Proceedings of the 3rd International Conference on New Media Development and Modernized Education, NMDME 2023, October 13–15, 2023, Xi’an, China

Research Article

Design of Multilevel Speech Automatic Recognition and Translation Software Based on Internet

Download26 downloads
  • @INPROCEEDINGS{10.4108/eai.13-10-2023.2341325,
        author={Xin  Xiong and Qing  Xu and Mingjing  Guo},
        title={Design of Multilevel Speech Automatic Recognition and Translation Software Based on Internet},
        proceedings={Proceedings of the 3rd International Conference on New Media Development and Modernized Education, NMDME 2023, October 13--15, 2023, Xi’an, China},
        publisher={EAI},
        proceedings_a={NMDME},
        year={2024},
        month={1},
        keywords={internet; english; mfcc; translator; speech recognition system},
        doi={10.4108/eai.13-10-2023.2341325}
    }
    
  • Xin Xiong
    Qing Xu
    Mingjing Guo
    Year: 2024
    Design of Multilevel Speech Automatic Recognition and Translation Software Based on Internet
    NMDME
    EAI
    DOI: 10.4108/eai.13-10-2023.2341325
Xin Xiong1, Qing Xu1,*, Mingjing Guo2
  • 1: Naval University of Engineer
  • 2: East China University of Technology
*Contact email: xuq235@126.com

Abstract

Speech recognition is an important research field that involves converting spoken language into textual form for computers to understand and process. In recent years, with the continuous development of embedded technology, embedded speech recognition technology has become the main research direction in the field of speech recognition. In this context, this article designs and implements an English translator speech recognition system. Our experimental results indicate that our designed translator performs well in terms of average speech recognition accuracy, reaching approximately 91.24%. In contrast, the average speech recognition accuracy of the traditional method 1 translation system is about 76.73%, while the average speech recognition accuracy of the traditional method 2 translation system is about 65.34%. These results indicate that the translator designed in this article performs better in speech recognition and has stronger speech recognition capabilities. This is crucial for the entire translation process, as more accurate speech recognition is expected to lead to higher quality translations, thereby improving translation effectiveness.