近幾年中AI已漸漸成為生活中密不可分的一環，其中在家庭安全方面更是有智慧住宅、智慧家庭的概念出現，並透過AI影音辨識提升居家安全、便利性以及管理效率。本篇論文使用深度學習技術並結合低成本低解析度無線網路攝影機擷取影像資訊作為視覺系統之輸入，並提出三個關鍵技術。(1)基於骨架資訊輔以頭部影像之多人身分辨識：本技術使用卷積神經網路進行骨架資訊擷取並輔以彩色RGB影像之頭部特徵來進行身分辨識。一般家庭中人數約在二至六人，在小規模資料比對中骨架資訊具有相當的可靠度，可彌補以往在身分辨識只能使用臉部進行辨識的問題。在身分辨識網路中，本篇論文提出一具分支之輕量型卷積神經網路，可有效結合骨架以及頭部特徵資訊，在減少參數的同時並提升分類準確率。(2)車庫門狀態辨識技術：本技術使用深度殘差網路進行訓練，並在網路訓練階段導入資料增強進行大量旋轉、調整大小、比例尺寸，以及改變亮度、翻轉等處理，透過這種學習方式可有效提升網路對於輸入圖片之空間與結構之注意力。(3)火焰偵測技術：本技術使用全卷積神經網路，使輸入圖像將不再受到特定大小的限制，在分類及回歸任務上由於不再依賴於全連接層，更是有效減少了參數量以及預測時間。實驗結果顯示在正面身分辨識的情況下可以得到87.74%的準確率，背面身分辨識的情況下則為80.24%的準確率。在車庫狀態辨識中準確率為95.01%。在火焰偵測則達到95.33%的偵測準確率。本系統實作多種關鍵技術確實地提升了居家安全。

In recent years, artificial intelligence has gradually become an indispensable part of life. Among them, the concept of smart home and smart home has appeared in the field of home security, and through AI image recognition to improve home security, convenience and management efficiency. This paper uses deep learning technology in combination with low-cost, low-resolution wireless network cameras to capture image information as input to the visual system, and proposes three key technologies. (1)Multi-person recognition based on skeleton information supplemented by head image: We uses a convolutional neural network for skeleton information extraction and supplemented by head features of color RGB images for identity recognition. There are about two to six people in a family. The skeleton information is quite reliable in the comparison of small-scale data, which can make up for the problem that only face recognition can be used for identification in the past. In the identity recognition network, a branched light-weight convolution neural network is proposed, which can effectively combine skeleton and head feature information, reduce parameters, and improve classification accuracy. (2) Garage door status classification: This technology uses a deep residual network for training, and uses the data augmentation function to rotate, resize, change brightness, and flip during network training. This learning method can effectively increase the input picture space and structural attention of the network. (3) Fire detection: This technology uses a fully convolution neural network, so the input image will no longer be limited by a specific size. Since it no longer depends on fully connected layers for classification and regression tasks, it can also effectively reduce the amount of parameters and implement time. The experimental results show that the accuracy rate of 87.74% can be achieved in the case of human body front recognition, and the accuracy rate of 80.24% can be achieved in the case of human body back recognition. The accuracy rate of garage door status classification is 95.01%. In fire detection, the detection accuracy is 95.33%. The system implements a variety of key technologies to improve home safety and security.

[1] H. Li, S. Gong, L. Lai, Z. Han, R. C. Qiu and D. Yang, "Efficient and Secure Wireless Communications for Advanced Metering Infrastructure in Smart Grids," in IEEE Transactions on Smart Grid, vol. 3, no. 3, pp. 1540-1551, Sept. 2012, doi: 10.1109/TSG.2012.2203156.
[2] Raj SV. Implementation of pervasive computing based high-secure smart home system. In: 2012 IEEE International Conference on computational intelligence & computing research (ICCIC), 18–20 December; 2012. p. 1–8.
[3] Lian, Kuang-Yow, Sung-Jung Hsiao, and Wen-Tsai Sung. "Smart home safety handwriting pattern recognition with innovative technology." Computers & Electrical Engineering 40.4 (2014): 1123-1142.
[4] Yang, Lili, Shuang-Hua Yang, and Fang Yao. "Safety and security of remote monitoring and control of intelligent home environments." 2006 IEEE International Conference on Systems, Man and Cybernetics. Vol. 2. IEEE, 2006.
