Description: 语音识别中经典的HMM算法，包括产生序列、测试和参数训练，由C语言编写。- In speech recognition classical HMM algorithm, including has
the sequence, the test and the parameter training, compiles by the C
language. Platform: |
Size: 13312 |
Author: |
Hits:
Description: 语音识别配套的VQ及DHMM模型训练程序，C语言，已经定点化，可直接移植到8位MCU或16位DSP中。与目前市面的语音识别玩具的算法基本一致，非常实用，仅供大家参考，别去抢人家饭碗才好。-speech recognition and matching VQ DHMM model training procedures, C language, has been positioned, can be directly transplanted to eight 16-bit MCU or DSP. And the current market speech recognition algorithm toys basically the same, very practical, is for your reference. stealing jobs other people a better life. Platform: |
Size: 6388736 |
Author:xubin |
Hits:
Description: 说话人识别是语音识别的一种特殊方式，其目的不是识别语音内容，而是识别说话人是谁，即从语音信号中提取个人特征。采用矢量量化（VQ）可避免困难的语音分段问题和时间归整问题，且作为一种数据压缩手段可大大减少系统所需的数据存储量。本文提出了识别特征选取采用复倒谱特征参数和对应用VQ的说话人识别系统改进的一种方法。当用于训练的数据量较小时，复倒谱特征可以得到比较稳定的识别性能。VQ的改进方法避免了说话人识别系统的训练时间与使用时间相差过长从而导致系统的性能明显下降以及若利用自相关函数带来的大量运算。-Speaker Recognition Speech Recognition is a special way, its purpose is not voice recognition, Who identification but said that the voice signal from extracting personal characteristics. Vector quantization (VQ) can avoid the difficulties subparagraph voice to the issues and the whole time, and as a means of data compression system can significantly reduce the required data storage capacity. This paper presents a selection of identifiers employ Cepstrum parameters and the application of VQ speaker recognition system to improve a side France. When training for the amount of data is smaller, rehabilitation Cepstrum be relatively stable recognition performance. VQ improved ways to avoid the speech recognition system of training and the use of the difference in time, resulting in excessive sys Platform: |
Size: 23552 |
Author:张开 |
Hits:
Description: 基于隐马尔科夫的语音识别及语音合成，包括两个处理过程，训练过程和识别过程-Hidden Markov based speech recognition and speech synthesis, including the two deal with the process, training process and the recognition process Platform: |
Size: 2319360 |
Author:李秀环 |
Hits:
Description: 智能语音识别避障机器人
// 语音识别+机器人+超声波测距综合应用方案
// 采用特定人识别技术，程序开始时用户需要对语音识别进行训练，每
// 条指令训练两次，训练成功后，才开始真正的语音辨识，针对用户发
// 出的不同语音指令，机器人执行不同的动作，在动作过程中进行超声
// 波测距，遇到障碍物停止动作，并发射飞盘-Intelligent voice recognition robot obstacle avoidance// speech recognition+ Robot+ Ultrasonic rangefinder integrated application program// using specific recognition technology, users need the beginning of the proceedings of the speech recognition training, each// instructions training twice , training success, the only real speech recognition, user-fat// a different voice commands, robot perform different actions, in the course of action to carry out ultrasound// wave ranging, encountered obstacles to stop action, and fired Frisbee Platform: |
Size: 183296 |
Author:会自 |
Hits:
Description: This code implements in C++ a basic left-right hidden Markov model
and corresponding Baum-Welch (ML) training algorithm. It is meant as
an example of the HMM algorithms described by L.Rabiner (1) and
others. Serious students are directed to the sources listed below for
a theoretical description of the algorithm. KF Lee (2) offers an
especially good tutorial of how to build a speech recognition system
using hidden Markov models. Platform: |
Size: 15360 |
Author:aaaaaaa |
Hits:
Description: Hidden_Markov_model_for_automatic_speech_recognition
This code implements in C++ a basic left-right hidden Markov model
and corresponding Baum-Welch (ML) training algorithm. It is meant as
an example of the HMM algorithms described by L.Rabiner (1) and
others. Serious students are directed to the sources listed below for
a theoretical description of the algorithm. KF Lee (2) offers an
especially good tutorial of how to build a speech recognition system
using hidden Markov models. Platform: |
Size: 23552 |
Author: |
Hits:
Description: 0-9语音识别，基于matlab的实现，包含了训练和识别的功能！-0-9 voice recognition, matlab based on the realization, including the training and identification of the function! Platform: |
Size: 93184 |
Author:yue |
Hits:
