arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2204.09883cs.SDeess.AS
收藏

端到端多口音语音识别的分层快速自适应算法

Layer-wise Fast Adaptation for End-to-End Multi-Accent Speech Recognition

Shanghai Jiao Tong University(上海交通大学)

Xun Gong, Yizhou Lu, Zhikai Zhou, Yanmin Qian

摘要:

口音变异对自动语音识别建模提出了巨大的挑战。虽然基于独热口音向量的自适应系统是常用的,但它们需要关于目标口音的先验知识,并且不能处理看不见的口音。此外,简单地连接重音嵌入并不能很好地利用重音知识,其改进有限。在这项工作中,我们的目标是解决这些问题,注入到E2 E ASR模型编码器的一种新的逐层自适应结构。适配器层在口音空间中编码任意口音,并帮助ASR模型识别口音语音。对于给定的语音,自适应结构提取相应的口音信息,并通过所有口音基的线性组合将输入的声学特征转换为口音相关特征。我们进一步探索适应层的注入位置、重音基础的数量以及不同类型的重音基础,以实现更好的重音适应。实验结果表明,在AESRC 2020口音数据集和Librispeech数据集上,与基线相比,该自适应结构分别使相对单词出错率降低了12%和10%.

英文摘要:

Accent variability has posed a huge challenge to automatic speech recognition~(ASR) modeling. Although one-hot accent vector based adaptation systems are commonly used, they require prior knowledge about the target accent and cannot handle unseen accents. Furthermore, simply concatenating accent embeddings does not make good use of accent knowledge, which has limited improvements. In this work, we aim to tackle these problems with a novel layer-wise adaptation structure injected into the E2E ASR model encoder. The adapter layer encodes an arbitrary accent in the accent space and assists the ASR model in recognizing accented speech. Given an utterance, the adaptation structure extracts the corresponding accent information and transforms the input acoustic feature into an accent-related feature through the linear combination of all accent bases. We further explore the injection position of the adaptation layer, the number of accent bases, and different types of accent bases to achieve better accent adaptation. Experimental results show that the proposed adaptation structure brings 12\% and 10\% relative word error rate~(WER) reduction on the AESRC2020 accent dataset and the Librispeech dataset, respectively, compared to the baseline.

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。

↑