搜索资源列表
EndPoint
- 一个语音信号端点检测的程序，算法可能需要根据具体的应用调整一些参数。请使用者自己调整，根据不同的噪声环境。
vad1
- 语音端点检测算法，在matlab里面实现语音端点的准确检测。为读者开发噪声环境下的精确VAD提供思路。-voice endpoint detection algorithm in Matlab voice inside the precise endpoint detection. Readers development under noisy environments to provide accurate VAD ideas.
航空噪声背景下的语音端点检测和语音增强算法
- 航空噪声背景下的语音端点检测和语音增强算法，在信噪比很低的强航空噪声背景下，效果特别好
ASKBPSk
- 噪声背景下的语音端点检测和语音增强算法，在信噪比很低的强航空噪声背景下，-aviation noise background voice endpoint detection and voice enhancement algorithms, and very low signal to noise ratio in the strong aviation background noise, particularly effective
enframe
- 语音端点检测算法，在matlab里面实现语音端点的准确检测。为读者开发噪声环境下的精确VAD提供思路.-voice endpoint detection algorithm in Matlab voice inside the precise endpoint detection. Readers development under noisy environments to provide accurate VAD ideas.
lee
- 语音采集加噪声程序 语音端点检测 已经调试很好用 -The voice acquisition plus the speech endpoint detection of the noise program debugging has been very good,
zaoshenghuajing-xiayuyin-jiance
- 语音的端点检测 噪声可以自己加 使用阈值进行检测 能量的阈值跟据经验来进行决定-Speech endpoint detection noise can add to the threshold detection energy threshold according to experience to the decision
CS
- CS端点检测,基于语音压缩感知的语音在不同噪声下的语音端点检测程序-compressed sensing
duandianjiance
- 基于支持向量机的语音信号端点检测，主要是去掉首位的噪声。-Voice signal endpoint detection based on support vector machines, mainly to get rid of the first noise.
wiener
- 一种基于非稳态噪声估计的语音增强算法。语音端点检测VAD。维纳语音增强-winner speech enhancement VAD
frequency-band-variance-of-VAD
- 语音和噪声在频谱域中的特性差异较大，基于此给出频带方差的端点检测程序。-Speech and noise have differences in the spectral domain feature, based on the given frequency band variance program applied in the speech endpoint detection.
log--spectral-distance-for-VAD
- 语音增强之前先进行端点检测，对有话帧和噪声帧作不同处理，利用对数频谱距离进行端点检测。-Speech enhancement for endpoint detection, before to have words for different frame and noise frame processing, using logarithmic spectrum distance for endpoint detection.
vad
- 双门限端点检测。因为低噪声导致假过零率的产生，进行了算法改进。可以直接调用函数运行，亲测-Dual threshold endpoint detection. Because of the low noise cause false over-produce zero rate, the algorithm was improved. It can directly call the function run, pro-test
vad
- 关于端点检测的几种方法，语音样本是自己录制的，对传统算法做了一些改进，加入了去噪，去噪之后再进行端点检测，均调通 vad0303：自己设置调整门限为一定值 vad0310：根据能量值和过零率设置门限，自适应门限值 vad0310_2：基于比例因子的门限自调整 vad0310.m加入了噪声，端点检测前都噪声进行了处理 entropy.m：基于自适应子带频谱熵的稳健性语音端点检测 可用于语音增强及端点检测 dbdoor.m：双门限算法，用于语音端点检测。可以通过调整门限值，并
用短时平均能量和短时平均过零率双门限检测VAD
- 用短时平均能量和过零率的特征参数对语音信号进行端点检测， 以消除静默帧和噪声帧的干扰