Description: 基于小波包的孤立词语音识别技术，构建了一个关于方向信息的孤立词非特定人语音识别系统。给出了从模型训练到识别的实现过程。-Based on Wavelet Packet isolated word speech recognition technology, to build an information on the direction of non-specific people isolated word speech recognition system. From the model are given training to realize the process of identification. Platform: |
Size: 18776064 |
Author:wangyan |
Hits:
Description: 完成基于MFCC的语音识别系统仿真，文件中带有完成语音识别训练的源文件-Completion of MFCC-based speech recognition system simulation, document the completion of speech recognition training with the source file Platform: |
Size: 807936 |
Author:zhao |
Hits:
Description: 数字的语音识别matlab程序
训练及识别数据太大,少量上传一点,大家可以用COOLEDIT等软件仿照自己制作.(*.wav文件采样率8000,单声道,采样精度16位,16bit Motorola PCM,*.lab文件存的是语音的起止点和语音内容). -Digit speech recognition matlab program too much training and recognition data, a small amount of upload that you can use software such as CoolEdit modeled to produce their own. (*. Wav file sampling rate 8000, mono, 16-bit sampling precision, 16bit Motorola PCM,*. lab files maintained by the beginning and ending points of speech and voice content). Platform: |
Size: 314368 |
Author:er |
Hits:
Description: 一种简单有效的基于动态时变语音识别源码
对于大多数研究者来说,寻找能够匹配二重时间序列信号的最佳途径是很重要的,因为它有许多重要的应用需求.DTW是实现这项工作的显著技术,尤其在语音识别技术领域,在这里一个测试信号被按照参照模板拉伸或压缩,
-Searching for the best path that matches two time-series signals is the main task for many researchers, because of its importance in these applications. Dynamic Time-Warping (DTW) is one of the prominent techniques to accomplish this task, especially in speech recognition systems. DTW is a cost minimisation matching technique, in which a test signal is stretched or compressed according to a reference template.
Although there are other advanced techniques in speech recognition such as the hidden Markov modelling (HMM) and artificial neural network (ANN) techniques, the DTW is widely used in the small-scale embedded-speech recognition systems such as those embedded in cell phones. The reason for this is owing to the simplicity of the hardware implementation of the DTW engine, which makes it suitable for many mobile devices. Additionally, the training procedure in DTW is very simple and fast, as compared with the HMM and ANN rivals.
Platform: |
Size: 2658304 |
Author:宋小小 |
Hits:
Description: ：为了使应力变异在顽健语音识别系统中能够达到较好的识别效果，研究了基于隐马
尔可夫模型(HMM)的自适应技术，提出了将最大后验概率(MAP)和最大似然回归方法(MLLR)用
于应力变异语音的自适应中。实验结果表明，与基本系统相比，两种方法均有效地提高系统识别
率。以SD为初始模型的最大后验概率方法在150个训练样本时识别效果最好，可以达到90．4％ 。-: In order to stress variation in the robustness of speech recognition system can achieve better recognition results, based on Hidden Markov Model (HMM) of adaptive technology, put forward a maximum a posteriori probability (MAP) and Maximum Likelihood regression (MLLR) for the stress of the adaptive variation in voice. The experimental results show that compared with the basic system, both methods are effective to improve the system recognition rate. SD as the initial model to the maximum a posteriori probability method in 150 training samples to identify the best, can reach 90.4 . Platform: |
Size: 234496 |
Author:尹江波 |
Hits:
Description: This Microsoft Speech Recognition training upload. It can add training directly to speech engine. I have found this code at some other forum and looking for converting it to csharp for use in my project. Platform: |
Size: 2777088 |
Author:Umaid |
Hits:
Description: DTW（Dynamic Time Warping，动态时间归整）算法，该算法基于动态规划（DP）的思想，解决了发音长短不一的模板匹配问题，是语音识别中出现较早、较为经典的一种算法。用于孤立词识别，DTW算法与HMM算法在训练阶段需要提供大量的语音数据，通过反复计算才能得到模型参数，而DTW算法的训练中几乎不需要额外的计算。所以在孤立词语音识别中，DTW算法仍然得到广泛的应用。 -DTW (Dynamic Time Warping, dynamic time warping) algorithm based on dynamic programming (DP) ideas, sounds of varying lengths to solve the template matching problem, is speech recognition appeared earlier, more classic kind of algorithm. For isolated word recognition, DTW algorithm and HMM algorithm in the training phase need to provide a large number of voice data, obtained by repeated calculations to model parameters, while the DTW algorithm is almost no additional training calculations. Therefore, in isolated word speech recognition, DTW algorithm is still widely used. Platform: |
Size: 6144 |
Author:fujuan |
Hits:
Search - training speech recognition