The conventional sub-band methods can improve the recognition accuracy of noisy speech, but degrade that of clean speech.
传统的子带特征方法虽然能提高噪声下的语音识别性能，但通常会使无噪声情况下的识别性能下降。
Speech enhancement tries to extract clean speech signal from original one with noise and improve the SNR of speech signal.
语音增强则从含噪信号中提取干净的语音信号，提高语音信号的信噪比。
In order to extract the underlying clean speech signal from the noise speech signal and reduce the noise disturbance to the minimum, speech enhancement is needed.
为了从带噪语音信号中获得尽可能纯净的语音信号，减少噪音的干扰，就需进行语音增强。
Owing to the lack of telephone speech data, this paper proposes a software simulation implementation of converting clean speech sounds into telephone-quality ones.
针对电话语料比较缺乏的问题，提出了一种完全由软件模拟实现由纯净语音向电话质量语音转换的算法。
In this paper, a parallel sub-state hidden Markov model, which integrates the clean speech and noise information, and each state of the model has several parallel sub-states, is presented.
提出了一种平行子状态隐马尔可夫模型用作噪声鲁棒语音识别的声学模型。
In this paper, a parallel sub-state hidden Markov model, which integrates the clean speech and noise information, and each state of the model has several parallel sub-states, is presented.
提出了一种平行子状态隐马尔可夫模型用作噪声鲁棒语音识别的声学模型。
应用推荐