arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.04702cs.SDeess.AS
收藏

基于CIF的非自回归端到端ASR的边界和上下文感知训练

Boundary and Context Aware Training for CIF-based Non-Autoregressive End-to-end ASR

Northwestern Polytechnical University(西北工业大学);China Mobile Research Institute(中国移动研究院)

Fan Yu, Haoneng Luo, Pengcheng Guo, Yuhao Liang, Zhuoyuan Yao, Lei Xie, Yingying Gao, Leijing Hou, Shilei Zhang

摘要:

基于连续积分激发(CIF)模型的非自回归(NAR)语音识别方法采用了一种软的、单调的对齐机制,与其他NAR方法相比,具有很好的性能。然而,这样的对准学习策略可能遭受错误的声学边界估计,严重阻碍收敛速度以及系统性能。在本文中,我们提出了一个边界和上下文感知的训练方法CIF基于NAR模型。首先,利用连接主义时间分类(CTC)锋电位信息指导CIF中的声学边界学习。此外,在CIF解码器之后引入了额外的上下文解码器,旨在捕获句子内的语言依赖。最后,我们采用了最近提出的Conformer架构,以提高声学建模的能力。在开源汉语AISHELL-1语料库上的实验表明,该方法与现有的AR模型相比,在只有1/24延迟的情况下,字符出错率达到了4.9%.然而,当评估内部7500小时的汉语语料库,我们的模型仍然优于其他NAR方法,甚至达到AR Conformer模型在一个具有挑战性的现实世界的噪音测试集。

英文摘要:

Continuous integrate-and-fire (CIF) based models, which use a soft and monotonic alignment mechanism, have been well applied in non-autoregressive (NAR) speech recognition with competitive performance compared with other NAR methods. However, such an alignment learning strategy may suffer from an erroneous acoustic boundary estimation, severely hindering the convergence speed as well as the system performance. In this paper, we propose a boundary and context aware training approach for CIF based NAR models. Firstly, the connectionist temporal classification (CTC) spike information is utilized to guide the learning of acoustic boundaries in the CIF. Besides, an additional contextual decoder is introduced behind the CIF decoder, aiming to capture the linguistic dependencies within a sentence. Finally, we adopt a recently proposed Conformer architecture to improve the capacity of acoustic modeling. Experiments on the open-source Mandarin AISHELL-1 corpus show that the proposed method achieves a comparable character error rates (CERs) of 4.9% with only 1/24 latency compared with a state-of-the-art autoregressive (AR) Conformer model. Futhermore, when evaluating on an internal 7500 hours Mandarin corpus, our model still outperforms other NAR methods and even reaches the AR Conformer model on a challenging real-world noisy test set.

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。

↑