{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T05:48:05Z","timestamp":1782193685623,"version":"3.54.5"},"reference-count":37,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T00:00:00Z","timestamp":1782172800000},"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. Neurorobot."],"abstract":"<jats:p>Driver fatigue constitutes a critical safety hazard in Level 2-3 (L2-3) conditionally automated vehicles, where the paradoxical demand for sustained supervisory vigilance despite minimal active engagement accelerates cognitive underload and impairs timely takeover readiness. Existing vision-based driver monitoring systems are constrained by fixed temporal analysis windows and binary classifiers that neither quantify prediction uncertainty nor adapt to the heterogeneity of real-world fatigue dynamics, resulting in elevated false alarm rates and poor cross-domain generalization. This study introduces a confidence-driven adaptive time window (CDATW) framework: a closed-loop neuro-computational pipeline that couples a lightweight MobileNetV3-CBAM-BiLSTM spatial\u2013temporal encoder with a Monte Carlo Dropout uncertainty estimator to produce simultaneous fatigue probability and epistemic confidence outputs at each inference step. The confidence signal governs a window controller that contracts observation periods to 5\u201310\u202fs under high certainty (confidence &amp;gt;0.85) for rapid warning, and extends them to 20\u201330\u202fs under low certainty (confidence &amp;lt;0.60) to suppress spurious alarms\u2014instantiating the feedback-driven adaptive sensing principle central to neurorobotic perception. The framework was validated on four heterogeneous public datasets (NTHU-DDD, YawDD, UTA-RLDD, and DROZY) under single-dataset, cross-dataset transfer, and mixed-dataset training protocols. Single-dataset accuracy ranged from 88.6 to 91.8% with AUC of 0.92\u20130.95, while the adaptive mechanism reduced false alarm rates by 35.2% relative to fixed 15-s baselines. The architecture sustains 38\u201345 FPS on an NVIDIA Jetson Xavier NX automotive embedded platform, and confidence calibration achieves an Expected Calibration Error of 0.078, with high-confidence predictions (&amp;gt;0.9) attaining 95.6% accuracy. These results demonstrate that uncertainty-aware adaptive temporal reasoning embedded in a deployable neurorobotic architecture constitutes a computationally efficient and practically viable strategy for driver state monitoring in L2-3 autonomous vehicles, with broader implications for closed-loop perception in safety-critical human-machine systems.<\/jats:p>","DOI":"10.3389\/fnbot.2026.1857548","type":"journal-article","created":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T05:26:14Z","timestamp":1782192374000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Confidence-driven adaptive time window for real-time driver fatigue detection in Level 2-3 autonomous vehicles: a multi-dataset validation study"],"prefix":"10.3389","volume":"20","author":[{"given":"Wantong","family":"Xie","sequence":"first","affiliation":[{"name":"School of Computer and Network Security, Chengdu University of Technology","place":["Chengdu, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peng","family":"Xiao","sequence":"additional","affiliation":[{"name":"College of Computer Science and Cyber Security (Pilot Software College), Chengdu University of Technology","place":["Chengdu, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,6,23]]},"reference":[{"key":"ref1","first-page":"24","author":"Abtahi","year":"2014"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"19743","DOI":"10.1109\/TITS.2021.3134222","article-title":"Intelligent driver drowsiness detection for traffic safety based on multi CNN deep model and facial subsampling","volume":"23","author":"Ahmed","year":"2021","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref3","doi-asserted-by":"publisher","first-page":"9","DOI":"10.3390\/safety10010009","article-title":"A semantic hybrid temporal approach for detecting driver mental fatigue","volume":"10","author":"Ansari","year":"2024","journal-title":"Safety"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"15840","DOI":"10.1038\/s41598-025-92701-6","article-title":"Dynamic cross-domain transfer learning for driver fatigue monitoring: multi-modal sensor fusion with adaptive real-time personalizations","volume":"15","author":"Aravinth","year":"2025","journal-title":"Sci. Rep."