端点检测是语音识别中重要的一环。
提出了一种基于时频方差和的语音端点检测算法。
An algorithm for speech endpoint detection based on time-frequency-variance-summation was proposed.
提出了一种基于指数门限(et)的端点检测方法。
A new endpoint detection method based on the exponential threshold (et) is proposed.
语音端点检测的准确性直接影响着语音识别系统性能。
The accuracy of the speech endpoint detection is important to the recognition performance.
提出了一种基于滑动窗口的综合语音端点检测方法。
Optimal algorithm of data streams clustering on sliding window model;
不采用端点检测，在语音帧内及噪声帧内都进行噪声更新。
The noise estimation is updated in both speech frame and noise frame without the voice activity detection.
实验结果表明，该方法可以得到较高正确率的端点检测结果。
The experiment result proves that the method has better detection result.
研究了噪声环境下，利用短时能量为特征进行语音端点检测的问题。
This paper analyzes speech endpoint detection based on short-term energy feature in the presence of noise.
不同背景噪声下的实验结果表明，用该熵可以得到高正确率的端点检测。
The experiments in different noise backgrounds show that high endpoint detection accuracy can be obtained by this entropy.
提出了基于DCT(离散余弦变换)增强和改进谱熵的语音端点检测方法。
In the paper, a speech endpoint detection method based on DCT (Discrete Cosine Transform) enhancement and improved spectral entropy is proposed.
语音段起止端点检测是语音分析、语音合成和语音识别中的一个必要环节。
Speech endpoint detection is a paragraph beginning and end speech analysis, speech synthesis and speech recognition of a necessary link.
通过噪音评估，调节录音增益，调整端点检测方法参 数来提高语音识别率。
The invention is characterized in that: by evaluating noise, adjusting recording gain and adjusting port detection parameters to improve speech rate.
为提高浊音端点检测的准确率和效率，提出一种基于循环自相关函数的检测方法。
To enhance the accuracy and efficiency of endpoint detection, a detection method based on Circular Autocorrelation Function(CACF) is proposed.
对不同类型噪声环境下的语音进行了端点检测，并对检测效果进行了评价和分析。
Endpoint detection is made to speech in different kinds of noise environments, evaluation and analysis are given to detection results.
介绍了短时平均能量法、短时平均过零率法和短时能零积法三种语音端点检测法。
Short-time average energy, short-time average zero-crossing rate and short-time energy-zero-product are introduced.
二是在端点检测上进行了算法改进，分别采用了动态窗长及零能积差的阈值判决法。
Second, the end of the algorithm to improve the detection, long and dynamic Windows were used to zero threshold from poor judgment algorithms.
研究了数字语音短时能量和过零率特点，提出了基于有限状态机的端点检测新算法。
The characteristics of digital voice short energy and ZRC is studied, and a new voice activity detection algorithm based on finite state machine is presented.
本文提出了基于谱熵和谱减法相结合的带噪语音端点检测改进算法以及端点检测的判决准则。
In this paper, we propose a new approach based on spectral entropy and spectral subtraction for noisy speech endpoint detection, and discriminative rules with robustness.
另外，在前端的端点检测中采用一种新的对数能量特征作为判别依据，以进一步改善识别效果。
In addition, to gain a higher rate, a new logarithmic energy feature is adopted in endpoint detection.
为提高实时通信中语音端点检测系统的性能，提出了一种基于能量和鉴别信息的端点检测算法。
A new algorithm based on the energy and discrimination information was developed to improve the performance of the voice activity detection system in real-time speech communications.
为提高实时通信中语音端点检测系统的性能，提出了一种基于能量和鉴别信息的端点检测算法。
On discussing the defects of the traditional voice activity detection method based on cepstrum distance, this paper proposes an improved project.
提出一种采用累积量矩阵的最大奇异值来实现语音端点检测的方法，并引入一种自适应的实现方法。
The proposed method uses the maximum singular value of an cumulant matrix to distinguish between voiced parts of the speech signal and noise.
在此基础上提出了IC A能量(ICAE)和滤波icae (FICAE)特征来进行端点检测。
On the basis, the characteristics called the ICA Energy (ICAE) and the Filtered ICAE (FICAE) are used for the noisy signal endpoint detection.
本文提出了两种特征处理方法：特征的似然度加权和基于散度的维数缩减，来提高噪声下端点检测的性能。
This paper proposes two new methods: feature weighted likelihood and divergence based dimension reduction to improve detecting performance in noise.
因此在如何提高语音端点检测系统的鲁棒性的前提下，加强检测系统的稳定性是当今语音端点检测的目标方向。
Therefore, how to improve endpoint detection system's robustness and increasing the stability of the system is ours direction.
在较低信噪比情况下，基于语音信号的短时相对自相关序列的短时平均幅度的端点检测能够获得较高的检测精度。
The endpoint detection based on short-time average magnitude of speech signals relative autocorrelation sequences can be detected in high accuracy under the low signal-to noise ratio.
语音信号的端点检测技术就是从包含语音的一段信号中准确地确定语音的起始点和终止点，区分语音和非语音信号。
The endpoint detection technology of speech signal is to accurately determine starting point and ending point from a section of speech signal. Thus it can distinguish speech and non-speech signal.
该方法利用了分形维数在噪声情况下作为语音端点检测参数的优越性，克服了在噪声情况下判决门限难以估计的问题。
The first is the endpoint detection based on fractal dimension. It utilizes fractal dimension superiority and overcomes the difficulty of decision threshold in noise environment.
语音识别系统的实用化，需要对噪声有很强的鲁棒性，而噪声环境下的端点检测对整个识别系统性能起着关键的作用。
While speech recognition system is put into use, it must be robust to noise. The endpoint detection in noisy background plays an important role in the whole recognition system.
语音识别系统的实用化，需要对噪声有很强的鲁棒性，而噪声环境下的端点检测对整个识别系统性能起着关键的作用。
While speech recognition system is put into use, it must be robust to noise. The endpoint detection in noisy background plays an important role in the whole recognition system.
应用推荐