arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1902.01951eess.AScs.CLcs.LGcs.SD
收藏

使用多任务学习来提高声学到词和常规混合模型的性能

Using multi-task learning to improve the performance of acoustic-to-word and conventional hybrid models

Thai-Son Nguyen, Sebastian Stueker, Alex Waibel

摘要:

声学到单词(A2W)模型允许从声学信号到单词序列的直接映射,由于其简单性而成为端到端自动语音识别的有吸引力的方法。然而,先前的工作已经表明,A2W建模通常会遇到数据稀疏性问题,这会阻止直接训练这样的模型。到目前为止,预训练初始化是唯一提出来处理这个问题的方法。在这项工作中,我们建议建立一个共享的神经网络,并以多任务的方式优化A2W和传统的混合模型。我们的研究结果表明,在没有预训练初始化的情况下,使用我们的多任务模型训练A2W模型更加稳定,并且与基线模型相比有显著的改进。实验还表明,当与序列级优化标准(例如声学到字)联合训练时,混合声学模型的性能可以进一步提高。

英文摘要:

Acoustic-to-word (A2W) models that allow direct mapping from acoustic signals to word sequences are an appealing approach to end-to-end automatic speech recognition due to their simplicity. However, prior works have shown that modelling A2W typically encounters issues of data sparsity that prevent training such a model directly. So far, pre-training initialization is the only approach proposed to deal with this issue. In this work, we propose to build a shared neural network and optimize A2W and conventional hybrid models in a multi-task manner. Our results show that training an A2W model is much more stable with our multi-task model without pre-training initialization, and results in a significant improvement compared to a baseline model. Experiments also reveal that the performance of a hybrid acoustic model can be further improved when jointly training with a sequence-level optimization criterion such as acoustic-to-word.

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。

↑