{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T10:40:05Z","timestamp":1773916805051,"version":"3.50.1"},"reference-count":33,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,3,17]],"date-time":"2026-03-17T00:00:00Z","timestamp":1773705600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Science Foundation","award":["21223749"],"award-info":[{"award-number":["21223749"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["BDCC"],"abstract":"<jats:p>Wearable fall detection systems face a fundamental challenge: while gyroscope data provide valuable orientation cues, naively combining raw gyroscope and accelerometer signals can degrade performance due to noise contamination. To overcome this challenge, we present a dual-stream transformer architecture that incorporates (i) Kalman-based sensor fusion to convert noisy gyroscope angular velocities into stable orientation estimates (roll, pitch, yaw), maintaining an internal state of body pose, and (ii) processing accelerometer and orientation streams in separate encoder pathways before fusion to prevent cross-modal interference. Our architecture further integrates Squeeze-and-Excitation channel attention and Temporal Attention Pooling to focus on fall-critical temporal patterns. Evaluated on the SmartFallMM dataset using 21-fold leave-one-subject-out cross-validation, the dual-stream Kalman transformer achieves 91.10% F1, outperforming single-stream Kalman transformers (89.80% F1) by 1.30% and single-stream baseline transformers (88.96% F1) by 2.14%. We further evaluate the model in real time using a watch-based SmartFall App on five participants, maintaining an average F1 score of 83% and an accuracy of 90%. These results indicate robust performance in both offline and real-world deployment settings, establishing a new state-of-the-art for inertial-measurement-unit-based fall detection on commodity smartwatch devices.<\/jats:p>","DOI":"10.3390\/bdcc10030090","type":"journal-article","created":{"date-parts":[[2026,3,17]],"date-time":"2026-03-17T12:42:16Z","timestamp":1773751336000},"page":"90","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Dual-Stream Transformer with Kalman-Based Sensor Fusion for Wearable Fall Detection"],"prefix":"10.3390","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-3820-6458","authenticated-orcid":false,"given":"Abheek","family":"Pradhan","sequence":"first","affiliation":[{"name":"Department of Computer Science, Texas State University, San Marcos, TX 78666, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6472-7570","authenticated-orcid":false,"given":"Sana","family":"Alamgeer","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Texas State University, San Marcos, TX 78666, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-7462-2286","authenticated-orcid":false,"given":"Rakesh","family":"Suvvari","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Texas State University, San Marcos, TX 78666, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-8024-9896","authenticated-orcid":false,"given":"Syed Tousiful","family":"Haque","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Texas State University, San Marcos, TX 78666, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5877-0230","authenticated-orcid":false,"given":"Anne H. H.","family":"Ngu","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Texas State University, San Marcos, TX 78666, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,3,17]]},"reference":[{"key":"ref_1","unstructured":"WHO (2024, December 01). Falls: Fact Sheet, Available online: https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/falls."},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Yasmin, A., Mahmud, T., Haque, S.T., Alamgeer, S., and Ngu, A.H.H. (2025). Enhancing Real-World Fall Detection Using Commodity Devices: A Systematic Study. Sensors, 25.","DOI":"10.3390\/s25175249"},{"key":"ref_3","unstructured":"SmartFall Group, Texas State University (2026, January 13). SmartFallMM: A Multimodal Dataset Collected with Commodity Devices. Available online: https:\/\/github.com\/txst-cs-smartfall\/SmartFallMM-Dataset."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Xuan, J., Zhu, T., Peng, G., Sun, F., and Dong, D. (2024). A Review on the Inertial Measurement Unit Array of Microelectromechanical Systems. Sensors, 24.","DOI":"10.3390\/s24227140"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"35","DOI":"10.1115\/1.3662552","article-title":"A New Approach to Linear Filtering and Prediction Problems","volume":"82","author":"Kalman","year":"1960","journal-title":"J. Basic Eng."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Hu, J., Shen, L., and Sun, G. (2018, January 18\u201323). Squeeze-and-Excitation Networks. Proceedings of the 2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA.","DOI":"10.1109\/CVPR.2018.00745"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Mauldin, T.R., Canby, M.E., Metsis, V., Ngu, A.H., and Rivera, C.C. (2018). SmartFall: A Smartwatch-Based Fall Detection System Using Deep Learning. Sensors, 18.","DOI":"10.3390\/s18103363"},{"key":"ref_8","unstructured":"SmartFall Group, Texas State University (2026, January 13). SmartFall Dataset, 2018. Available online: https:\/\/userweb.cs.txstate.edu\/~hn12\/data\/SmartFallDataSet\/."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1145\/3428666","article-title":"Ensemble Deep Learning on Wearables Using Small Datasets","volume":"2","author":"Mauldin","year":"2021","journal-title":"ACM Trans. Comput. Healthcare"},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Haque, S.T., Debnath, M., Yasmin, A., Mahmud, T., and Ngu, A.H.H. (2024). Experimental Study of Long Short-Term Memory and Transformer Models for Fall Detection on Smartwatches. Sensors, 24.","DOI":"10.3390\/s24196235"},{"key":"ref_11","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, \u0141., and Polosukhin, I. (2017, January 4\u20139). Attention is all you need. Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"e101439","DOI":"10.1136\/bmjhci-2025-101439","article-title":"Real-time activity and fall detection using transformer-based deep learning models for elderly care applications","volume":"32","author":"Zafar","year":"2025","journal-title":"BMJ Health Care Informatics"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Vavoulas, G., Chatzaki, C., Malliotakis, T., Pediaditis, M., and Tsiknakis, M. (2016, January 21\u201322). The MobiAct Dataset: Recognition of Activities of Daily Living using Smartphones. Proceedings of the 2nd International Conference on Information and Communication Technologies for Ageing Well and e-Health, Rome, Italy.","DOI":"10.5220\/0005792401430151"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Yhdego, H., Li, J., Paolini, C., and Audette, M. (2021, January 9\u201312). Wearable Sensor Gait Analysis of Fall Detection using Attention Network. Proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Virtual.","DOI":"10.1109\/BIBM52615.2021.9669795"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"10635","DOI":"10.1038\/s41598-025-95508-7","article-title":"Fall recognition using a three stream spatio temporal GCN model with adaptive feature aggregation","volume":"15","author":"Shin","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Rehouma, H., and Boukadoum, M. (2025). Fall Detection by Deep Learning-Based Bimodal Movement and Pose Sensing with Late Fusion. Sensors, 25.","DOI":"10.3390\/s25196035"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"109304","DOI":"10.1016\/j.bspc.2025.109304","article-title":"Efficient fall detection using Kalman filter-enhanced triaxial accelerometer signals and machine learning","volume":"114","author":"Liu","year":"2026","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Shi, Y., Zhang, Y., Li, Z., Yuan, S., and Zhu, S. (2023). IMU\/UWB Fusion Method Using a Complementary Filter and a Kalman Filter for Hybrid Upper Limb Motion Estimation. Sensors, 23.","DOI":"10.3390\/s23156700"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"An, G., Zhou, W., Wu, Y., Zheng, Z., and Liu, Y. (2018, January 12\u201316). Squeeze-and-Excitation on Spatial and Temporal Deep Feature Space for Action Recognition. Proceedings of the 2018 14th IEEE International Conference on Signal Processing (ICSP), Beijing, China.","DOI":"10.1109\/ICSP.2018.8652287"},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"110867","DOI":"10.1016\/j.knosys.2023.110867","article-title":"Temporal-channel convolution with self-attention network for human activity recognition using wearable sensors","volume":"278","author":"Essa","year":"2023","journal-title":"Knowl.-Based Syst."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"33389","DOI":"10.1038\/s41598-025-18444-6","article-title":"Adaptive temporal attention mechanism and hybrid deep CNN model for wearable sensor-based human activity recognition","volume":"15","author":"Wang","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Lin, T.Y., Goyal, P., Girshick, R., He, K., and Doll\u00e1r, P. (2017, January 22\u201329). Focal Loss for Dense Object Detection. Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy.","DOI":"10.1109\/ICCV.2017.324"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Student (1908). The Probable Error of a Mean. Biometrika, 6, 1\u201325.","DOI":"10.2307\/2331554"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"80","DOI":"10.2307\/3001968","article-title":"Individual Comparisons by Ranking Methods","volume":"1","author":"Wilcoxon","year":"1945","journal-title":"Biom. Bull."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"239","DOI":"10.1023\/A:1024068626366","article-title":"Inference for the Generalization Error","volume":"52","author":"Nadeau","year":"2003","journal-title":"Mach. Learn."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"e56750","DOI":"10.2196\/56750","article-title":"An Effective Deep Learning Framework for Fall Detection: Model Development and Study Design","volume":"26","author":"Zhang","year":"2024","journal-title":"J. Med. Internet Res."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Liu, C.P., Li, J.H., Chu, E.P., Hsieh, C.Y., Liu, K.C., Chan, C.T., and Tsao, Y. (2023, January 14\u201316). Deep Learning-based Fall Detection Algorithm Using Ensemble Model of Coarse-fine CNN and GRU Networks. Proceedings of the 2023 IEEE International Symposium on Medical Measurements and Applications (MeMeA), Jeju, Republic of Korea.","DOI":"10.1109\/MeMeA57477.2023.10171944"},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Wu, J., Wang, J., Zhan, A., and Wu, C. (2021). Fall Detection with CNN-Casual LSTM Network. Information, 12.","DOI":"10.3390\/info12100403"},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"Mart\u00ednez-Villase\u00f1or, L., Ponce, H., Brieva, J., Moya-Albor, E., N\u00fa\u00f1ez-Mart\u00ednez, J., and Pe\u00f1afort-Asturiano, C. (2019). UP-fall detection dataset: A multimodal approach. Sensors, 19.","DOI":"10.3390\/s19091988"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Fula, V., and Moreno, P. (2024). Wrist-based fall detection: Towards generalization across datasets. Sensors, 24.","DOI":"10.3390\/s24051679"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Yasmin, A., Mahmud, T., Debnath, M., and Ngu, A.H. (2024, January 2\u20134). An empirical study on ai-powered edge computing architectures for real-time iot applications. Proceedings of the 2024 IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC), Osaka, Japan.","DOI":"10.1109\/COMPSAC61105.2024.00187"},{"key":"ref_32","unstructured":"SmartFall Group, Texas State University (2026, January 29). Optimizing Real-Time Fall Detection: Integrating NATS.io for Low-Latency IoT Edge Applications. Available online: https:\/\/smartfall.github.io\/assets\/docs\/SayaliNATS.pdf."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"2250048","DOI":"10.1142\/S0129065722500484","article-title":"Personalized Watch-Based Fall Detection Using a Collaborative Edge-Cloud Framework","volume":"32","author":"Ngu","year":"2022","journal-title":"Int. J. Neural Syst."}],"container-title":["Big Data and Cognitive Computing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2504-2289\/10\/3\/90\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T05:23:33Z","timestamp":1773897813000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2504-2289\/10\/3\/90"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,17]]},"references-count":33,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3]]}},"alternative-id":["bdcc10030090"],"URL":"https:\/\/doi.org\/10.3390\/bdcc10030090","relation":{},"ISSN":["2504-2289"],"issn-type":[{"value":"2504-2289","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,17]]}}}