{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,21]],"date-time":"2026-11-21T10:54:21Z","timestamp":1795258461140,"version":"4.10.3"},"reference-count":21,"publisher":"BMJ","issue":"1","license":[{"start":{"date-parts":[[2025,9,17]],"date-time":"2025-09-17T00:00:00Z","timestamp":1758067200000},"content-version":"unspecified","delay-in-days":16,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["BMJ Health Care Inform"],"accepted":{"date-parts":[[2025,9,5]]},"published-print":{"date-parts":[[2025,9]]},"abstract":"<jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>This study aims to develop a transformer-based deep learning model for real-time activity recognition and fall detection, addressing the limitations of existing methods in terms of accuracy and real-time applicability.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>The proposed system uses sliding window segmentation technique to process wearable sensor data, including accelerometer, gyroscope and orientation signals. The transformer encoder models temporal dependencies through a self-attention mechanism, enabling the extraction of global and local temporal patterns. The performance of the model is evaluated on an updated version of the MobiAct data set, which includes over 14\u2009million sensor records collected from 66 participants and 16 activities, including four types of falls and multiple scenario-based activities of daily living.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Result<\/jats:title>\n                  <jats:p>The transformer model achieved an accuracy of over 98% and demonstrated excellent precision and recall for difficult fall categories such as forward-lying and sideward-lying. Comparative analysis shows that transformers outperform convolutional neural networks long short-term memory (CNN-LSTM) and temporal convolutional networks in terms of classification metrics, confusion matrix results and training stability.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>The results highlight the effectiveness of the transformer model in capturing complex temporal dependencies, addressing key challenges such as misclassification and false positives. Compared with traditional models, its parallel processing capabilities improve real-time deployment efficiency.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>This research establishes transformer-based models as powerful solutions for activity recognition and fall detection, providing reliable applications for elderly care and fall prevention. Future work will focus on optimising edge devices and validating on real-world data sets.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1136\/bmjhci-2025-101439","type":"journal-article","created":{"date-parts":[[2025,9,18]],"date-time":"2025-09-18T05:05:21Z","timestamp":1758171921000},"page":"e101439","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":22,"title":["Real-time activity and fall detection using transformer-based deep learning models for elderly care applications"],"prefix":"10.1136","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7223-7977","authenticated-orcid":false,"given":"Raja Omman","family":"Zafar","sequence":"first","affiliation":[{"name":"Dalarna University\u2014Campus Borlange, Borl\u00e4nge, Sweden"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Farhan","family":"Zafar","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Capital University of Science and Technology, Islamabad, Islamabad Capital Territory, Pakistan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"239","published-online":{"date-parts":[[2025,9,17]]},"reference":[{"key":"2025091722051589000_32.1.e101439.1","unstructured":"World Health Organization . Falls: key facts. WHO; 2021. Available: https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/falls [Accessed 4 Jan 2025]."},{"key":"2025091722051589000_32.1.e101439.2","doi-asserted-by":"crossref","first-page":"257","DOI":"10.1007\/s12062-020-09267-6","article-title":"Developing a Smart Home Solution Based on Personalized Intelligent Interior Units to Promote Activity and Customized Healthcare for Aging Society","volume":"13","author":"Hu","year":"2020","journal-title":"Population Ageing"},{"key":"2025091722051589000_32.1.e101439.3","doi-asserted-by":"crossref","first-page":"498","DOI":"10.1016\/j.procs.2019.12.212","article-title":"Fall Prevention as Personal Learning and Changing Behaviors: Systems and Technologies","volume":"164","author":"Aidemark","year":"2019","journal-title":"Procedia Comput Sci"},{"key":"2025091722051589000_32.1.e101439.4","doi-asserted-by":"crossref","first-page":"31314","DOI":"10.3390\/s151229858","article-title":"Physical Human Activity Recognition Using Wearable Sensors","volume":"15","author":"Attal","year":"2015","journal-title":"Sensors (Basel)"},{"key":"2025091722051589000_32.1.e101439.5","doi-asserted-by":"crossref","first-page":"149","DOI":"10.14283\/jfa.2024.19","article-title":"Wearable Technologies for Healthy Ageing: Prospects, Challenges, and Ethical Considerations","volume":"13","author":"Canali","year":"2024","journal-title":"J Frailty Aging"},{"key":"2025091722051589000_32.1.e101439.6","doi-asserted-by":"crossref","unstructured":"Daghistani T , Alshammari R . Improving Accelerometer-Based Activity Recognition by Using Ensemble of Classifiers. Ijacsa 2016;7. doi:10.14569\/IJACSA.2016.070520","DOI":"10.14569\/IJACSA.2016.070520"},{"key":"2025091722051589000_32.1.e101439.7","doi-asserted-by":"crossref","first-page":"132","DOI":"10.1016\/j.ins.2020.05.070","article-title":"Fall