【摘要】：近年來(lái),隨著人工智能的不斷發(fā)展,語(yǔ)音識(shí)別技術(shù),已經(jīng)逐漸從研究階段進(jìn)入到實(shí)際應(yīng)用階段,是一項(xiàng)潛在研究?jī)r(jià)值較高的技術(shù)。但是,在語(yǔ)音識(shí)別系統(tǒng)的研究中,如何優(yōu)化系統(tǒng)性能,仍然是現(xiàn)在討論的焦點(diǎn)。文中詳細(xì)介紹了整個(gè)系統(tǒng)的基本構(gòu)成及其原理,對(duì)語(yǔ)音識(shí)別系統(tǒng)的個(gè)別關(guān)鍵技術(shù)進(jìn)行了深入研究,并提出了相應(yīng)的改進(jìn)算法。語(yǔ)音識(shí)別大致流程包括:語(yǔ)音端點(diǎn)檢測(cè)、特征參數(shù)的提取以及語(yǔ)音模型的訓(xùn)練與識(shí)別算法。首先,本文對(duì)語(yǔ)音識(shí)別系統(tǒng)的部分關(guān)鍵技術(shù),包括語(yǔ)音信號(hào)的預(yù)處理、端點(diǎn)檢測(cè)以及特征提取算法,進(jìn)行了深入的研究。在低信噪比噪聲環(huán)境下,對(duì)信號(hào)的端點(diǎn)檢測(cè)和基音周期提取這兩個(gè)關(guān)鍵技術(shù)提出了相應(yīng)的改進(jìn)算法,分別是:基于經(jīng)驗(yàn)?zāi)Ｊ椒纸?EMD)和改進(jìn)小波熵的端點(diǎn)檢測(cè)算法和一種基于小波包變換加權(quán)自相關(guān)的基音周期提取算法,并與原始算法相比較。其次,本文選取Mel倒譜系數(shù)為特征參數(shù),并仔細(xì)研究了MFCC特征參數(shù)的提取過(guò)程,提出了一種基于小波包變換的抗噪語(yǔ)音特征參數(shù)-WPTMFCC特征參數(shù)。實(shí)驗(yàn)表明,新的特征參數(shù)能提高系統(tǒng)的魯棒性,在不同信噪比噪聲環(huán)境下識(shí)別率相比傳統(tǒng)LPCC特征參數(shù)和MFCC特征參數(shù)分都有所提高。本文在MATLAB平臺(tái)上搭建了一個(gè)基于隱馬爾科夫模型(HMM)的識(shí)別系統(tǒng)。通過(guò)對(duì)比仿真實(shí)驗(yàn),證明了改進(jìn)的端點(diǎn)檢測(cè)技術(shù)和WPTMFCC特征參數(shù)能提高系統(tǒng)的識(shí)別率。最后,設(shè)計(jì)出識(shí)別系統(tǒng)的GUI界面,通過(guò)此界面可以對(duì)語(yǔ)音庫(kù)中的語(yǔ)音進(jìn)行實(shí)時(shí)識(shí)別演示。
[Abstract]:In recent years, with the continuous development of artificial intelligence, speech recognition technology has gradually moved from the research stage to the practical application stage, is a potential research value of the technology. However, in the research of speech recognition system, how to optimize the system performance is still the focus of discussion. In this paper, the basic structure and principle of the whole system are introduced in detail, some key technologies of speech recognition system are deeply studied, and corresponding improved algorithms are put forward. The general flow of speech recognition includes speech endpoint detection, feature parameter extraction, speech model training and recognition algorithm. Firstly, some key technologies of speech recognition system, including speech signal preprocessing, endpoint detection and feature extraction algorithm, are studied in this paper. In the environment of low SNR noise, two key techniques of signal endpoint detection and pitch period extraction are proposed. They are: an endpoint detection algorithm based on empirical mode decomposition (EMD) and improved wavelet entropy and an algorithm of pitch period extraction based on wavelet packet transform weighted autocorrelation and compared with the original algorithm. Secondly, the Mel cepstrum coefficient is selected as the feature parameter, and the extraction process of MFCC feature parameter is studied carefully, and a feature parameter -WPTMFCC for anti-noise speech based on wavelet packet transform is proposed. The experimental results show that the new feature parameters can improve the robustness of the system, and the recognition rate in different SNR noise environments is higher than that of the traditional LPCC feature parameters and MFCC feature parameters. In this paper, a recognition system based on Hidden Markov Model (hmm) is built on MATLAB platform. The simulation results show that the improved endpoint detection technique and the characteristic parameters of WPTMFCC can improve the recognition rate of the system. Finally, the GUI interface of the recognition system is designed, through which the speech in the speech database can be recognized and demonstrated in real time.
【學(xué)位授予單位】：安徽工業(yè)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN912.34

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張思陽(yáng);徐敏強(qiáng);王日新;高晶波;;EMD與樣本熵在往復(fù)壓縮機(jī)氣閥故障診斷中的應(yīng)用[J];哈爾濱工程大學(xué)學(xué)報(bào);2014年06期

2 王琳;李成榮;;一種基于自適應(yīng)譜熵的端點(diǎn)檢測(cè)改進(jìn)方法[J];計(jì)算機(jī)仿真;2010年12期

3 趙毅;尹雪飛;陳克安;;一種新的基于倒譜的共振峰頻率檢測(cè)算法[J];應(yīng)用聲學(xué);2010年06期

4 鄭繼明;王勁松;;語(yǔ)音基音周期檢測(cè)方法[J];計(jì)算機(jī)工程;2010年10期

5 李宏梅;伍小芹;;有關(guān)語(yǔ)音識(shí)別技術(shù)的研究[J];現(xiàn)代電子技術(shù);2010年08期

6 陳磊;吳小培;呂釗;;基于線性預(yù)測(cè)與歸一化互相關(guān)的基音檢測(cè)[J];電子測(cè)量技術(shù);2009年10期

7 張軍;李學(xué)斌;;一種基于DTW的孤立詞語(yǔ)音識(shí)別算法[J];計(jì)算機(jī)仿真;2009年10期

8 劉建新;曹榮;趙鶴鳴;;一種LPC改進(jìn)算法在提取耳語(yǔ)音共振峰中的應(yīng)用[J];西華大學(xué)學(xué)報(bào)(自然科學(xué)版);2008年03期

9 胡瑛;陳寧;;基于小波變換的清濁音分類及基音周期檢測(cè)算法[J];電子與信息學(xué)報(bào);2008年02期

10 張玲華;鄭寶玉;楊震;;基于LPC分析的語(yǔ)音特征參數(shù)研究及其在說(shuō)話人識(shí)別中的應(yīng)用[J];南京郵電學(xué)院學(xué)報(bào);2005年06期

相關(guān)碩士學(xué)位論文 前10條

1 廖振東;基于DTW的孤立詞語(yǔ)音識(shí)別系統(tǒng)研究[D];云南大學(xué);2015年

2 王一蒙;語(yǔ)音識(shí)別關(guān)鍵技術(shù)研究[D];電子科技大學(xué);2015年

3 劉方洲;語(yǔ)音識(shí)別關(guān)鍵技術(shù)及其改進(jìn)算法研究[D];長(zhǎng)安大學(xué);2014年

4 李W毞，

本文編號(hào)：2128734