arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2111.00844cs.CLcs.MM
收藏

探索非自回归端到端神经建模用于英语发音错误检测和诊断

Exploring Non-Autoregressive End-To-End Neural Modeling For English Mispronunciation Detection And Diagnosis

National Taiwan Normal University(台湾师范大学);Chunghwa Telecom Laboratories(中华电信研究院);EZAI

Hsin-Wei Wang, Bi-Cheng Yan, Hsuan-Sheng Chiu, Yung-Chang Hsu, Berlin Chen

摘要:

端到端(E2 E)神经建模已经成为开发计算机辅助语言训练(CAPT)系统的一个主要思想流派,与传统的基于发音评分的方法相比表现出了竞争力。然而,目前的E2 E CAPT神经方法面临着至少两个关键的挑战。一方面,大多数E2 E方法以自回归的方式操作,具有从左到右的束搜索来指示L2学习者的发音。然而,这导致非常慢的推理速度,这不可避免地阻碍了它们的实际使用。另一方面,E2 E神经方法通常是数据贪婪的,同时非本地训练数据的量不足常常会降低它们在错误发音检测和诊断(MD&D)上的功效。作为回应,我们提出了一种新的MD&D方法,该方法利用非自回归(NAR)E2 E神经建模来显着加快推理时间,同时保持与传统E2 E神经方法一致的性能。此外,我们设计和开发了一个发音建模网络,堆叠在我们的方法的NAR E2 E模型之上,以进一步提高MD&D的有效性。在L2-ARCTIC英语数据集上进行的实证实验似乎验证了我们方法的可行性,与一些顶级的E2 E模型和基于DNN-HMM声学模型的基于图标发音评分的方法相比。

英文摘要:

End-to-end (E2E) neural modeling has emerged as one predominant school of thought to develop computer-assisted language training (CAPT) systems, showing competitive performance to conventional pronunciation-scoring based methods. However, current E2E neural methods for CAPT are faced with at least two pivotal challenges. On one hand, most of the E2E methods operate in an autoregressive manner with left-to-right beam search to dictate the pronunciations of an L2 learners. This however leads to very slow inference speed, which inevitably hinders their practical use. On the other hand, E2E neural methods are normally data greedy and meanwhile an insufficient amount of nonnative training data would often reduce their efficacy on mispronunciation detection and diagnosis (MD&D). In response, we put forward a novel MD&D method that leverages non-autoregressive (NAR) E2E neural modeling to dramatically speed up the inference time while maintaining performance in line with the conventional E2E neural methods. In addition, we design and develop a pronunciation modeling network stacked on top of the NAR E2E models of our method to further boost the effectiveness of MD&D. Empirical experiments conducted on the L2-ARCTIC English dataset seems to validate the feasibility of our method, in comparison to some top-of-the-line E2E models and an iconic pronunciation-scoring based method built on a DNN-HMM acoustic model.

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。

↑