{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T06:03:34Z","timestamp":1789625014705,"version":"build-2803163510"},"reference-count":41,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T00:00:00Z","timestamp":1789603200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Traditional methods like passwords and PINs are increasingly vulnerable, making continuous authentication essential. Behavioral biometrics such as mouse dynamics provide a non-intrusive way to verify users through their unique interaction patterns. This study proposes a secure and privacy-preserving framework for mouse dynamics authentication using a Hybrid Transformer-CNN model, where the Transformer captures global behavioral dependencies and the Convolutional Neural Network (CNN) extracts local movement patterns. The framework includes a comprehensive preprocessing pipeline with feature engineering, SMOTE-based class balancing, and stratified cross-validation. To enhance security and privacy, the system integrates adversarial robustness evaluation and federated learning with differential privacy, enabling decentralized training without sharing raw user data. The model is further evaluated under non-IID data distributions and open-set authentication scenarios with per-user thresholds to reflect realistic deployment conditions. Experimental results demonstrate high performance, achieving 99.51% accuracy, 99.53% precision, low FPR (0.000256), and EER (0.000068). The model maintains robustness under adversarial attacks and achieves efficient real-time detection. Overall, the proposed framework provides a reliable and privacy-aware solution for continuous authentication in security-sensitive applications.<\/jats:p>","DOI":"10.3389\/frai.2026.1914353","type":"journal-article","created":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T05:31:02Z","timestamp":1789623062000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Privacy-preserving and robust mouse dynamics authentication using hybrid transformer-CNN and federated learning"],"prefix":"10.3389","volume":"9","author":[{"given":"K.","family":"Sasikumar","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sivakumar","family":"Nagarajan","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Vellore Institute of Technology","place":["Vellore, India"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,17]]},"reference":[{"key":"B1","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/ICCA62237.2024.10928042","article-title":"\u201cDeep learning-driven user legitimacy prediction using keystroke and mouse behavioural dynamics,\u201d","volume-title":"2024 International Conference on Computer and Applications (ICCA)","author":"AbouRida","year":"2024"},{"key":"B2","doi-asserted-by":"publisher","first-page":"224","DOI":"10.25195\/ijci.v51i2.612","article-title":"User authentication based on mouse dynamics using an efficient-net model","volume":"51","author":"Aljoubory","year":"2025","journal-title":"Iraqi J. Comp. Inform"},{"key":"B3","first-page":"205","article-title":"\u201cKeystroke dynamics for continuous authentication,\u201d","year":"2018","journal-title":"2018 8th International Conference on Cloud Computing, Data Science"},{"key":"B4","doi-asserted-by":"publisher","first-page":"8262252","DOI":"10.1049\/bme2\/8262252","article-title":"A deepconvlstm approach for continuous authentication using operating system performance counters","volume":"2025","author":"Andrade","year":"2025","journal-title":"IET Biometrics"},{"key":"B5","doi-asserted-by":"publisher","first-page":"39","DOI":"10.2478\/ausi-2020-0003","article-title":"Mouse dynamics based user recognition using deep learning","volume":"12","author":"Antal","year":"2020","journal-title":"Acta Univ. Sapientiae, Informatica"},{"key":"B6","doi-asserted-by":"publisher","first-page":"108039","DOI":"10.1016\/j.jfranklin.2025.108039","article-title":"Robust and imperceptible medical image watermarking for telemedicine applications based on transform-domain and neural clustering techniques","volume":"362","author":"Beggari","year":"2025","journal-title":"J. Franklin Institute"},{"key":"B7","doi-asserted-by":"publisher","first-page":"5","DOI":"10.7454\/mst.v27i1.1557","article-title":"User classification based on mouse dynamic authentication using k-nearest neighbor","volume":"27","author":"Chandranegara","year":"2023","journal-title":"Makara J. Technol"},{"key":"B8","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1613\/jair.953","article-title":"Smote: Synthetic minority over-sampling technique","volume":"16","author":"Chawla","year":"2002","journal-title":"J. Artif. Intellig. Res"},{"key":"B9","doi-asserted-by":"publisher","first-page":"1086","DOI":"10.1109\/TIFS.2019.2930429","article-title":"User