人體活動辨識 (Human Activity Recognition, HAR) 在健康照護、運動分析和輔助生活等各個領域中扮演著重要的角色。隨著Wi-Fi技術的普及，Wi-Fi通道狀態資訊 (Wi-Fi Channel State Information, CSI) 因其非侵入性的特性和廣泛可用性而成為HAR的寶貴資源。本研究探討了在HAR中應用深度學習技術，利用Wi-Fi CSI進行活動辨識。系統架構包含對CSI資訊進行預處理、特徵提取模組從CSI數據中提取相關特徵，以及使用深度學習模型 (如LSTM) 進行活動辨識的分類模組。本研究的發現有助於推進使用Wi-Fi CSI的HAR技術，並為發展堅固且即時的活動辨識系統提供了深入洞察。

Human Activity Recognition (HAR) plays a vital role in various domains such as healthcare, sports analysis, and assisted living. With the proliferation of Wi-Fi technology, Wi-Fi Channel State Information (CSI) has emerged as a valuable resource for HAR due to its non-intrusive nature and widespread availability. In this study, we investigate the application of deep learning techniques in HAR using Wi-Fi CSI. The system's architecture consists of preprocesses CSI information, a feature extraction module that extracts relevant features from the CSI data, and a classification module that utilizes a deep learning model, such as LSTM, to perform activity recognition. The findings of this study contribute to the advancement of HAR techniques using Wi-Fi CSI and provide insights into the development of robust and real-time activity recognition systems.

[1] Daniel Halperin, Wenjun Hu, Anmol Shethy, and David Wetherall, “Two Antennas are Better than One:A Measurement Study of 802.11n”.
[2] Andrii Zhuravchaka, Oleg Kapshiib, Evangelos Pournarasc, “ Human Activity Recognition based on Wi-Fi CSI Data -A Deep Neural Network Approach,” 於 The 11th International Conference on Emerging Ubiquitous Systems and Pervasive Networks, 2021.
[3] M. H. R. ,. W. S. ,. M. HUMAYUN KABIR, “CSI-IANet: An Inception Attention Network for Human-Human Interaction Recognition Based on CSI Signal,” 於 IEEE Acess, 2021.
[4] J. D. C. W. Nan Bao 且 D. H. J. C. R. N. Z. L. , “Wi-Breath: A WiFi-Based Contactless and Real-Time Respiration Monitoring Scheme for Remote Healthcare,” 於 IEEE Journal of Biomedical and Health Informatics, 2022.
[5] X. Z. Y. L. S. Z. J. W. Haihan Li, “Convolutional neural networks based indoor Wi-Fi localization with a novel kind of CSI images,” 於 China Communications, 2019.
[6] K. T. K. N. T. M. S. O. Osamu Muta, “Device-Free WLAN Based Indoor Localization Scheme With Spatially Concatenated CSI and Distributed Antennas,” 於 IEEE Transactions on Vehicular Technology, 2022.
[7] PENGPENG CHEN , FEN LIU , SHOUWAN GAO , PEIHAO LI ,XU YANG AND QIANG NIU, “Smartphone-Based Indoor Fingerprinting Localization Using Channel State Information,” 於 IEEE Access, 2019.
[8] M. B. K. Y. a. T. O. Takashi Nakamura, “Wi-Fi-CSI-based Fall Detection by Spectrogram Analysis with CNN,” 於 IEEE Global Communications Conference, 2021.
[9] M. B. K. Y. T. O. Takashi Nakamura, “Wi-Fi-Based Fall Detection Using Spectrogram Image of Channel State Information,” 於 IEEE IoT, 2022.
[10] Siamak Yousefi, Hirokazu Narui, Sankalp Dayal, Stefano Ermon, and Shahrokh Valaee, “A Survey on Behavior Recognition Using WiFi Channel State Information,” 於 IEEE, 2017.
[11] SAMEERA PALIPANA, DAVID ROJAS, PIYUSH AGRAWAL, DIRK PESCH, “FallDeFi: Ubiquitous Fall Detection using Commodity Wi-Fi Devices,” 於 ACM, 2018.
[12] Jing Bi, Xiang Zhang, Haitao Yuan , Jia Zhang, and MengChu Zhou, “A Hybrid Prediction Method for Realistic Network Traffic With Temporal Convolutional Network and LSTM,” 於 IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING, VOL. 19, NO. 3, 2022.
[13] PeterHillyard, AnhLuong, AlemayehuSolomon Abrar, NealPatwari, NealPatwari, RobertFarney, JasonBurch, ChristinaA.Porucznik, SarahHatchPollard, “Experience:Cross-TechnologyRadioRespiratory MonitoringPerformanceStudy,” 於 MobiCom’18, 2018.
[14] BeiMing Yan, Wei Cheng, GeTong Huang, Zhong Shang Zhu, Xiang Gao, “Activity Recognition Using the Joint of Wi-Fi 2.4G and 5G Frequency Bands,” 於 IEEE 21st ICCT, 2021.