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"12240","DOI":"10.1038\/s41598-025-86709-1","article-title":"Optimized driver fatigue detection method using multimodal neural networks","volume":"15","author":"Cao","year":"2025","journal-title":"Sci. Rep."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"692","DOI":"10.1080\/10447318.2015.1070549","article-title":"Investigating the importance of trust on adopting an autonomous vehicle","volume":"31","author":"Choi","year":"2015","journal-title":"Int. J. Hum. Comput. Interact."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"104851","DOI":"10.1016\/j.micpro.2023.104851","article-title":"Real-time driver fatigue detection system with deep learning on a low-cost embedded system","volume":"99","author":"Civik","year":"2023","journal-title":"Microprocess. Microsyst."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"7921","DOI":"10.1109\/TNNLS.2022.3147208","article-title":"EEG-based cross-subject driver drowsiness recognition with an interpretable convolutional neural network","volume":"34","author":"Cui","year":"2022","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref9","doi-asserted-by":"publisher","first-page":"3089","DOI":"10.1109\/TITS.2020.2979527","article-title":"Vision-based fatigue driving recognition method integrating heart rate and facial features","volume":"22","author":"Du","year":"2020","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref10","doi-asserted-by":"publisher","first-page":"13559","DOI":"10.1109\/TITS.2021.3125737","article-title":"Detection of train driver fatigue and distraction based on forehead EEG: a time-series ensemble learning method","volume":"23","author":"Fan","year":"2022","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"397","DOI":"10.1016\/j.eswa.2016.06.042","article-title":"Dynamic driver fatigue detection using hidden Markov model in real driving condition","volume":"63","author":"Fu","year":"2016","journal-title":"Expert Syst. Appl."},{"key":"ref12","first-page":"1050","author":"Gal","year":"2016"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"1513","DOI":"10.1007\/s10462-023-10562-9","article-title":"A survey of uncertainty in deep neural networks","volume":"56","author":"Gawlikowski","year":"2023","journal-title":"Artif. Intell. Rev."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"15518","DOI":"10.1038\/s41598-025-99441-7","article-title":"A dense multi-pooling convolutional network for driving fatigue detection","volume":"15","author":"Han","year":"2025","journal-title":"Sci. Rep."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"17493","DOI":"10.1038\/s41598-025-02111-x","article-title":"Real-time driver drowsiness detection using transformer architectures: a novel deep learning approach","volume":"15","author":"Hassan","year":"2025","journal-title":"Sci. Rep."},{"key":"ref16","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3786319","article-title":"A survey on uncertainty quantification methods for deep learning","volume":"57","author":"He","year":"2025","journal-title":"ACM Comput. Surv."},{"key":"ref17","first-page":"1314","author":"Howard","year":"2019"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"1234","DOI":"10.1109\/TITS.2023.3348517","article-title":"EEG-based driver fatigue detection using spatio-temporal fusion network with brain region partitioning strategy","volume":"26","author":"Hu","year":"2024","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref19","doi-asserted-by":"publisher","first-page":"17075","DOI":"10.1038\/s41598-024-67131-5","article-title":"Efficient detection of driver fatigue state based on all-weather illumination scenarios","volume":"14","author":"Hu","year":"2024","journal-title":"Sci. Rep."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"505","DOI":"10.1007\/s11760-023-02733-6","article-title":"Computer vision-based driver fatigue detection framework with personalization threshold and multi-feature fusion","volume":"18","author":"Li","year":"2024","journal-title":"Signal Image Video Process."},{"key":"ref21","doi-asserted-by":"publisher","first-page":"e0304669","DOI":"10.1371\/journal.pone.0304669","article-title":"Advancing driver fatigue detection in diverse lighting conditions for assisted driving vehicles with enhanced facial recognition technologies","volume":"19","author":"Lin","year":"2024","journal-title":"PLoS One"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"576","DOI":"10.1049\/ipr2.12373","article-title":"Real time detection of driver fatigue based on CNN-LSTM","volume":"16","author":"Liu","year":"2022","journal-title":"IET Image Process."