arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.09127eess.AS
收藏

面向德语端到端语音识别的大规模语料CTC切分

CTC-Segmentation of Large Corpora for German End-to-end Speech Recognition

Ludwig Kürzinger, Dominik Winkelbauer, Lujun Li, Tobias Watzel, Gerhard Rigoll

摘要:

最近的端到端自动语音识别(ASR)系统展示了优于传统混合DNN/ HMM ASR的能力。除了这些系统的结构改进外,这些模型在深度、参数和模型容量方面都有所增加。然而,这些模型也需要更多的训练数据来实现可比的性能。在这项工作中,我们结合了德国语音识别,包括尚未标记的语音数据,超过$1700$ h的语音数据的大数据集免费提供的语料库。对于数据准备,我们提出了一种两阶段的方法,该方法使用用连接主义时间分类(CTC)预训练的ASR模型,从未分割或未标记的训练数据中引导更多的训练数据。然后,从CTC训练的网络获得的标签概率中提取话语,以确定片段对齐。利用这些训练数据,我们训练了一个混合CTC/attention Transformer模型,该模型在Tuda-DE测试集上实现了$12.8\%$ WER,超过了传统混合DNN/HMM ASR的$14.4\%$的先前基线。

英文摘要:

Recent end-to-end Automatic Speech Recognition (ASR) systems demonstrated the ability to outperform conventional hybrid DNN/ HMM ASR. Aside from architectural improvements in those systems, those models grew in terms of depth, parameters and model capacity. However, these models also require more training data to achieve comparable performance. In this work, we combine freely available corpora for German speech recognition, including yet unlabeled speech data, to a big dataset of over $1700$h of speech data. For data preparation, we propose a two-stage approach that uses an ASR model pre-trained with Connectionist Temporal Classification (CTC) to boot-strap more training data from unsegmented or unlabeled training data. Utterances are then extracted from label probabilities obtained from the network trained with CTC to determine segment alignments. With this training data, we trained a hybrid CTC/attention Transformer model that achieves $12.8\%$ WER on the Tuda-DE test set, surpassing the previous baseline of $14.4\%$ of conventional hybrid DNN/HMM ASR.

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。

↑