Speech Recognition for Air Traffic Control Utilizing a Multi-Head State-Space Model and Transfer Learning

IF 2.1 3区 工程技术 Q2 ENGINEERING, AEROSPACE
Haijun Liang, Hanwen Chang, Jianguo Kong
{"title":"Speech Recognition for Air Traffic Control Utilizing a Multi-Head State-Space Model and Transfer Learning","authors":"Haijun Liang, Hanwen Chang, Jianguo Kong","doi":"10.3390/aerospace11050390","DOIUrl":null,"url":null,"abstract":"In the present study, a novel end-to-end automatic speech recognition (ASR) framework, namely, ResNeXt-Mssm-CTC, has been developed for air traffic control (ATC) systems. This framework is built upon the Multi-Head State-Space Model (Mssm) and incorporates transfer learning techniques. Residual Networks with Cardinality (ResNeXt) employ multi-layered convolutions with residual connections to augment the extraction of intricate feature representations from speech signals. The Mssm is endowed with specialized gating mechanisms, which incorporate parallel heads that acquire knowledge of both local and global temporal dynamics in sequence data. Connectionist temporal classification (CTC) is utilized in the context of sequence labeling, eliminating the requirement for forced alignment and accommodating labels of varying lengths. Moreover, the utilization of transfer learning has been shown to improve performance on the target task by leveraging knowledge acquired from a source task. The experimental results indicate that the model proposed in this study exhibits superior performance compared to other baseline models. Specifically, when pretrained on the Aishell corpus, the model achieves a minimum character error rate (CER) of 7.2% and 8.3%. Furthermore, when applied to the ATC corpus, the CER is reduced to 5.5% and 6.7%.","PeriodicalId":48525,"journal":{"name":"Aerospace","volume":null,"pages":null},"PeriodicalIF":2.1000,"publicationDate":"2024-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Aerospace","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.3390/aerospace11050390","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, AEROSPACE","Score":null,"Total":0}
引用次数: 0

Abstract

In the present study, a novel end-to-end automatic speech recognition (ASR) framework, namely, ResNeXt-Mssm-CTC, has been developed for air traffic control (ATC) systems. This framework is built upon the Multi-Head State-Space Model (Mssm) and incorporates transfer learning techniques. Residual Networks with Cardinality (ResNeXt) employ multi-layered convolutions with residual connections to augment the extraction of intricate feature representations from speech signals. The Mssm is endowed with specialized gating mechanisms, which incorporate parallel heads that acquire knowledge of both local and global temporal dynamics in sequence data. Connectionist temporal classification (CTC) is utilized in the context of sequence labeling, eliminating the requirement for forced alignment and accommodating labels of varying lengths. Moreover, the utilization of transfer learning has been shown to improve performance on the target task by leveraging knowledge acquired from a source task. The experimental results indicate that the model proposed in this study exhibits superior performance compared to other baseline models. Specifically, when pretrained on the Aishell corpus, the model achieves a minimum character error rate (CER) of 7.2% and 8.3%. Furthermore, when applied to the ATC corpus, the CER is reduced to 5.5% and 6.7%.
利用多头状态空间模型和迁移学习进行空中交通管制语音识别
本研究为空中交通管制(ATC)系统开发了一种新型端到端自动语音识别(ASR)框架,即 ResNeXt-Mssm-CTC。该框架以多头状态空间模型(Mssm)为基础,并结合了迁移学习技术。残差网络(ResNeXt)采用多层卷积与残差连接,以增强从语音信号中提取复杂特征表征的能力。Mssm 具有专门的门控机制,它结合了并行头,可获取序列数据中局部和全局时间动态的知识。在序列标注中使用了连接主义时序分类(CTC),从而消除了强制对齐的要求,并适应不同长度的标注。此外,迁移学习的使用已被证明可以利用从源任务中获得的知识来提高目标任务的性能。实验结果表明,与其他基线模型相比,本研究提出的模型表现出更优越的性能。具体来说,在 Aishell 语料库上进行预训练时,该模型的最小字符错误率 (CER) 为 7.2% 和 8.3%。此外,当应用于 ATC 语料库时,CER 降至 5.5% 和 6.7%。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Aerospace
Aerospace ENGINEERING, AEROSPACE-
CiteScore
3.40
自引率
23.10%
发文量
661
审稿时长
6 weeks
期刊介绍: Aerospace is a multidisciplinary science inviting submissions on, but not limited to, the following subject areas: aerodynamics computational fluid dynamics fluid-structure interaction flight mechanics plasmas research instrumentation test facilities environment material science structural analysis thermophysics and heat transfer thermal-structure interaction aeroacoustics optics electromagnetism and radar propulsion power generation and conversion fuels and propellants combustion multidisciplinary design optimization software engineering data analysis signal and image processing artificial intelligence aerospace vehicles'' operation, control and maintenance risk and reliability human factors human-automation interaction airline operations and management air traffic management airport design meteorology space exploration multi-physics interaction.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信