},{"key":"ref23","doi-asserted-by":"publisher","first-page":"545","DOI":"10.1007\/s44230-025-00113-6","article-title":"Driver fatigue prediction method based on BiLSTM and novel comprehensive fatigue index","volume":"5","author":"Lv","year":"2025","journal-title":"Hum. Centric Intell. Syst."},{"key":"ref24","first-page":"1","author":"Massoz","year":"2016"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"682","DOI":"10.1080\/1463922X.2019.1621406","article-title":"A multi-level model on automated vehicle acceptance (MAVA): a review-based study","volume":"20","author":"Nordhoff","year":"2019","journal-title":"Theor. Issues Ergon. Sci."},{"key":"ref26","doi-asserted-by":"publisher","first-page":"1319","DOI":"10.1109\/TIV.2022.3224690","article-title":"Challenges of driver drowsiness prediction: the remaining steps to implementation","volume":"8","author":"Perkins","year":"2023","journal-title":"IEEE Trans. Intell. Veh."},{"key":"ref27","doi-asserted-by":"publisher","first-page":"e24112","DOI":"10.1016\/j.heliyon.2024.e24112","article-title":"Getting back in the loop: does autonomous driving duration affect driver's takeover performance?","volume":"10","author":"Portron","year":"2024","journal-title":"Heliyon"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"3711","DOI":"10.1007\/s00521-023-09255-9","article-title":"An embedded device-oriented fatigue driving detection method based on a YOLOv5s","volume":"36","author":"Qu","year":"2024","journal-title":"Neural Comput. & Applic."},{"key":"ref29","doi-asserted-by":"publisher","first-page":"1317","DOI":"10.1109\/JAS.2024.124290","article-title":"Uncertainty-aware deep learning: a promising tool for trustworthy fault diagnosis","volume":"11","author":"Ren","year":"2024","journal-title":"IEEE CAA J. Autom. Sin."},{"key":"ref30","first-page":"4510","author":"Sandler","year":"2018"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"913","DOI":"10.3390\/smartcities7020038","article-title":"Gauging public acceptance of conditionally automated vehicles in the United States","volume":"7","author":"Saravanos","year":"2024","journal-title":"Smart Cities"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"2339","DOI":"10.1109\/TITS.2018.2868499","article-title":"Driver fatigue detection systems: a review","volume":"20","author":"Sikander","year":"2019","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"9811","DOI":"10.1007\/s10489-024-05696-4","article-title":"A real-time and lightweight driver fatigue detection model using anchor-free and visual-attention mechanisms","volume":"54","author":"Wang","year":"2024","journal-title":"Appl. Intell."},{"key":"ref34","doi-asserted-by":"publisher","first-page":"2884","DOI":"10.3390\/electronics12132884","article-title":"A multi-feature fusion and situation awareness-based method for fatigue driving level determination","volume":"12","author":"Wei","year":"2023","journal-title":"Electronics"},{"key":"ref35","first-page":"117","author":"Weng","year":"2016"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"4999","DOI":"10.1109\/TIE.2023.3288182","article-title":"Real-time driver cognitive workload recognition: attention-enabled learning with multimodal information fusion","volume":"71","author":"Yang","year":"2024","journal-title":"IEEE Trans. Ind. Electron."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"247","DOI":"10.3390\/a18050247","article-title":"A multi-feature fusion algorithm for fatigue driving detection considering individual driver differences","volume":"18","author":"Zhou","year":"2025","journal-title":"Algorithms"}],"container-title":["Frontiers in Neurorobotics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2026.1857548\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,23]],"date-time":"2026-06-23T05:26:14Z","timestamp":1782192374000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2026.1857548\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,23]]},"references-count":37,"alternative-id":["10.3389\/fnbot.2026.1857548"],"URL":"https:\/\/doi.org\/10.3389\/fnbot.2026.1857548","relation":{},"ISSN":["1662-5218"],"issn-type":[{"value":"1662-5218","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,23]]},"article-number":"1857548"}}