Search - first and follow algorithm

Search - first and follow algorithm - List

[Speech/Voice recognition/combine] DTWspeech DL : 0: 本文首先介绍了语音识别的研究和发展状况，然后循着语音识别系统的处理过程，介绍了语音识别的各个步骤，并对每个步骤可用的几种方法在实验基础上进行了分析对比。研究了语音信号的预处理和特征参数提取，包括语音信号的数字化、分帧加窗、预加重滤波、端点检测及时域特征向量和变换域特征向量.其中端点检测采用双门限法.通过实验比对特征参数的选取，采用12阶线性预测倒谱系数作为识别参数。详细分析了特定人孤立词识别算法，选定动态时间弯折为识别算法，并重点介绍其设计实现。在 Vi su alC++环境下，设计并实现一个特定人、孤立词语音识别系统，系统可以识别数字0-9等简单指令。该系统还具备演示、学习功能，可以演示语音处理的各个步骤，还可以根据需要添加新的指令。最后，重点从端点检测算法和动态时间弯折识别算法对系统进行改进。实验表明，改进后的系统识别率有很大提高，达到95 ，为进一步开发实用性语音识别系统产品打下了基础。-This article introduced the first speech recognition research and development, and then follow the voice recognition system Processing, speech recognition, introduced the various steps, each step of the methods available in the real A post-mortem conducted on the basis of the analysis and comparison. Research on the speech signal pre-processing and feature extraction, including Digitized voice signals, sub-frame window, pre-emphasis filtering, endpoint detection feature vector in time domain and variable Eigenvector for the domain. One endpoint detection method using dual-threshold. Through experiments over the selection of characteristic parameters, The use of 12-order linear prediction cepstral coefficients as recognition parameters. Detailed analysis of the specific operator who isolated word recognition Law, selected Dynamic Time Warping Algorithm for identifying and focusing on the achievement of its design. In Vi su alC++ environment, design and realization of a s
Date : 2025-12-26 Size : 2.38mb User : 周文超
[Speech/Voice recognition/combine] JLDATA DL : 0: 摘要：本论文主要研究了语音识别的基本原理,对语音识别系统的构成进行分析处理,其中包括预处理、特征参数提取、建立模块库、识别匹配几大部分。预处理又包括语音采样、预加重、加窗（汉明窗）、端点检测；特征提取的参数是梅尔频率倒谱系数MFCC。该语音系统采用的是动态时间伸缩算法(DTW)，研究对象是特定人的语音识别，并在MATLAB平台上实现。为了进行后续研究，首先使用电脑中的录音系统录制了阿拉伯数字0—9的语音文件，并转化成 “.wav”格式的文件。-Abstract: This thesis mainly studied the basic principle of speech recognition, to analyze the composition of the speech recognition system, including the preprocessing, feature extraction, to set up the module library, identify several most matches. Pretreatment, including speech sampling, pre-emphasis, add window (hamming window), endpoint detection Feature extraction of MFCC MEL frequency cepstrum coefficient. The voice system USES a dynamic time scale (DTW) algorithm, the research object is the speaker-dependent speech recognition, and realized in MATLAB platform.To carry out the follow-up study, the first to use the recording in a computer system to record the audio files of Arabic Numbers 0-9, and translated into . Wav format file.
Date : 2025-12-26 Size : 9kb User : silver teng