authentication based on mouse dynamics using deep neural networks: A comprehensive study","volume":"15","author":"Chong","year":"2020","journal-title":"IEEE Trans. Inform. Forens. Security"},{"key":"B10","doi-asserted-by":"publisher","first-page":"12","DOI":"10.1109\/MITP.2013.50","article-title":"Continuous authentication using behavioral biometrics","volume":"15","author":"Deutschmann","year":"2013","journal-title":"IT Professional"},{"key":"B11","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1109\/IRI58017.2023.00010","article-title":"\u201cOn the improvements of mouse dynamics based continuous user authentication,\u201d","volume-title":"2023 IEEE 24th International Conference on Information Reuse and Integration for Data Science (IRI)","author":"Durmaz","year":"2023"},{"key":"B12","article-title":"Efficient and privacy-preserving federated learning based on full homomorphic encryption","author":"Guo","year":"2024","journal-title":"arXiv"},{"key":"B13","unstructured":"Cost of a Data Breach Report 2025\n          \n          2025"},{"key":"B14","unstructured":"\u201cMouse dynamics for user authentication,\u201d\n          \n          \n            \n              Jaafar\n              N.\n            \n          \n          Kaggle Dataset\n          \n          2021"},{"key":"B15","doi-asserted-by":"publisher","first-page":"323","DOI":"10.1007\/s00530-025-01897-0","article-title":"Captcha farm detection and user authentication via mouse-trajectory similarity measurement","volume":"31","author":"Jin","year":"2025","journal-title":"Multimedia Syst"},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2006.09365","article-title":"Byzantine-robust learning on heterogeneous datasets via bucketing","author":"Karimireddy","year":"2020","journal-title":"arXiv"},{"key":"B17","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3640311","article-title":"Mouse dynamics behavioral biometrics: a survey","volume":"56","author":"Khan","year":"2024","journal-title":"ACM Comp. Surv"},{"key":"B18","doi-asserted-by":"publisher","first-page":"50","DOI":"10.1109\/MSP.2020.2975749","article-title":"Federated learning: Challenges, methods, and future directions","volume":"37","author":"Li","year":"2020","journal-title":"IEEE Signal Proc. Magaz"},{"key":"B19","doi-asserted-by":"publisher","first-page":"9128","DOI":"10.1109\/JIOT.2020.3004077","article-title":"Behavioral biometrics for continuous authentication in the internet-of-things era: an artificial intelligence perspective","volume":"7","author":"Liang","year":"2020","journal-title":"IEEE Intern. Things J"},{"key":"B20","doi-asserted-by":"publisher","first-page":"6895","DOI":"10.1109\/TIFS.2025.3585435","article-title":"A mouse dynamics authentication system with a recurrence plot image representation and a vision transformer framework","volume":"20","author":"Mazumdar","year":"2025","journal-title":"IEEE Trans. Inform. Forens. Security"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.neucom.2016.11.031","article-title":"A study on continuous authentication using a combination of keystroke and mouse biometrics","volume":"230","author":"Mondal","year":"2017","journal-title":"Neurocomputing"},{"key":"B22","first-page":"15","article-title":"\u201cA review of evaluation metrics in machine learning algorithms,\u201d","volume-title":"Artificial Intelligence Application in Networks and Systems, volume 724 of Lecture Notes in Networks and Systems","author":"Naidu","year":"2023"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1002\/wics.1500","article-title":"Deep learning: computational aspects","author":"Polson","year":"2020","journal-title":"Wiley Interdiscipl. Rev.: Comput. Statist"},{"key":"B24","doi-asserted-by":"publisher","first-page":"6086","DOI":"10.1038\/s41598-024-56706-x","article-title":"Evaluation metrics and statistical tests for machine learning","volume":"14","author":"Rainio","year":"2024","journal-title":"Sci. Rep"},{"key":"B25","doi-asserted-by":"publisher","first-page":"3622","DOI":"10.1109\/TIFS.2023.3245418","article-title":"Attacking mouse dynamics authentication using novel wasserstein conditional dcgan","volume":"18","author":"Roy","year":"2023","journal-title":"IEEE Trans. Inform. Forens. Security"},{"key":"B26","doi-asserted-by":"publisher","first-page":"100280","DOI":"10.1016\/j.ibmed.2025.100280","article-title":"Fdct-based watermarking for robust and imperceptible medical image protection","volume":"2025","author":"Sa\u00efd","year":"","journal-title":"Intellig.