BriefGPT.xyz
Sep, 2016
一种用于语音增强的全卷积神经网络
A Fully Convolutional Neural Network for Speech Enhancement
HTML
PDF
Se Rim Park, Jinwon Lee
TL;DR
本文提出了使用全卷积神经网络来解决低信噪比下听力辅助设备中存在的背景噪音干扰人类语音理解的问题。通过有监督学习找到含噪语音的频谱和纯净语音频谱之间的映射，提出了冗余卷积编码解码（R-CED）网络，该网络大小为递归网络的1/12，表现更好，能够应用于嵌入式系统：听力辅助设备。
Abstract
In
hearing aids
, the presence of
babble noise
degrades hearing intelligibility of human speech greatly. However, removing the babble without creating artifacts in human speech is a challenging task in a low SNR e
→