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 2109.05977eess.AScs.SD
收藏

CNN中用于说话人确认的压缩激励方法研究

Studying squeeze-and-excitation used in CNN for speaker verification

LIA, Avignon University(阿维尼翁大学实验室信息学与自动化研究所(LIA))

Mickael Rouvier, Pierre-Michel Bousquet

摘要:

在说话人确认中,语音表示的提取主要基于残差神经网络(ResNet)架构。ResNet建立在卷积层之上,卷积层学习过滤器来捕获所有输入的局部空间模式,然后生成联合编码空间和通道信息的特征图。不幸的是,卷积层中的所有特征图都是独立学习的(卷积层不利用特征图之间的依赖关系)和本地学习的。这个问题首先在图像处理中得到解决。最近在卷积层中提出了一种称为挤压和激励(SE)的通道注意机制,并将其应用于说话人确认。该机制重新加权跨特征图提取的信息。在本文中,我们首先提出了一个原始的定性研究的影响和作用的SE机制应用于说话人确认任务在不同阶段的ResNet,然后评估几个SE架构。最后,我们建议使用基于均值和标准差合并的新合并变体来改进SE方法。结果表明,仅在ResNet的第一阶段应用SE可以更好地捕获用于验证任务的说话人信息,并且使用所提出的池化变体,已经注意到Voxceleb 1-E,Voxceleb 1-H和SITW评估任务的显著区分增益。

英文摘要:

In speaker verification, the extraction of voice representations is mainly based on the Residual Neural Network (ResNet) architecture. ResNet is built upon convolution layers which learn filters to capture local spatial patterns along all the input, then generate feature maps that jointly encode the spatial and channel information. Unfortunately, all feature maps in a convolution layer are learnt independently (the convolution layer does not exploit the dependencies between feature maps) and locally. This problem has first been tackled in image processing. A channel attention mechanism, called squeeze-and-excitation (SE), has recently been proposed in convolution layers and applied to speaker verification. This mechanism re-weights the information extracted across features maps. In this paper, we first propose an original qualitative study about the influence and the role of the SE mechanism applied to the speaker verification task at different stages of the ResNet, and then evaluate several SE architectures. We finally propose to improve the SE approach with a new pool- ing variant based on the concatenation of mean- and standard- deviation-pooling. Results showed that applying SE only on the first stages of the ResNet allows to better capture speaker information for the verification task, and that significant discrimination gains on Voxceleb1-E, Voxceleb1-H and SITW evaluation tasks have been noted using the proposed pooling variant.

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。

↑