DiffCSI:Enhance Channel State Infomation-based Activities Recognition with Diffusion

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：78

、訪客IP：3.145.7.187

姓名

劉威廷(Wei-Ting Liu) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

(DiffCSI:Enhance Channel State Infomation-based Activities Recognition with Diffusion)

相關論文

★ 基於edX線上討論板社交關係之分組機制	★ 利用Kinect建置3D視覺化之Facebook互動系統
★ 利用 Kinect建置智慧型教室之評量系統	★ 基於行動裝置應用之智慧型都會區路徑規劃機制
★ 基於分析關鍵動量相關性之動態紋理轉換	★ 基於保護影像中直線結構的細縫裁減系統
★ 建基於開放式網路社群學習環境之社群推薦機制	★ 英語作為外語的互動式情境學習環境之系統設計
★ 基於膚色保存之情感色彩轉換機制	★ 一個用於虛擬鍵盤之手勢識別框架
★ 分數冪次型灰色生成預測模型誤差分析暨電腦工具箱之研發	★ 使用慣性傳感器構建即時人體骨架動作
★ 基於多台攝影機即時三維建模	★ 基於互補度與社群網路分析於基因演算法之分組機制
★ 即時手部追蹤之虛擬樂器演奏系統	★ 基於類神經網路之即時虛擬樂器演奏系統

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2026-7-5以後開放)

摘要(中)

在物聯網 (IoT) 應用領域中，基於通道狀態信息 (CSI) 的活動識別因其在實現精確的無線定位、行為識別和健康監測系統方面的潛力而引起了極大關注。 CSI 利用無線信號在物理環境中的傳播細節，為開發先進的識別系統提供了基礎。然而，收集涵蓋各種環境條件和特定行動（如跌倒）的全面數據集具有相當大的挑戰。這一難度在嘗試捕捉罕見或複雜事件時更為顯著，此時需要多樣且廣泛的數據集變得尤為重要。為應對這些挑戰，本研究倡導採用擴散生成模型作為合成原始 CSI 數據的新方法。我們的實驗結果表明，這種方法顯著減少了傳統數據收集所需的時間和資源，同時提升了活動識別系統的性能。本研究的主要貢獻包括： (1) 引入一種創新技術來生成原始 CSI 數據，該技術能夠產生多樣化的合成樣本，促進訓練穩健的分類器；以及 (2) 開發了一種獨特的分類器架構，通過整合來自不同時間段的數據並使用 Transformer 編碼器，克服了傳統時間卷積網絡 (TCN) 的限制。這種整合，結合優化的 TCN 核大小，確保了在活動識別任務中的卓越準確性。

摘要(英)

In the realm of Internet of Things (IoT) applications, activity recognition based on Channel State Information (CSI) has garnered significant interest due to its potential for enabling precise wireless localization, behavior recognition, and health monitoring systems. CSI leverages the intricacies of wireless signal propagation in physical environments, providing a foundation for developing advanced recognition systems. Nonetheless, amassing a comprehensive dataset covering a wide array of environmental conditions and specific actions, such as falls, poses substantial challenges. This difficulty is exacerbated when attempting to capture rare or complex events, where the need for a diverse and extensive dataset becomes paramount.
In response to these challenges, this study advocates for the adoption of diffusion generative models as a novel methodology for synthesizing raw CSI data. Our experimental findings reveal that this approach markedly mitigates the time and resources traditionally required for data collection, while simultaneously enhancing the performance of activity recognition systems. Main contributions of this research include:(1) The introduction of an innovative technique for generating raw CSI data, capable of producing a varied set of synthetic samples to facilitate the training of robust classifiers; and (2) The development of a unique classifier architecture that surpasses the limitations of conventional Temporal Convolutional Networks (TCNs) by amalgamating data from disparate time segments through a Transformer encoder. This integration, when combined with an optimized TCN kernel size, ensures exceptional accuracy in activity recognition tasks.