detection in older adults with mobile IoT devices and machine learning in the cloud and on the edge","volume":"537","author":"Mrozek","year":"2020","journal-title":"Inf Sci (Ny)"},{"key":"2025091722051589000_32.1.e101439.8","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/j.patrec.2018.02.010","article-title":"Deep learning for sensor-based activity recognition: A survey","volume":"119","author":"Wang","year":"2019","journal-title":"Pattern Recognit Lett"},{"key":"2025091722051589000_32.1.e101439.9","doi-asserted-by":"crossref","unstructured":"Zafar RO , Latif I . Exploring advanced deep learning architectures for older adults activity recognition. In: Miesenberger K , Pe\u0148\u00e1z P , Kobayashi M , eds. Computers Helping People with Special Needs: 19th International Conference, ICCHP 2024, Linz, Austria, July 8\u201312, 2024, Proceedings, Part II; Cham, 2024:320\u20137.","DOI":"10.1007\/978-3-031-62849-8_39"},{"key":"2025091722051589000_32.1.e101439.10","doi-asserted-by":"crossref","unstructured":"Zhang X , Abbasi QH , Fioranelli F , et al . Elderly care - human activity recognition using radar with an open dataset and hybrid maps. In: Rehman M UR , Zoha A , eds. Body Area Networks. Smart IoT and Big Data for Intelligent Health Management: 16th EAI International Conference, BODYNETS 2021, Virtual Event, October 25-26, 2021, Proceedings; Cham, 2022:39\u201351.","DOI":"10.1007\/978-3-030-95593-9_4"},{"key":"2025091722051589000_32.1.e101439.11","doi-asserted-by":"crossref","DOI":"10.3390\/s22051911","article-title":"Wearable Sensor-Based Human Activity Recognition with Transformer Model","volume":"22","author":"Dirgov\u00e1 Lupt\u00e1kov\u00e1","year":"2022","journal-title":"Sensors (Basel)"},{"key":"2025091722051589000_32.1.e101439.12","doi-asserted-by":"crossref","first-page":"238","DOI":"10.1016\/j.procs.2018.10.324","article-title":"Real-time Detection of Human Falls in Progress: Machine Learning Approach","volume":"140","author":"Serpen","year":"2018","journal-title":"Procedia Comput Sci"},{"key":"2025091722051589000_32.1.e101439.13","doi-asserted-by":"crossref","DOI":"10.3390\/s18041101","article-title":"Real-Life\/Real-Time Elderly Fall Detection with a Triaxial Accelerometer","volume":"18","author":"Sucerquia","year":"2018","journal-title":"Sensors (Basel)"},{"key":"2025091722051589000_32.1.e101439.14","doi-asserted-by":"crossref","unstructured":"Muheidat F , Tawalbeh L , Tyrer H . Context-aware, accurate, and real time fall detection system for elderly people. 2018 IEEE 12th International Conference on Semantic Computing (ICSC); Laguna Hills, CA, 2018:329\u201333. doi:10.1109\/ICSC.2018.00068","DOI":"10.1109\/ICSC.2018.00068"},{"key":"2025091722051589000_32.1.e101439.15","doi-asserted-by":"crossref","first-page":"106159","DOI":"10.1016\/j.neunet.2024.106159","article-title":"A novel physical activity recognition approach using deep ensemble optimized transformers and reinforcement learning","volume":"173","author":"Ahmadian","year":"2024","journal-title":"Neural Netw"},{"key":"2025091722051589000_32.1.e101439.16","first-page":"44","article-title":"The MobiFall Dataset","volume":"2","author":"Vavoulas","year":"2014","journal-title":"Int J Monit Surveill Technol Res"},{"key":"2025091722051589000_32.1.e101439.17","doi-asserted-by":"crossref","first-page":"85231","DOI":"10.1109\/ACCESS.2024.3415359","article-title":"Human Activity Recognition Based on Self-Attention Mechanism in WiFi Environment","volume":"12","author":"Ge","year":"2024","journal-title":"IEEE Access"},{"key":"2025091722051589000_32.1.e101439.18","doi-asserted-by":"crossref","first-page":"6390","DOI":"10.1109\/TNNLS.2021.3136503","article-title":"DeepSMOTE: Fusing Deep Learning and SMOTE for Imbalanced Data","volume":"34","author":"Dablain","year":"2023","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"2025091722051589000_32.1.e101439.19","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3681797","article-title":"A Systematic Review of Digital Twin Technology for Home Care","volume":"5","author":"Zafar","year":"2024","journal-title":"ACM Trans Comput Healthcare"},{"key":"2025091722051589000_32.1.e101439.20","doi-asserted-by":"crossref","DOI":"10.3390\/s17010198","article-title":"SisFall: A Fall and Movement Dataset","volume":"17","author":"Sucerquia","year":"2017","journal-title":"Sensors (Basel)"},{"key":"2025091722051589000_32.1.e101439.21","doi-asserted-by":"crossref","first-page":"4267","DOI":"10.1109\/TNNLS.2024.3360990","article-title":"Deep Reinforcement Learning in Human Activity Recognition: A Survey and Outlook","volume":"36","author":"Nikpour","year":"2024","journal-title":"IEEE Trans Neural Netw Learning Syst"}],"container-title":["BMJ Health &amp; Care Informatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bmjhci-2025-101439","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,18]],"date-time":"2025-09-18T05:05:25Z","timestamp":1758171925000},"score":1,"resource":{"primary":{"URL":"https:\/\/informatics.bmj.com\/lookup\/doi\/10.1136\/bmjhci-2025-101439"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9]]},"references-count":21,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,9,17]]},"published-print":{"date-parts":[[2025,9]]}},"alternative-id":["10.1136\/bmjhci-2025-101439"],"URL":"https:\/\/doi.org\/10.1136\/bmjhci-2025-101439","relation":{},"ISSN":["2632-1009"],"issn-type":[{"value":"2632-1009","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9]]}}}