This paper mainly discusses the speech enhancement algorithm based on the Minimum Mean Square Error (MMSE) estimation of speech short time spectrum.
本文主要讨论基于语音短时谱估计的语音增强算法。
And to compensate the enhanced signal's energy loss and spectrum distortion, several reasonable schemes of Speech compensation are touched upon in the paper.
以补偿增强后信号的能量损失和频谱失真为目的，浅谈了几种语音补偿可实现的算法及设想。
Objective \ to analyze the spectrum of vowels in the sequence therapy of patient with cleft palate in order to provide an objective basis for speech evaluation and correction.
目的分析腭裂病人序列治疗中元音的频谱特点，为腭裂病人的语音评价及语音矫治提供理论及临床上的指导。
This paper proposes a new speech enhancement scheme based on modified Minimum Mean Square Error (MMSE) estimation of speech Short Time Spectrum Amplitude (STSA).
本文提出了一种改进型语音短时谱最小均方误差(MMSE)估计的增强方法。
Labels like ADD, ADHD, speech and language disorders, learning disabilities, and the autistic spectrum disorders may actually represent an increasing severity of sensory integration dysfunction.
像ADD，ADHD，口头与语言失调，学习困难，自闭也许实际上代表着感觉统合失调的严重性。
Traditional speech signal processing is based on FFT power spectral analysis, in noisy condition, the spectrum of noise and speech signal are processed in the same way.
传统语音信号谱特征的提取是基于FFT 的能谱分析方法，在噪音环境情况下，对噪音的频谱成分与语音信号的频谱成分的处理采用“平均主义”的原则。
By the debugging, the coding of speech is realized; m sequence is produced. Moreover, the spread of signal frequency spectrum is realized.
通过调试实现了语音的编码、产生了m序列，并实现了信号频谱的扩展。
Noise power spectrum estimation is a fundamental component of speech enhancement.
噪声功率谱估计是语音增强系统的一个重要组成部分。
The proposed approach is tested on linear spectrum pair (LSP) parameters of speech signals.
提出了一种线谱对参数预测多级矢量量化联合优化算法。
The frequency pseudocolor-coded speech spectrogram is then obtained by means of a sampling narrow slit placed at Fourier plane in white-light optical spectrum analyzer.
然后在白光光学频谱分析系统中，利用放置在傅里叶频谱面上的取样狭缝得到按频率假彩色编码的频谱图。
Based on speech transformation and representation using adaptive interpolation of weighted spectrum (STRAIGHT), a wideband speech coding algorithm is presented.
提出了一种基于自适应加权谱内插(STRAIGHT)的宽带语音编码算法。
During the noise estimation, the estimation of its spectrum is updated by tracking the speech-absent frames.
噪声估计过程中通过跟踪带噪语音帧来更新噪声估计。
Correct noise power spectrum estimation and apriori signal-to-noise ratio (SNR) estimation are all essential to good quality of the enhanced speech.
噪声功率谱估计和先验信噪比估计的准确性有助于提高增强后的语音质量。
The spectrum of auditory model and LPC-CEP shows that this auditory model not only represents speech signal well, but also is noise robust.
通过对听觉谱和L PC倒谱对比分析，得到了听觉谱适宜用作语音识别并具有良好的噪声鲁棒性的结论。
Autism lies on the more severe end of the spectrum whereas Asperger's is a milder condition that doesn't involve speech impairment (as classic autism does);
自闭症位于谱系较为严重的一端；阿斯伯格综合症不包括语言障碍（而典型自闭症包括），是较为缓和的谱系障碍；
Autism lies on the more severe end of the spectrum whereas Asperger's is a milder condition that doesn't involve speech impairment (as classic autism does);
自闭症位于谱系较为严重的一端；阿斯伯格综合症不包括语言障碍（而典型自闭症包括），是较为缓和的谱系障碍；
应用推荐