[5] Hwang, J.Y., Kang, J.M., Jang, Y.W., Kim, H.C.: Development of Novel Algorithm and Real-time Monitoring Ambulatory System Using Bluetooth Module for Fall Detection in the Elderly. Proceedings of the 26th Annual International Conference of the IEEE EMBS, San Francisco (2004) 2204-2207
[6] de Leeuw, Karl; Bergstra, Jan (2007). The History of Information Security: A Comprehensive Handbook. Amsterdam:Elsevier.pp. 264265. ISBN 9780444516084.
[7] A. J. Goldstein, L. D. Harmon and A. B. Lesk, "Identification of human faces," in Proceedings of the IEEE, vol. 59, no. 5, pp. 748-760, May 1971, doi: 10.1109/PROC.1971.8254.
[8] Sirovich, Lawrence & Kirby, M. (1987). Low-Dimensional Procedure for the Characterization of Human Faces. Journal of the Optical Society of America. A, Optics and image science. 4. 519-24. 10.1364/JOSAA.4.000519.
[9] Turk, Matthew, and Alex Pentland. "Eigenfaces for recognition." Journal of cognitive neuroscience 3.1 (1991): 71-86.
[10] Ojala, Timo, Matti Pietikäinen, and David Harwood. "A comparative study of texture measures with classification based on featured distributions." Pattern recognition 29.1 (1996): 51-59.
[11] G. B. Huang, M. Ramesh, T. Berg, and E. Learned-Miller, "Labeled faces in the wild: A database for studying face recognition in unconstrained environments," Technical Report 07-49, University of Massachusetts, Amherst2007.
[12] Y. Taigman, M. Yang, M. A. Ranzato, and L. Wolf, "Deepface: Closing the gap to human-level performance in face verification," in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1701-1708, 2014.
[13] Lu, Chaochao, and Xiaoou Tang. "Surpassing human-level face verification performance on LFW with GaussianFace." Twenty-ninth AAAI conference on artificial intelligence. 2015.
[14] F. Schroff, D. Kalenichenko, and J. Philbin, "Facenet: A unified embedding for face recognition and clustering," in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 815-823, 2015.
[15] Scardino G., Infantino I., Vella F. (2013) Recognition of Human Identity by Detection of User Activity. In: Marinos L., Askoxylakis I. (eds) Human Aspects of Information Security, Privacy, and Trust. HAS 2013. Lecture Notes in Computer Science, vol 8030. Springer, Berlin, Heidelberg.
[16] Sim JY, Noh HW, Goo W, Kim N, Chae SH, Ahn CG. Identity Recognition Based on Bioacoustics of Human Body [published online ahead of print, 2019 Oct 4]. IEEE TransCybern.2019;10.1109/TCYB.2019.2941281.doi:10.1109/TCYB.2019.2941281.
[17] Wang, Liang, et al. "Silhouette analysis-based gait recognition for human identification." IEEE transactions on pattern analysis and machine intelligence 25.12 (2003): 1505-1518.
[18] Zeng, Xianhui & Deng, Jinwen & Liu, Mujun & Wang, Yue & Zhang, Xuejun. (2016). A Human Identity Recognition System Based on Kinect Skeletal Tracking and ANN Classifier. 10.2991/aiie-16.2016.57.
[19] A. Glandon et al., "3D Skeleton Estimation and Human Identity Recognition Using Lidar Full Motion Video," 2019 International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 2019, pp. 1-8, doi: 10.1109/IJCNN.2019.8852370.
[20] Copîndean, Romul, Rodica Holonec, and Florin Drăgan. "Phone Call Controlled Garage Door." Acta Electrotehnica 53.4 (2012): 330-333.
[21] R. G. J. Wijnhoven and P. H. N. de With, "Identity verification using computer vision for automatic garage door opening," in IEEE Transactions on Consumer Electronics, vol. 57, no. 2, pp. 906-914, May 2011, doi: 10.1109/TCE.2011.5955239.
[22] Lu, Jiang, and Fangqin Liu. "In-vehicle garage door monitor." U.S. Patent Application No. 15/340,738.
[23] Styers, Justin R., and Ryan H. McDowell. "Remote garage door monitoring system." U.S. Patent No. 7,468,676. 23 Dec. 2008.
