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 2108.10752eess.AScs.AIcs.SD
收藏

通过稀疏自注意层将RNN转换器推广到域外音频

Generalizing RNN-Transducer to Out-Domain Audio via Sparse Self-Attention Layers

SK Telecom(SK电讯);Kakao Enterprise Corporation(Kakao企业公司)

Juntae Kim, Jeehye Lee

摘要:

递归神经网络转换器(RNN-T)是一个端到端的语音识别框架,它将输入的声学帧转换为字符序列。RNN-T最先进的编码器网络是Conformer,它可以通过卷积和自注意层有效地对局部-全局上下文信息进行建模。尽管Conformer RNN-T表现出了出色的性能,但大多数研究都是在训练和测试数据来自同一领域的情况下进行验证的。Conformer RNN-T的域失配问题是产品级语音识别系统中的一个重要问题,但目前还没有得到深入研究。在这项研究中,我们发现,完全连接的自我注意层的一致性导致高删除错误,特别是在长形式的域外话语。为了解决这个问题,我们为基于Conformer的编码器网络引入了稀疏自注意层,它可以通过修剪大多数域内拟合的全局连接来利用局部和广义全局信息。此外,我们提出了一个状态重置方法的预测网络的泛化,以应付长形式的话语。将所提出的方法应用于域外测试,与基于全连接自注意层的构象相比,我们获得了27.6%的相对字符错误率(CER)降低。

英文摘要:

Recurrent neural network transducer (RNN-T) is an end-to-end speech recognition framework converting input acoustic frames into a character sequence. The state-of-the-art encoder network for RNN-T is the Conformer, which can effectively model the local-global context information via its convolution and self-attention layers. Although Conformer RNN-T has shown outstanding performance, most studies have been verified in the setting where the train and test data are drawn from the same domain. The domain mismatch problem for Conformer RNN-T has not been intensively investigated yet, which is an important issue for the product-level speech recognition system. In this study, we identified that fully connected self-attention layers in the Conformer caused high deletion errors, specifically in the long-form out-domain utterances. To address this problem, we introduce sparse self-attention layers for Conformer-based encoder networks, which can exploit local and generalized global information by pruning most of the in-domain fitted global connections. Also, we propose a state reset method for the generalization of the prediction network to cope with long-form utterances. Applying proposed methods to an out-domain test, we obtained 27.6% relative character error rate (CER) reduction compared to the fully connected self-attention layer-based Conformers.

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。

↑