{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T13:00:48Z","timestamp":1785416448109,"version":"3.56.0"},"reference-count":25,"publisher":"MDPI AG","issue":"12","license":[{"start":{"date-parts":[[2026,6,9]],"date-time":"2026-06-09T00:00:00Z","timestamp":1780963200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"General Program of the National Natural Science Foundation of China","award":["52072293"],"award-info":[{"award-number":["52072293"]}]},{"award":["52072293"],"award-info":[{"award-number":["52072293"]}],"id":[{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>To address the issues of low perception accuracy and poor robustness in traditional motion recognition methods within complex walking environments for visually impaired individuals, this study utilizes multi-modal data, including ECG, PPG, and IMU, for classification. Regarding the low filtering efficiency of multi-modal data, an improved wavelet filtering algorithm based on LSTM is proposed. To further enhance classification accuracy, this paper introduces a motion recognition method for the blindfolded mobility simulation based on an Attention-based Two-Stream Deep Fusion Convolutional Neural Network (ATS-DFCNN). The proposed method constructs a two-stream heterogeneous feature extraction architecture by synchronously collecting tri-axial motion signals and physiological signals from subjects. A 1D-CNN is employed to capture the spatial geometric features of limb movements, while a hybrid CNN-GRU network is utilized to mine the temporal evolution patterns of physiological stress. Furthermore, an attention mechanism is introduced to achieve dynamic weighted fusion at the feature level, which strengthens critical motion features and suppresses environmental noise. Experiments were conducted with 10 subjects simulating the movements of visually impaired individuals, covering typical actions such as walking, standing, climbing stairs, descending stairs, and falling. The results demonstrate that the proposed adaptive filtering algorithm achieves an AUC of 0.942, significantly improving feature distinctiveness compared to traditional algorithms. The ATS-DFCNN model achieved an average recognition accuracy of 92.2% across five activity categories, representing a 4.8% performance increase over single IMU modal classification. Particularly in fall detection, the model effectively reduces false alarms through physiological feedback and accurately infers motion intentions, providing reliable technical support for the safety monitoring of intelligent walking-aid systems.<\/jats:p>","DOI":"10.3390\/s26123681","type":"journal-article","created":{"date-parts":[[2026,6,10]],"date-time":"2026-06-10T06:33:44Z","timestamp":1781073224000},"page":"3681","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Research on Two-Stream Networks Integrating Physiological Features and Attention Mechanisms for Motion Classification in Visually Impaired Individuals"],"prefix":"10.3390","volume":"26","author":[{"given":"Wentong","family":"Wang","sequence":"first","affiliation":[{"name":"School of Optoelectronic Engineering, Xi\u2019an Technological University, Xi\u2019an 710021, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Changyuan","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Optoelectronic Engineering, Xi\u2019an Technological University, Xi\u2019an 710021, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zehui","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Optoelectronic Engineering, Xi\u2019an Technological University, Xi\u2019an 710021, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenbo","family":"Huang","sequence":"additional","affiliation":[{"name":"Mechanical Manufacturing and Automation Program, Shaanxi Polytechnic University, Xianyang 712000, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,6,9]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"de Oliveira, S.T., Bozo, J.V., and Okimoto, M.L.L.R. (2016). Assistive Technology for People with Low Vision: Equipment for Accessibility of Visual Information. Advances in Ergonomics in Design, Springer.","DOI":"10.1007\/978-3-319-41983-1_63"},{"key":"ref_2","first-page":"62","article-title":"Omni-Directional Obstacle Detection System for the Visually Impaired Based on Deep Learning","volume":"36","author":"Liu","year":"2026","journal-title":"Comput. Technol. Dev."},{"key":"ref_3","first-page":"34","article-title":"Intelligent recognition of human motion posture based on Hu invariant moments of machine vision images","volume":"29","author":"Wang","year":"2026","journal-title":"J. Xi\u2019an Univ. (Nat. Sci. Ed.)"},{"key":"ref_4","unstructured":"World Health Organization (2019). World Report on Vision, WHO Press."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"74","DOI":"10.1145\/1964897.1964918","article-title":"Activity recognition using cell phone accelerometers","volume":"12","author":"Kwapisz","year":"2011","journal-title":"ACM SIGKDD Explor. Newsl."},{"key":"ref_6","first-page":"4564","article-title":"Current Trends in Complex Human Activity Recognition","volume":"96","author":"Sakr","year":"2018","journal-title":"J. Theor. Appl. Inf. Technol."