arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.09963eess.AScs.LGcs.SDeess.SP
收藏

TSTNN:基于两级Transformer的时域语音增强神经网络

TSTNN: Two-stage Transformer based Neural Network for Speech Enhancement in the Time Domain

Concordia University(康考迪亚大学)

Kai Wang, Bengbeng He, Wei-Ping Zhu

摘要:

在本文中,我们提出了一种基于变换的架构,称为两级Transformer神经网络(TSTNN)的端到端的语音去噪在时域。该模型由编码器、两级Transformer模块(TSTM)、掩蔽模块和解码器组成。编码器将输入的噪声语音映射为特征表示。TSTM利用四个堆叠的两级Transformer块来逐级有效地从编码器输出中提取局部和全局信息。掩码模块创建将与编码器输出相乘的掩码。最后,解码器使用掩蔽的编码器特征来重建增强的语音。在基准数据集上的实验结果表明,TSTNN在时域或频域上优于大多数最先进的模型,同时具有显着降低的模型复杂度。

英文摘要:

In this paper, we propose a transformer-based architecture, called two-stage transformer neural network (TSTNN) for end-to-end speech denoising in the time domain. The proposed model is composed of an encoder, a two-stage transformer module (TSTM), a masking module and a decoder. The encoder maps input noisy speech into feature representation. The TSTM exploits four stacked two-stage transformer blocks to efficiently extract local and global information from the encoder output stage by stage. The masking module creates a mask which will be multiplied with the encoder output. Finally, the decoder uses the masked encoder feature to reconstruct the enhanced speech. Experimental results on the benchmark dataset show that the TSTNN outperforms most state-of-the-art models in time or frequency domain while having significantly lower model complexity.

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。

↑