- 學位論文
結合語音和圖像識別技術實現智能家居安全系統
Combining Voice and Image Recognition for Smart Home Security System
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
由於技術和硬體的進步，人類可以掌握比以前更大量的數據。機器學習是實現人工智能的一種方式，使用不同的演算法處理不同類型的數據，讓系統能夠像人類一樣分析、學習、處理數據。人工神經網絡有很多種類，如回歸神經網絡，擅長處理連續性數據，語音識別等情況，由於語音是連續性的，不同的字母組合具有不同的含義，所以連續性數據對於語音識別至關重要。 樹莓派是一種微型電腦，您可以在其中安裝操作系統。樹莓派可作為一個基礎;你可以擴充不同的模組，如果我們想要收集圖像信息，可以安裝鏡頭模組，如果要獲取聲音數據，只需要安裝麥克風模組。目前的樹莓派已經內建WIFI和藍牙的模組，所以樹莓派目前是物聯網設備最受歡迎的選擇之一。在本文中，我們希望將人臉識別和語音識別技術應用到生活中，使用樹莓派建置居家安全系統。我們提出了一種結合人臉識別和語音識別的架構來建置居家安全系統。透過臉部和語音識別，檢查身份是否正確，以提高智能家居環境的便利性和安全性。 本篇論文是以樹莓派為居家安全系統的核心，使用鏡頭模組和麥克風模組來收集人臉及語音的資料。 使用樹莓派將數據上傳到臉部API和語音API。臉部API用於辨識臉部信息，語音API用於辨識語音信息。運用模糊邏輯結合臉部API及語音API判斷身分，驗證通過系統將執行下一步控制設備，如電子鎖或支持遠程控制的其他設備。
並列摘要
Through technique and hardware improvements, the human can handle more significant data than before. Machine learning is a way to achieve artificial intelligence, using difference algorithm to process difference type data, let the system analyzing, learning, processing data like the way of a human. The Artificial neural network has many types, such as the Recurrent Neural Network that good at handling continuity data, the situation like speech recognition, since speech is continuity, different letters combination with the various meaning, so continuity data is essential to speech recognition. Raspberry Pi is an excellent Internet of Things device because it is a kind of micro-computer, you can install the operating system in it. The function is as a base; you can expand different modules. For example, if we want to collect image information, we can install the lens module, if we're going to obtain sound data, we need to install the microphone module. The current raspberry pi already has WIFI and Bluetooth, so the raspberry pie is currently one of the favorite choices for IoT devices. In this paper, we want to apply the technique of face recognition and speech recognition in life, using a raspberry pi to build a home security system. We proposed an architecture that combines face recognition and voice recognition to develop a home security system. Through face and voice recognition, we double-checking the identity is correct or not to improve convenience and security of the smart home environment. Raspberry Pi is the home security system's kernel and uses lens module to collect face information and microphone module to collect voice information. Raspberry Pi upload the data to the face API and speaker recognition system. Using fuzzy logic to combine face API and speech API, improve the performance of the system. When completing the verification requirement; the system will continue next step to control devices, like an electronic lock or other devices that supported remote control.
參考文獻
[1]. Ben-Yacoub, S., Abdeljaoued, Y., & Mayoraz, E. (1999). Fusion of face and speech data for person identity verification. IEEE Transactions on Neural Networks, 10(5), 1065-1074.
[2]. Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M. (2013). Internet of Things (IoT): A vision, architectural elements, and future directions. Future Generation Computer Systems, 29(7), 1645-1660.
[3]. Jain, S., Vaibhav, A., & Goyal, L. (2014). Raspberry Pi based interactive. Paper presented at the 2014 International Conference on Reliability, Optimization and Information Technology, India.
[4]. Chihaoui, M., Elkefi, A., Bellil, W., & Ben Amar, C. (2016). A survey of 2D face recognition techniques. Computers, 5(4), 21.
[5]. Sirovich, L., & Kirby, M. (1986). Low-dimensional procedure for the characterization of human faces. Journal of the Optical Society of America A, 4, 519-524.