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"10691","DOI":"10.3390\/s140610691","article-title":"Detecting falls with wearable sensors using machine learning techniques","volume":"14","author":"Ozdemir","year":"2014","journal-title":"Sensors"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Gaya-Morey, F.X., Manresa-Yee, C., and Buades-Rubio, J.M. (2024). Deep learning for computer vision based activity recognition and fall detection of the elderly: A systematic review. arXiv.","DOI":"10.1007\/s10489-024-05645-1"},{"key":"ref_9","unstructured":"Wu, H. (2024). Research on Action Recognition Methods Based on Human Skeleton Feature Extraction Models. [Master\u2019s Thesis, Nanjing University of Posts and Telecommunications]. (In Chinese)."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Reyes Leiva, K.M., Gato, M.\u00c1.C., and Olmedo, J.J.S. (2023). Estimation of Spatio-Temporal parameters of GAIT and posture of visually impaired people using wearable sensors. Sensors, 23.","DOI":"10.3390\/s23125564"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"1192","DOI":"10.1109\/SURV.2012.110112.00192","article-title":"A survey on human activity recognition using wearable sensors","volume":"15","author":"Lara","year":"2013","journal-title":"IEEE Commun. Surv. Tutor."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"13029","DOI":"10.1109\/JSEN.2021.3069927","article-title":"Human activity recognition with smartphone and wearable sensors using deep learning techniques: A review","volume":"21","author":"Ramanujam","year":"2021","journal-title":"IEEE Sens. J."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Lin, Y., Wang, K., Yi, W., and Lian, S. (2019). Deep learning based wearable assistive system for visually impaired people. Proceedings of the IEEE International Conference on Wearable and Implantable Body Sensor Networks, IEEE.","DOI":"10.1109\/ICCVW.2019.00312"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Ord\u00f3\u00f1ez, F.J., and Roggen, D. (2016). Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition. Sensors, 16.","DOI":"10.3390\/s16010115"},{"key":"ref_15","unstructured":"Ravi, N., Dandanayan, A., Mysore, P., and Littman, L.M. (2005). Activity recognition from accelerometer data. Proceedings of the AAAI Conference on Artificial Intelligence, AAAI Press."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"204","DOI":"10.1016\/j.ipm.2014.07.008","article-title":"Human activity data discovery from triaxial accelerometer sensor: Non-supervised learning sensitivity to feature extraction parametrization","volume":"51","author":"Machado","year":"2015","journal-title":"Inf. Process. Manag."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"241","DOI":"10.1016\/j.inffus.2021.11.006","article-title":"Multi-sensor information fusion based on machine learning for real applications in human activity recognition: State-of-the-art and research challenges","volume":"80","author":"Qiu","year":"2022","journal-title":"Inf. Fusion"},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"230","DOI":"10.1109\/TBME.1985.325532","article-title":"A real-time QRS detection algorithm","volume":"32","author":"Pan","year":"1985","journal-title":"IEEE Trans. Biomed. Eng."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Reiss, A., and Stricker, D. (2012). Introducing a new benchmarked dataset for activity monitoring. Proceedings of the IEEE International Symposium on Wearable Computers, IEEE.","DOI":"10.1109\/ISWC.2012.13"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Cacciola, M., Angiulli, G., Burrascano, P., Lagan\u00e0, F., and Versaci, M. (2026). A Prototypical Fuzzy Similarity-Based Classification Framework for Ultrasonic Defect Detection in Concrete. Eng, 7.","DOI":"10.3390\/eng7020088"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"210816","DOI":"10.1109\/ACCESS.2020.3037715","article-title":"Human activity recognition using inertial, physiological and environmental sensors: A comprehensive survey","volume":"8","author":"Demrozi","year":"2020","journal-title":"IEEE Access"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Afzali Arani, M.S., Costa, D.E., and Shihab, E. (2021). Human activity recognition: A comparative study of accelerometer, ECG and PPG signals. Sensors, 21.","DOI":"10.3390\/s21216997"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"1018","DOI":"10.1016\/j.asoc.2015.01.025","article-title":"On the use of ensemble of classifiers for accelerometer-based activity recognition","volume":"37","author":"Catal","year":"2015","journal-title":"Appl. Soft Comput."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Lagan\u00e0, F., Prattic\u00f2, D., Quattrone, M.F., Pullano, S.A., and Calcagno, S. (2026). Hybrid AI\u2013Taguchi\u2013ANOVA Approach for Thermographic Monitoring of Electronic Devices. Eng, 7.","DOI":"10.3390\/eng7010028"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Lagan\u00e0, F. (2026). Design and Simulation-Based Validation of an Embedded Acquisition Architecture for In Situ PCB Integrity Monitoring in Biomedical Devices. Electronics, 15.","DOI":"10.3390\/electronics15040833"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/12\/3681\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,10]],"date-time":"2026-06-10T06:44:08Z","timestamp":1781073848000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/12\/3681"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,9]]},"references-count":25,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2026,6]]}},"alternative-id":["s26123681"],"URL":"https:\/\/doi.org\/10.3390\/s26123681","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,9]]}}}