-Based Med"},{"key":"B27","doi-asserted-by":"publisher","first-page":"200390","DOI":"10.1016\/j.sasc.2025.200390","article-title":"Robust medical image watermarking based on ridgelet transform and ant colony optimization for telemedicine security","volume":"2025","author":"Sa\u00efd","year":"","journal-title":"Syst. Soft Comp"},{"key":"B28","doi-asserted-by":"publisher","first-page":"963","DOI":"10.24996\/ijs.2018.59.2B.18","article-title":"User authentication via mouse dynamics","volume":"59","author":"Salman","year":"2018","journal-title":"Iraqi J. Sci"},{"key":"B29","doi-asserted-by":"publisher","first-page":"653","DOI":"10.1007\/s00530-022-00997-5","article-title":"User authentication method based on keystroke dynamics and mouse dynamics using HDA","volume":"29","author":"Shi","year":"2023","journal-title":"Multimedia Syst"},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.1109\/ICECET52533.2021.9698532","article-title":"\u201cContinuous user authentication using mouse dynamics, machine learning, and minecraft,\u201d","volume-title":"2021 International Conference on Electrical, Computer and Energy Technologies (ICECET)","author":"Siddiqui","year":"2021"},{"key":"B31","doi-asserted-by":"publisher","first-page":"502","DOI":"10.3390\/make4020023","article-title":"Machine and deep learning applications to mouse dynamics for continuous user authentication","volume":"4","author":"Siddiqui","year":"2022","journal-title":"Mach. Learn. Knowl. Extract"},{"key":"B32","doi-asserted-by":"publisher","first-page":"479","DOI":"10.1109\/JSAIT.2021.3054610","article-title":"Turbo-aggregate: breaking the quadratic aggregation barrier in secure federated learning","volume":"2","author":"So","year":"2021","journal-title":"IEEE J. Select. Areas Inform. Theory"},{"key":"B33","doi-asserted-by":"crossref","DOI":"10.1109\/LINDI63813.2024.10820440","article-title":"\u201cElevating security: Mouse dynamics in behavior biometrics for user identity authentication,\u201d","volume-title":"2024 IEEE 6th International Symposium on Logistics and Industrial Informatics (LINDI)","author":"Szil\u00e1gyi","year":"2024"},{"key":"B34","doi-asserted-by":"publisher","first-page":"227","DOI":"10.1504\/IJBM.2025.145901","article-title":"Feature ranking for effective continuous user authentication using keystroke and mouse dynamics with the cat recurrent neural model","volume":"17","author":"Thomas","year":"2025","journal-title":"Int. J. Biometrics"},{"key":"B35","doi-asserted-by":"publisher","first-page":"401","DOI":"10.1007\/s11277-021-09363-6","article-title":"An efficient optimized mouse and keystroke dynamics framework for continuous non-intrusive user authentication","volume":"124","author":"Thomas","year":"2022","journal-title":"Wireless Pers. Commun"},{"key":"B36","doi-asserted-by":"publisher","first-page":"6627","DOI":"10.3390\/s22176627","article-title":"User authentication method based on keystroke dynamics and mouse dynamics with scene-independent features in hybrid scenes","volume":"22","author":"Wang","year":"2022","journal-title":"Sensors"},{"key":"B37","doi-asserted-by":"publisher","first-page":"513","DOI":"10.1007\/s13042-022-01647-y","article-title":"A survey on federated learning: challenges and applications","volume":"14","author":"Wen","year":"2023","journal-title":"Int. J. Mach. Learn. Cybernet"},{"key":"B38","doi-asserted-by":"publisher","first-page":"24569","DOI":"10.1109\/JIOT.2024.3382875","article-title":"Efficiency optimization techniques in privacy-preserving federated learning with homomorphic encryption: a brief survey","volume":"11","author":"Xie","year":"2024","journal-title":"IEEE Intern. Things J"},{"key":"B39","doi-asserted-by":"crossref","first-page":"340","DOI":"10.1109\/AIIoT65859.2025.11105233","article-title":"\u201cMulti-modal adversarial activity detection using keyboard and mouse dynamics,\u201d","volume-title":"2025 IEEE World AI IoT Congress (AIIoT)","author":"Xiong","year":"2025"},{"key":"B40","doi-asserted-by":"publisher","first-page":"106775","DOI":"10.1016\/j.knosys.2021.106775","article-title":"A survey on federated learning","volume":"216","author":"Zhang","year":"2021","journal-title":"Knowl.-Based Syst"},{"key":"B41","doi-asserted-by":"publisher","first-page":"104555","DOI":"10.1016\/j.dsp.2024.104555","article-title":"An intelligent identity authentication method based on mouse trajectory and wireless signal","volume":"151","author":"Zhang","year":"2024","journal-title":"Digit. Signal Proc"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1914353\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T05:31:07Z","timestamp":1789623067000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1914353\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,17]]},"references-count":41,"alternative-id":["10.3389\/frai.2026.1914353"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1914353","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,17]]},"article-number":"1914353"}}