[24] (2019). 108年12月火災概況暨1至12月全國火災次數、起火原因及火災損失統計表Available: https://www.nfa.gov.tw/cht/index.php?code=list&flag=detail&
ids=220&article_id=7256.
[25] Töreyin, B. Uğur, et al. "Computer vision based method for real-time fire and flame detection." Pattern recognition letters 27.1 (2006): 49-58.
[26] Albers, Burt W., and Ajay K. Agrawal. "Schlieren analysis of an oscillating gas-jet diffusion flame." Combustion and flame 119.1-2 (1999): 84-94.
[27] T. Celik and Kai-Kuang Ma, "Computer vision based fire detection in color images," 2008 IEEE Conference on Soft Computing in Industrial Applications, Muroran, 2008, pp. 258-263, doi: 10.1109/SMCIA.2008.5045970.
[28] Yuan, Fei-niu, et al. "Feature extraction for computer vision based fire detection." Journal of University of Science and Technology of China 36.1 (2006): 39-43.
[29] R. Girshick, J. Donahue, T. Darrell, and J. Malik, "Region-based convolutional networks for accurate object detection and segmentation," IEEE transactions on pattern analysis and machine intelligence, vol. 38, no. 1, pp. 142-158, 2015.
[30] K. He, X. Zhang, S. Ren, and J. Sun, "Spatial pyramid pooling in deep convolutional networks for visual recognition," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 37, no. 9, pp. 1904-1916, 2014, doi: 10.1109/TPAMI.2015.2389824.
[31] R. Girshick, "Fast R-CNN," presented at the IEEE International Conference on Computer Vision, Washington, DC, USA, Dec. 2015, 2015.
[32] S. Ren, K. He, R. Girshick, and J. Sun, "Faster r-cnn: Towards real-time object detection with region proposal networks," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, pp. 1137-1149, 2015, doi: 10.1109/TPAMI.2016.2577031.
[33] J. R. Uijlings, K. E. Van De Sande, T. Gevers, and A. W. Smeulders, "Selective search for object recognition," International journal of computer vision, vol. 104, no. 2, pp. 154-171, 2013..
[34] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You only look once: Unified, real-time object detection," in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 779-788.
[35] W. Liu et al., "Ssd: Single shot multibox detector," in European conference on computer vision, 2016: Springer, pp. 21-37.
[36] K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," presented at the International Conference on Learning Representations, San Diego, CA, USA, May 2015, 2015.
[37] S. Frizzi, R. Kaabi, M. Bouchouicha, J. Ginoux, E. Moreau and F. Fnaiech, "Convolutional neural network for video fire and smoke detection," IECON 2016 - 42nd Annual Conference of the IEEE Industrial Electronics Society, Florence, 2016, pp. 877-882, doi: 10.1109/IECON.2016.7793196.
[38] K. Muhammad, J. Ahmad, Z. Lv, P. Bellavista, P. Yang and S. W. Baik, "Efficient Deep CNN-Based Fire Detection and Localization in Video Surveillance Applications," in IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 49, no. 7, pp. 1419-1434, July 2019, doi: 10.1109/TSMC.2018.2830099.
[39] Cao, Zhe, et al. "Realtime multi-person 2d pose estimation using part affinity fields." Proceedings of the IEEE conference on computer vision and pattern recognition. 2017.
[40] Lin, Tsung-Yi, et al. "Microsoft coco: Common objects in context." European conference on computer vision. Springer, Cham, 2014.
[41] S.-E. Wei, V. Ramakrishna, T. Kanade, and Y. Sheikh. Convolutional pose machines. In CVPR, 2016.
[42] He, Kaiming, et al. "Deep residual learning for image recognition." Proceedings of the IEEE conference on computer vision and pattern recognition. 2016.
[43] Redmon, Joseph, and Ali Farhadi. "Yolov3: An incremental improvement." arXiv preprint arXiv:1804.02767 (2018).
[44] Figga’s Fire Detection Dataset. Available: https://mivia.unisa.it/datasets/video-analysis-datasets/fire-detection-dataset/.
[45] Jadon, Arpit, et al. "Firenet: A specialized lightweight fire & smoke detection model for real-time iot applications." arXiv preprint arXiv:1905.11922 (2019).