Postfiltering Using an Adversarial Denoising Autoencoder with Noise-aware Training
Postfiltering Using an Adversarial Denoising Autoencoder with Noise-aware Training
复制标题
DOI：
10.1109/icassp.2019.8682684
复制
发表时间：
2019-05
期刊：
影响因子：
--
通讯作者：
Naohiro Tawara;Hikari Tanabe;Tetsunori Kobayashi;Masaru Fujieda;Kazuhiro Katagiri;T. Yazu;Tetsuji Ogawa
中科院分区：
文献类型：
--
作者：
Naohiro Tawara;Hikari Tanabe;Tetsunori Kobayashi;Masaru Fujieda;Kazuhiro Katagiri;T. Yazu;Tetsuji Ogawa
提出了一种具有噪声感知训练的对抗性去噪自动编码器（ADAE），并成功地应用于线性降噪的后滤波。ADAE对于衰减干扰声音是有效的，然而，很难学会处理其各种意想不到的有害影响（例如，各种类型的噪声）。传统语音增强被引入作为预处理器，以通过减少ADAE输入中的意外变化来有效地训练ADAE。时频掩蔽能很好地抑制信号的变异性，但也会引起令人不快的失真，这是ADAE难以弥补的。在本文中，最小方差无失真响应（MVDR）波束形成器，它可以避免麻烦的非线性失真，被利用作为预处理器，和MVDR输出被用作输入到基于ADAE的后置滤波器。此外，从MVDR波束形成器导出的噪声主导信号可以提高基于ADE的后置滤波器的精度，因为残余噪声取决于原始噪声信号。使用多通道语音增强进行的实验比较表明，基于ADAE的后滤波产生显着的改善MVDR和基于ADAE的语音增强系统，和噪声感知训练的ADAE工作得很好。
An adversarial denoising autoencoder (ADAE) with noise-aware training is proposed and successfully applied to post-filtering for linear noise reduction. The ADAE is effective for attenuating interference sounds, however, it is difficult to learn to handle its various unexpected harmful effects (e.g., various types of noise) using a single network. Legacy speech enhancement was introduced as a pre-processor to make it possible to efficiently train the ADAEs by reducing the unexpected variabilities in the inputs to the ADAEs. Time-frequency masking performed well to suppress the variabilities, however, it induced unpleasant distortion, which is difficult for the ADAE to complement. In this paper, a minimum variance distortionless response (MVDR) beam-former, which can avoid troublesome non-linear distortions, is exploited as a preprocessor, and the MVDR outputs are used as the inputs to the ADAE-based post-filter. In addition, noise-dominant signals derived from the MVDR beamformer can improve the accuracy of the ADAE-based post-filter because the residual noise depends on the original noise signals. Experimental comparisons conducted using multichannel speech enhancement demonstrate that ADAE-based post-filtering yields significant improvements over the MVDR-and ADAE-based speech enhancement systems, and noise-aware training of ADAE works well.