Rostov-on-. In: Physics and Mechanics of New Materials and Their Applications (PHENMA, : 239-240.

Cyber Skylight .2019 .-kill-you/. , : .

Dabouei Soleymani S.,Kazemi A.,Dawson H.,Nasrabadi J.,N.M. J. .2018 .Multi-level feature abstraction from convolutional neural networks for multimodal biometric identification. 24th International Conference on Pattern Recognition (ICPR), : 3469-3476.

Torfi Soleymani S.,Dawson A.,Nasrabadi J.,N.M. J. .2018 .Generalized bilinear deep convolutional neural networks for multimodal biometric identification. 25th IEEE International Conference on Image Processing, : 763-767.

Topnikov Stefanidi A.F.,Priorov A.I.,A.L. A.I. .2020 .Bimodal identification of personality based on facial and speech biometrics. Proceedings of the 17th International Conference, : 125-129.

Moore R.K Varga A.P.,Processing Signal .1990 .Hidden Markov model decomposition of speech and noise. , : 845-848.