当前位置:
X-MOL 学术
›
arXiv.cs.SD
›
论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A Recursive Network with Dynamic Attention for Monaural Speech Enhancement
arXiv - CS - Sound Pub Date : 2020-03-29 , DOI: arxiv-2003.12973 Andong Li, Chengshi Zheng, Cunhang Fan, Renhua Peng, Xiaodong Li
arXiv - CS - Sound Pub Date : 2020-03-29 , DOI: arxiv-2003.12973 Andong Li, Chengshi Zheng, Cunhang Fan, Renhua Peng, Xiaodong Li
A person tends to generate dynamic attention towards speech under complicated
environments. Based on this phenomenon, we propose a framework combining
dynamic attention and recursive learning together for monaural speech
enhancement. Apart from a major noise reduction network, we design a separated
sub-network, which adaptively generates the attention distribution to control
the information flow throughout the major network. To effectively decrease the
number of trainable parameters, recursive learning is introduced, which means
that the network is reused for multiple stages, where the intermediate output
in each stage is correlated with a memory mechanism. As a result, a more
flexible and better estimation can be obtained. We conduct experiments on TIMIT
corpus. Experimental results show that the proposed architecture obtains
consistently better performance than recent state-of-the-art models in terms of
both PESQ and STOI scores.
中文翻译:
用于单耳语音增强的具有动态注意的递归网络
一个人倾向于在复杂环境下对语音产生动态注意。基于这种现象,我们提出了一个将动态注意力和递归学习相结合的框架,用于增强单声道语音。除了一个主要的降噪网络,我们还设计了一个分离的子网络,它自适应地生成注意力分布以控制整个主要网络的信息流。为了有效地减少可训练参数的数量,递归学习被引入,这意味着网络被重复用于多个阶段,其中每个阶段的中间输出与记忆机制相关。结果,可以获得更灵活和更好的估计。我们在 TIMIT 语料库上进行实验。
更新日期:2020-04-02
中文翻译:
用于单耳语音增强的具有动态注意的递归网络
一个人倾向于在复杂环境下对语音产生动态注意。基于这种现象,我们提出了一个将动态注意力和递归学习相结合的框架,用于增强单声道语音。除了一个主要的降噪网络,我们还设计了一个分离的子网络,它自适应地生成注意力分布以控制整个主要网络的信息流。为了有效地减少可训练参数的数量,递归学习被引入,这意味着网络被重复用于多个阶段,其中每个阶段的中间输出与记忆机制相关。结果,可以获得更灵活和更好的估计。我们在 TIMIT 语料库上进行实验。