中原大學電機工程學系
Loading
專題成果首頁>專題成果
希爾伯特黃轉換與類神經網路於聲音辨識之應用
指導老師:洪穎怡   組長:曾怡嘉   組員:曾怡嘉
探討到聲音辨識系統,目前台灣社會動盪不安,每個人都擔心自己住的地方是否安全?因此,若有一套需要有特定聲音或是有特定暗號才能進入屋內的語音辨識系統,想必對民眾的生命財產安全保護上有很大助益。研究語音辨識的門禁系統,防止竊盜等犯罪現象,才能確保每個人的生命財產安全。另外,目前市面上有許多家電產品,也試圖加上語音控制,來增加使用上的便利性。由此可知,語音辨識勢必是一個趨勢。

本文提出一個新的辨識方法,結合希爾伯特黃轉換(Hilbert-Huang Transform,縮寫HHT)之特徵與類神經網路(Neural Network,縮寫NN)之訓練,應用於聲音之辨識。
In this unstable society, a lot of people are worried about whether their home is safe enough to live in or not. Therefore, if there is a security system that recognizes specific voice or specific signals, it will without a doubt prevent some dangers and strengthens the protection on people’s lives and properties. Voice recognition system can prevent theft and other criminal activities, and that is what we need to work on to ensure everyone’s safety.
  In addition, in order to make our lives more convenient, there are already many home appliances that have added an alternate controlling system which can recognize voices. In conclusion, voice recognition is certainly on its way to become a trend.
  This paper presents a new identification method, combined with the Hilbert Huang transform (Hilbert-Huang Transform, abbreviation HHT) features and artificial neural network’s (Neural Network, abbreviations, NN) training, applied to the recognition of voices.
我們決定將壓電陶瓷晶片和鞋子做結合。
   
 
32023 桃園市中壢區中北路200號 No-.200, Zhongbei Rd., Zhongli City, Taoyuan County 320, Taiwan(ROC) TEL:03-2654801 FAX:03-2654899