期刊文献+

环境噪声对语言识别率的影响 被引量:2

Influence of environmental noise on speech discrimination score
在线阅读 下载PDF
导出
摘要 本文比较了4种干扰声源声级(dBA)和语言干扰级(SIL)的关系,观察了不同dBA、SIL和信噪比(S/N)对语言识别率影响的规律。结果表明dBA与SIL有很好的线性关系,但不同性质的噪声源,两者间的差值相差很大。4种干扰声源下的语言识别率都有随dBA和SIL的增加而减少,随SdBA/NdBA和SdBA/NSIL的增加而提高的趋势,但语言识别率与dBA、SdBA/NdBA的相关性因噪声的性质不同有很大差异,而与SIL、SdBA/NSIL的相关性受噪声性质的影响很小。 Comparisons were made of four kinds of dBA and SIL and a study was undertaken to observe the influence of various dBA, SIL and S/N on speech discrimination score. The results showed that under the four kinds of noise source the speech discrimination score was decreased with increasing of dBA or SIL, and was raised with increasing of SdBA/NdBA or SdBA/NSIL. The correlation between speech discrimination score and dBA or SdBA/NdBA was varied due to different noise sources, but the noise characteristics showed little influence on correlation between speech discrimination score and SIL or SdBA/NSIL.
机构地区 海军医学研究所
出处 《声学技术》 CSCD 1999年第3期123-127,共5页 Technical Acoustics
关键词 声级 语言干扰级 信噪比 语言识别率 sound level speech interference level(SIL) signal to noise ratio(S/N) speech discrimination score
  • 相关文献

参考文献3

  • 1顾瑞.耳塞、噪声和语言识别率.全国耳科学术会议论文汇编[M].,1983.49.
  • 2何春田.噪声和带宽对语言感受的影响.第二届全国声学学术会议论文摘要[M].,1979.151.
  • 3王四德.机车司机室噪声对汉语清晰度影响的研究.第三届全国声学学术会议论文摘要[M].,1982.297.

同被引文献27

  • 1纪元法,欧阳缮,曾庆宁,施浒立.基于波束形成的谱相减语音增强[J].数据采集与处理,2006,21(1):80-85. 被引量:3
  • 2BAI, Rong Junmei, XU Zhengbo, ZHANG Shuwu. Robust speaker recognition intergrating pitch and wiener filter[A]. IEEE ISCSLP[C]. 2004, 7: 69-73.
  • 3Jia Ching Wang, Chung Hsien Yang, Jhing Fa Wang and Hsiao Ping Lee. Robust speaker identification and verification[J]. IEEE COMPUTATIONAL INTELLIGNCE, 2007, 5(2): 52-59.
  • 4Aliaa A. Youssif, Ebada A. Sarban. Development of automatic speaker identification system[A]. 21^st National Radio Science Conference[C]. 2004, C7: 1-8.
  • 5Naik D. Pole-filtered cepstral mean substraction[J]. IEEE ECASSP, 1995, 5(1): 157-160.
  • 6Reynolds D A. Channel robust speaker verification via feature mapping[A]. Proc. ICASSP[C]. 2003, 4( Ⅱ ): 53-56.
  • 7Stemmer G, Brugnara E Integration of Heteroscedastic Linear Discriminant Analysis (HLDA) into Adaptive Training[J]. !EEE ICASSP, 2006, 5(I): 1-4.
  • 8Reynolds D A. The super SID project: Exploiting high-level information for high-accuracy speaker recognition[A]. In Proc. IEEE ICASSP[C]. Hong Kong, 2003.
  • 9Reynolds D A. Speaker verification using adapted Gaussian mixture models[J]. Digital Signal Processing, 2000, (10): 19-41.
  • 10Barras C Gauvain. Feature and score normalization for speaker verification of cellular data[J]. IEEE ICASSP, 2003, 4(2): 49-52.

引证文献2

二级引证文献4

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部