Knowledge Commons of Institute of Automation,CAS
Improving End-to-End Contextual Speech Recognition with Fine-Grained Contextual Knowledge Selection | |
Minglun Han1,2,3; Linhao Dong3; Zhenlin Liang3; Meng Cai3; Shiyu Zhou1; Zejun Ma3; Bo Xu1,2 | |
2022 | |
会议名称 | ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) |
会议日期 | 2022.05 |
会议地点 | Singapore, Singapore |
摘要 | Nowadays, most methods for end-to-end contextual speech recognition bias the recognition process towards contextual knowledge. Since all-neural contextual biasing methods rely on phrase-level contextual modeling and attention-based relevance modeling, they may suffer from the confusion between similar context-specific phrases, which hurts predictions at the token level. In this work, we focus on mitigating confusion problems with fine-grained contextual knowledge selection (FineCoS). In FineCoS, we introduce fine-grained knowledge to reduce the uncertainty of token predictions. Specifically, we first apply phrase selection to narrow the range of phrase candidates, and then conduct token attention on the tokens in the selected phrase candidates. Moreover, we re-normalize the attention weights of most relevant phrases in inference to obtain more focused phrase-level contextual representations, and inject position information to help model better discriminate phrases or tokens. On LibriSpeech and an in-house 160,000-hour dataset, we explore the proposed methods based on an all-neural biasing method, collaborative decoding (ColDec). The proposed methods further bring at most 6.1% relative word error rate reduction on LibriSpeech and 16.4% relative character error rate reduction on the in-house dataset. |
关键词 | Automatic Speech Recognition Context Biasing Speech Recognition Customization Continuous Integrate-and-Fire Mechanism |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 是 |
七大方向——子方向分类 | 语音识别与合成 |
国重实验室规划方向分类 | 语音语言处理 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/51693 |
专题 | 复杂系统认知与决策实验室_听觉模型与认知计算 |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences 2.School of Artificial Intelligence, University of Chinese Academy of Sciences 3.ByteDance AI Lab |
第一作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Minglun Han,Linhao Dong,Zhenlin Liang,et al. Improving End-to-End Contextual Speech Recognition with Fine-Grained Contextual Knowledge Selection[C],2022. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
IMPROVING_END_TO_END(463KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论