期刊文献+

基于改进PNCC特征和两步区分性训练的录音设备识别方法 被引量:9

A Recording Device Identification Algorithm Based on Improved PNCC Feature and Two-Step Discriminative Training
在线阅读 下载PDF
导出
摘要 录音设备来源识别是通过分析已获取的数字语音信号从而确定其录制设备的一种技术,属于数字音频盲取证.本文提出了一种基于改进PNCC特征和两步区分性训练的录音设备识别方法,由于音频中的静音包含了完整的设备信息,且不受说话人和文本等因素的影响,因此从静音段提取改进的PNCC特征,利用了PNCC的长时帧分析去除背景噪声对设备信息的影响.在模型方面,以GMM-UBM为基准模型,并通过两步区分性训练调整集内设备模型和通用背景模型,提升模型区分能力.该方法对于30种设备闭集识别的平均正确识别率为90.23%;对于15个集内和15个集外设备的测试,等错误率为15.17%,集内平均正确识别率为96.65%,验证了本文算法的有效性. Recording device identification is a kind of blind digital audio forensic technique, which extracts digital evidence of device mechanism involved in the generation of the speech recording by analyzing the acoustic signal. This paper proposes a recording device identification algorithm which is based on improved PNCC feature and two-step discriminative training.Due to the fact that silence periods contain the device information and is not affected by speaker and texture factors, this paper extracts improved PNCC from silence periods, which uses long term analysis to remove the effect of background noise. GMM-UBM is set as the baseline system, which is improved by two steps discriminative training. The experimental result indicates that the average accuracy of recording device identification on 30 devices is 90.23% ;for 15 inset and 15 outset devices testing,the EER is 15.17% and ACC is 96.65 %, which proves the effectiveness of the proposed algorithm.
出处 《电子学报》 EI CAS CSCD 北大核心 2014年第1期191-198,共8页 Acta Electronica Sinica
基金 国家自然科学基金(No.60972132 No.61101160) 广东省自然科学基金(No.9351064101000003 No.10451064101004651)
关键词 数字音频取证 录音设备识别 GMM-UBM 区分性训练 PNCC digital audio forensics recording device identification GMM-UBM discfiminative training PNCC
作者简介 贺前华男.1965年2月生,湖南邵东人.现任华南理工大学电子与信息学院教授、博士研究生导师、副院长,广东省“千百十”人才工程培养对象.研究领域主要有多媒体信息检索技术、数字音频侦测技术、信息安全身份认证技术、音视频双模态语音识别技术和嵌入式系统设计与应用等. 王志锋(通讯作者)男.1985年2月生,湖北武汉人.华南理工大学信号与信息处理专业博士,美国卡内基梅隆大学计算机学院联合培养博士.研究方向为语音信号处理、数字取证技术以及生物特征识别.已发表相关论文近10篇.E—mail:eezfwang@gmail.com
  • 相关文献

参考文献6

二级参考文献36

  • 1张利鹏,曹犟,徐明星,郑方.防止假冒者闯入说话人识别系统[J].清华大学学报(自然科学版),2008,48(S1):699-703. 被引量:13
  • 2Weber F,Peskin B,et al.Speaker recognition on single and multi-speaker data[J].Digital Signal Processing,2000,10(1):75-92.
  • 3Nakagawa S,Zhang W,Takahashi M.Text independent speaker recognition by combining speaker-specific GMM with speaker adapted syllable-based HNM[A].Proceedings of the International Conference on Acoustic,Speech and Signal Processing (ICASSP)[C].Montreal,Canada:Institute of Electrical and Electronics Engineers Inc,2004.81-84.
  • 4McGurk H,MacDonald J.Hearing lips and seeing voices[J].Nature,1976,264(5588):746-748.
  • 5Brady K,Brandstein M.An evaluation of audio-visual person recognition on the XM2VTS corpus using the Lausanne protocols[A].Proceedings of the International Conference on Acoustic,Speech and Signal Processing(ICASSP)[C].Hawaii,USA:Institute of Electrical and Electronics Engineers Inc,2007.237-240.
  • 6Chen T.Audio-visual speech processing[J].IEEE Transactions on Signal Processing,2001,18(1):9-21.
  • 7Chu S M,Huang T S.Multi-model sensory fusion with application.to audio-visual speech recognition[A].Proceedings of European Conference on Speech Communication and Technology (Eurospeech)[C].Aalborg,Denmark:European Association for Signal Processing,2001.109-112.
  • 8Gowdy J N,Subramanya A,Bartels C,Bihnes J.DBN based multi-stream models for audio-visual speech recognition[A].Proceedings of the International Conference on Acoustic,Speech and Signal Processing(ICASSP)[C].Montreal,Canada:Institute of Electrical and Electronics Engineers Inc,2004.993-996.
  • 9Brownian C P,Goldstein L.Articulatory phonology:An overview[J].Phonetica,1992,49:155-180.
  • 10Bilmes J,Zweig G.The graphical models toolkit:An open source software system for speech and time-series processing[A].Proceedings of the International Conference on Acoustic,Speech and Signal Processing(ICASSP)[C].Florida,USA:Institute of Electrical and Electronics Engineers Inc,2002.3916-3919.

共引文献66

同被引文献36

引证文献9

二级引证文献34

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