搜索资源列表
Application
- Speaker recognition is the process of automatically recognizing who is speaking on the basis of individual information included in speech waves. This technique makes it possible to user s voice to verify their identity and control access to services
Final_Program
- Speaker recognition is the process of automatically recognizing who is speaking on the basis of individual information included in speech waves. This technique makes it possible to user s voice to verify their identity and control access to services
1234-bp
- bp神经网络识别语音信号，是个很好的学习资料-BP neural network recognition of speech signals, is a very good learning materials
fasl6-soal1
- speech recognition and synthesise
fasl6-soal5
- speech recognition practies
fasl8-soal6
- speech recognition and practies
TamrinCH3
- speech recognition and practies
Multichannel-Codec-Based-on-2812
- 基于TMS320F2812的多通道实时语音编解码-An Implementation of Multichannel Real Time Speech Codec Based on TMS320F2812
A-New-Word-Separation-Algorithm-for-Continuous.ra
- A New Word Separation Algorithm for Continuous Bangia Speech Recognition
dtw
- DTW算法 教程，可以用于语音识别，图像识别等等。 DTW algorithm tutorial, can be used for speech recognition, image recognition, etc-MATLAB voice processing program, speech enhancement, speech pretreatment, very rich, very useful for reference.
DTWword
- DTW算法用于语音识别的资料介绍，不错的参考资料。-DTW algorithm for speech recognition of information, a good reference.
nlms
- New algorithm for speech compression noise, a noise reduction of speech based vad method. This new algorithm derived vad technology and spectral subtraction, it firstly deal with the noise signal vad, then the masking threshold is based on perception
0712-MATLAB-adpcm_decoder
- 实现adpcm语音编码，已经试行过，在dm324系统，很好的一个编码- Implementation of ADPCM speech coding, a very good coding
eusipco_presentation
- multi-expert neural networks for speech recognition
hmmSVMhybrid
- Speech Recognition with Support Vector Machines in a Hybrid Syste
hmmSVMybrid2
- A Hybrid SVM/HMM Acoustic Modeling Approach to Automatic Speech Recognition
Research-on-MELP
- 主要介绍了一种新的低比特率MELP(Mixed Excitation Linear Prediction)混合激励线性预测语音编码技术,其中着重分析了该编码算法所采用的几项关键技术.给出了采用MELP压缩编码算法后的输出比特流在各个参数上的比特分配表,并通过计算机仿真,对MELP合成语音与原始语音的质量做了比较,最后就 MELP语音编码技术与现今其它几种不同的低速率语音编码技术的合成语音质量在DRT、DAM及MOS得分三个方面做了比较.-This paper mainly introduces
MELP
- 文基于LPC的自适应前后向量化技术,提出了一种可变速率的混合激励线性预测MELP语音编码算法.该算法中,采用当前语音帧(前向LPC)或前面某帧已合成语音帧(后向LPC)进行线性预测,当采用后向LPC时,只需传输时间序列编码,故减少了LPC系数的平均编码比特.计算机模拟表明,该算法与标准MELP算法合成的语音质量相当,但显著减少了LPC的传输带宽,从而明显降低了MELP平均编码速率.-VARIABLE-RATE MELP SPEECH CODER BASED ON ADAPTIVE FORWARD
MELPcodepaper
- 目前2.4kbps的混合激励线性预测 (MELP) 语音编码方法已经被确定为美国新的联邦语音编码标准.本文提出了一种改进的MELP语音编码方法,利用滤波器相似度和基于LPC系数分类的矢量量化技术,可以把MELP的码率降到1.7kbps以下,仍有较好的合成语音质量.-An Improved MELP Speech Coder
melp5
- 在混合激励线性预测(MELP)语音编码算法中,语音帧的突变转换导致合成语音质量的下降成为一个突出问题.为解决该问题,提出一种基于过渡帧判决算法的改进MELP模型,提高了参数估计的准确度,有效实现了语音的降噪处理.-In the hybrid excitation linear prediction (MELP) speech coding algorithm, the abrupt change of the speech frame leads to the degradation of sy