arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1903.08876eess.AScs.SD
收藏

基于DNN声源增强的数据驱动完美重构滤波器组设计

Data-driven design of perfect reconstruction filterbank for DNN-based sound source enhancement

Waseda University(早稻田大学);NTT Media Intelligence Laboratories(NTT媒体情报实验室)

Daiki Takeuchi, Kohei Yatabe, Yuma Koizumi, Yasuhiro Oikawa, Noboru Harada

摘要:

提出了一种基于深度神经网络(DNN)的声源增强(SSE)完全重构滤波器组(PRFB)的数据驱动设计方法。DNN已被用于在短时傅里叶变换(STFT)域中估计时频(T-F)掩模。当使用简单的代价函数均方误差(MSE)时,它们的训练比一些先进的代价(如客观音质评估)更稳定。然而,这样一个简单的代价函数继承了对目标和/或噪声的统计的强假设,这通常是不满足的,并且假设的失配导致性能下降。在本文中,我们提出了设计的频率范围的PRFB从训练数据,使MSE的假设是满意的。为了设计频率尺度,变形滤波器组框架(WFBF)被认为是PRFB。学习的WFBF的频率特性介于STFT和小波变换之间,并通过与标准的基于STFT的DNN(其输入特征被压缩到梅尔尺度)的比较来证实其有效性。

英文摘要:

We propose a data-driven design method of perfect-reconstruction filterbank (PRFB) for sound-source enhancement (SSE) based on deep neural network (DNN). DNNs have been used to estimate a time-frequency (T-F) mask in the short-time Fourier transform (STFT) domain. Their training is more stable when a simple cost function as mean-squared error (MSE) is utilized comparing to some advanced cost such as objective sound quality assessments. However, such a simple cost function inherits strong assumptions on the statistics of the target and/or noise which is often not satisfied, and the mismatch of assumption results in degraded performance. In this paper, we propose to design the frequency scale of PRFB from training data so that the assumption on MSE is satisfied. For designing the frequency scale, the warped filterbank frame (WFBF) is considered as PRFB. The frequency characteristic of learned WFBF was in between STFT and the wavelet transform, and its effectiveness was confirmed by comparison with a standard STFT-based DNN whose input feature is compressed into the mel scale.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