關鍵字(中)

★ 物聯網 (IoT)
★ WiFi
★ 通道狀態信息 (CSI)
★ 人類活動識別 (HAR)
★ 擴散模型

關鍵字(英)

★ —Internet of Things (IoT)
★ WiFi
★ Channel State Information (CSI)
★ Human Activity Recognition (HAR)
★ Diffusion Models

論文目次

1 Introduction 1
2 Relatedwork 6
2.1 CSI-Based HAR 6
2.1.1 Channel State Information 6
2.1.2 The models of CSI-based HAR 7
2.2 Diffusion 11
2.2.1 DDPM 11
2.2.2 Based on DDPM appraochs generate signal 13
3 Approach 17
3.1 System Architecture 17
3.2 DiffCSI 19
3.2.1 Conditional Generation with Classifier-Free Guidance 19
3.2.2 Efficient and Enriched 1D U-Net Architecture 21
3.3 TCN-Transformer 24
3.3.1 Causal Convolutions 25
3.3.2 Dilated Convolutions 26
3.3.3 Transformer 28
3.4 Data Preprocessing 29
3.4.1 Hampel filter 29
3.4.2 Butterworth filter 30
4 Experiment 33
4.1 Dataset 33
4.2 Data Augmentation 33
4.2.1 With Limited Original Data (50%) 34
4.2.2 With Sufficient Original Data (70%) 36
4.3 Verify Augmented Data 37
4.4 CSI Visualize 39
4.5 Overall experiment 39
5 Conclusion 43

