A probabilistic approach for phase estimation in single-channel speech enhancement using von mises phase priors
A probabilistic approach for phase estimation in single-channel speech enhancement using von mises phase priors
复制标题
使用 von Mises 相位先验进行单通道语音增强中相位估计的概率方法
DOI：
10.1109/mlsp.2014.6958861
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Mario Kaoru Watanabe
中科院分区：
文献类型：
--
作者：
Josef Kulmer;Pejman Mowlaee Begzade Mahale;Mario Kaoru Watanabe
在许多人工智能系统中，人类的声音被认为是信息传输的媒介。当语音与背景噪声混合时，通过语音进行人机通信变得困难。作为一种补救措施，单通道语音增强是必不可少的，以减少背景噪声从嘈杂的语音，使其适用于自动语音识别和电话语音。虽然传统的单通道语音增强技术在幅度估计和信号重建阶段都包含噪声相位，但在本文中，我们提出了一种概率方法来估计噪声观测中的干净语音相位。我们提出的方法包括相位展开，然后使用冯米塞斯相位先验的基于阈值的时间平滑。所提出的相位增强方法导致改善语音质量和可懂度预测的仪器措施没有明确纳入幅度增强。
In many artificial intelligence systems human voice is considered as the medium for information transmission. Human-machine communication by voice becomes difficult when speech is mixed with some background noise. As a remedy, a single-channel speech enhancement is indispensable for reducing background noise from noisy speech to make it suitable for automatic speech recognition and telephony speech. While the conventional techniques for single-channel speech enhancement incorporate noisy phase in both amplitude estimation and signal reconstruction stages, in this paper we propose a probabilistic method to estimate the clean speech phase from noisy observation. Our proposed method consists of phase unwrapping followed by threshold-based temporal smoothing using von Mises phase priors. The proposed phase enhancement method leads to improved speech quality and intelligibility predicted by instrumental measures without explicit incorporation of amplitude enhancement.