HGCN：用于语音增强的谐波门控补偿网络
HGCN: Harmonic gated compensation network for speech enhancement
摘要：
基于神经网络的时频域掩蔽处理已成为单通道语音增强的主流之一。然而，这是很难为大多数模型来处理的情况下，谐波被部分掩盖的噪声。为了应对这一挑战，我们提出了谐波门控补偿网络（HGCN）。为了提高谐波位置预测的准确性，我们设计了一种高分辨率的谐波积分谱。然后，我们加入语音活动检测（VAD）和浊音区域检测（VRD）的卷积递归网络（CRN）过滤谐波位置。最后，谐波门控机制用于指导补偿模型调整CRN的粗略结果，以获得精细增强的结果。我们的实验表明，HGCN在社区中的一些先进的方法实现了实质性的收益。
英文摘要：
Mask processing in the time-frequency (T-F) domain through the neural network has been one of the mainstreams for single-channel speech enhancement. However, it is hard for most models to handle the situation when harmonics are partially masked by noise. To tackle this challenge, we propose a harmonic gated compensation network (HGCN). We design a high-resolution harmonic integral spectrum to improve the accuracy of harmonic locations prediction. Then we add voice activity detection (VAD) and voiced region detection (VRD) to the convolutional recurrent network (CRN) to filter harmonic locations. Finally, the harmonic gating mechanism is used to guide the compensation model to adjust the coarse results from CRN to obtain the refinedly enhanced results. Our experiments show HGCN achieves substantial gain over a number of advanced approaches in the community.
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。