參考文獻

[1] F. Schneider, O. Kamal, Z. Jin, and B. Scholkopf, "Moˆusai: Text-to-music generation with long-context latent diffusion," 2023.
[2] Y. Lin, Y. Gao, B. Li, and W. Dong, "Revisiting indoor intrusion detection with wifi signals: Do not panic over a pet!," IEEE Internet of Things Journal, vol. 7, no. 10, pp. 10437–10449, 2020.
[3] V. Bianchi, M. Bassoli, G. Lombardo, P. Fornacciari, M. Mordonini, and I. De Munari, "Iot wearable sensor and deep learning: An integrated approach for personalized human activity recognition in a smart home environment," IEEE Internet of Things Journal, vol. 6, no. 5, pp. 8553–8562, 2019.
[4] S. An and U. Y. Ogras, "Mars: mmwave-based assistive rehabilitation system for smart healthcare," vol. 20, sep 2021.
[5] R. Espinosa, H. Ponce, S. Gutiérrez, L. Martínez-Villaseñor, J. Brieva, and E. Moya-Albor, "A vision-based approach for fall detection using multiple cameras and convolutional neural networks: A case study using the up-fall detection dataset," Computers in Biology and Medicine, vol. 115, p. 103520, 2019.
[6] A. Alarifi and A. Alwadain, "Killer heuristic optimized convolution neural network-based fall detection with wearable iot sensor devices," Measurement, vol. 167, p. 108258, 2021.
[7] F. Jin, A. Sengupta, and S. Cao, "mmfall: Fall detection using 4-d mmwave radar and a hybrid variational rnn autoencoder," IEEE Transactions on Automation Science and Engineering, vol. 19, no. 2, pp. 1245–1257, 2022.
[8] Z. Hussain, Q. Z. Sheng, and W. E. Zhang, "A review and categorization of techniques on device-free human activity recognition," Journal of Network and Computer Applications, vol. 167, p. 102738, 2020.
[9] I. Nirmal, A. Khamis, M. Hassan, W. Hu, and X. Zhu, "Deep learning for radio-based human sensing: Recent advances and future directions," IEEE Communications Surveys & Tutorials, vol. 23, pp. 995–1019, 2020.
[10] S. Yousefi, H. Narui, S. Dayal, S. Ermon, and S. Valaee, "A survey on behavior recognition using wifi channel state information," IEEE Communications Magazine, vol. 55, no. 10, pp. 98–104, 2017.
[11] W. Wang, A. X. Liu, M. Shahzad, K. Ling, and S. Lu, "Understanding and modeling of wifi signal based human activity recognition," in Proceedings of the 21st Annual International Conference on Mobile Computing and Networking, MobiCom ′15, (New York, NY, USA), p. 65–76, Association for Computing Machinery, 2015.
[12] J. Yang, X. Chen, D. Wang, H. Zou, C. X. Lu, S. Sun, and L. Xie, "Sensefi: A library and benchmark on deep-learning-empowered wifi human sensing," Patterns, vol. 4, no. 3, 2023.
[13] F. Wang, J. Feng, Y. Zhao, X. Zhang, S. Zhang, and J. Han, "Joint activity recognition and indoor localization with wifi fingerprints," IEEE Access, vol. 7, pp. 80058–80068, 2019.
[14] P. F. Moshiri, M. Nabati, R. Shahbazian, and S. A. Ghorashi, "CSI-based human activity recognition using convolutional neural networks," in 2021 11th International Conference on Computer Engineering and Knowledge (ICCKE), pp. 7–12, 2021.
[15] Z. Chen, L. Zhang, C. Jiang, Z. Cao, and W. Cui, "WiFi CSI based passive human activity recognition using attention based BLSTM," IEEE Transactions on Mobile Computing, vol. 18, no. 11, pp. 2714–2724, 2019.
[16] C.-Y. Lin, Y.-T. Liu, C.-Y. Lin, and T. K. Shih, "TCN AA: A WiFi based temporal convolution network for human to human interaction recognition with augmentation and attention," arXiv preprint arXiv:2305.18211, 2023.
[17] C. Xiao, D. Han, Y. Ma, and Z. Qin, "CSIGAN: Robust channel state information-based activity recognition with GANs," IEEE Internet of Things Journal, vol. 6, no. 6, pp. 10191–10204, 2019.
[18] D. Wang, J. Yang, W. Cui, L. Xie, and S. Sun, "Robust CSI-based human activity recognition using roaming generator," in 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), pp. 1329–1334, 2020.
[19] D. Wang, J. Yang, W. Cui, L. Xie, and S. Sun, "Multimodal CSI-based human activity recognition using GANs," IEEE Internet of Things Journal, vol. 8, no. 24, pp. 17345–17355, 2021.
[20] J. Ho, A. Jain, and P. Abbeel, "Denoising diffusion probabilistic models," ArXiv, vol. abs/2006.11239, 2020.
[21] Y. Ma, G. Zhou, and S. Wang, "WiFi sensing with channel state information: A survey," ACM Comput. Surv., vol. 52, jun 2019.
[22] S. Shang, Q. Luo, J. Zhao, R. Xue, W. Sun, and N. Bao, "LSTM-CNN network for human activity recognition using WiFi CSI data," Journal of Physics: Conference Series, vol. 1883, p. 012139, apr 2021.
[23] A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. u. Kaiser, and I. Polosukhin, "Attention is all you need," in Advances in Neural Information Processing Systems (I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, eds.), vol. 30, Curran Associates, Inc., 2017.
[24] Z. Kong, W. Ping, J. Huang, K. Zhao, and B. Catanzaro, "DiffWave: A versatile diffusion model for audio synthesis," arXiv preprint arXiv:2009.09761, 2020.
[25] G. Mariani, I. Tallini, E. Postolache, M. Mancusi, L. Cosmo, and E. Rodolà, "Multi-source diffusion models for simultaneous music generation and separation," arXiv preprint arXiv:2302.02257, 2023.
[26] J. Ho, A. Jain, and P. Abbeel, "Denoising diffusion probabilistic models," Advances in neural information processing systems, vol. 33, pp. 6840–6851, 2020.
[27] N. Chen, Y. Zhang, H. Zen, R. J. Weiss, M. Norouzi, and W. Chan, "WaveGrad: Estimating gradients for waveform generation," arXiv preprint arXiv:2009.00713, 2020.
[28] N. Neifar, A. Ben-Hamadou, A. Mdhaffar, and M. Jmaiel, "DiffECG: A versatile probabilistic diffusion model for ECG signals synthesis," 2023.
[29] E. Adib, A. S. Fernandez, F. Afghah, and J. J. Prevost, "Synthetic ECG signal generation using probabilistic diffusion models," IEEE Access, vol. 11, pp. 75818–75828, 2023.
[30] J. Ho, "Classifier-free diffusion guidance," ArXiv, vol. abs/2207.12598, 2022.
[31] O. Ronneberger, P. Fischer, and T. Brox, "U-net: Convolutional networks for biomedical image segmentation," ArXiv, vol. abs/1505.04597, 2015.
[32] R. Alazrai, A. Awad, B. Alsaify, M. Hababeh, and M. I. Daoud, "A dataset for Wi-Fi-based human-to-human interaction recognition," Data in Brief, vol. 31, p. 105668, 2020.
[33] S. Tan, Y. Ren, J. Yang, and Y. Chen, "Commodity WiFi sensing in ten years: Status, challenges, and opportunities," IEEE Internet of Things Journal, vol. 9, no. 18, pp. 17832–17843, 2022.
[34] F. Yu and V. Koltun, "Multi-scale context aggregation by dilated convolutions," in 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, May 2-4, 2016, Conference Track Proceedings (Y. Bengio and Y. LeCun, eds.), 2016.
[35] S. Huang, P.-Y. Chen, and J. McCann, "DiffAR: Adaptive conditional diffusion model for temporal-augmented human activity recognition," in Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, IJCAI-23 (E. Elkind, ed.), pp. 3812–3820, International Joint Conferences on Artificial Intelligence Organization, 8 2023. Main Track.

指導教授

施國琛

審核日期

2024-7-30

推文