Speech Replay Detection with x-Vector Attack Embeddings and Spectral Features

被引:8
|
作者
Williams, Jennifer [1 ]
Rownicka, Joanna [1 ]
机构
[1] Univ Edinburgh, Ctr Speech Technol Res, Edinburgh, Midlothian, Scotland
来源
基金
英国工程与自然科学研究理事会;
关键词
automatic speaker verification; spoofing countermeasures; speech replay detection; NOISE;
D O I
10.21437/Interspeech.2019-1760
中图分类号
R36 [病理学]; R76 [耳鼻咽喉科学];
学科分类号
100104 ; 100213 ;
摘要
We present our system submission to the ASVspoof 2019 Challenge Physical Access (PA) task. The objective for this challenge was to develop a countermeasure that identifies speech audio as either bona fide or intercepted and replayed. The target prediction was a value indicating that a speech segment was bona fide (positive values) or "spoofed" (negative values). Our system used convolutional neural networks (CNNs) and a representation of the speech audio that combined x-vector attack embeddings with signal processing features. The x-vector attack embeddings were created from mel-frequency cepstral coefficients (MFCCs) using a time-delay neural network (TDNN). These embeddings jointly modeled 27 different environments and 9 types of attacks from the labeled data. We also used sub-band spectral centroid magnitude coefficients (SCMCs) as features. We included an additive Gaussian noise layer during training as a way to augment the data to make our system more robust to previously unseen attack examples. We report system performance using the tandem detection cost function (tDCF) and equal error rate (EER). Our approach performed better that both of the challenge baselines. Our technique suggests that our x-vector attack embeddings can help regularize the CNN predictions even when environments or attacks are more challenging.
引用
下载
收藏
页码:1053 / 1057
页数:5
相关论文
共 50 条
  • [1] Teager Energy Operator Based Features with x-vector for Replay Attack Detection
    Zhang, Zhenchuan
    Zhou, Liming
    Yang, Yingchun
    Wu, Zhaohui
    BIOMETRIC RECOGNITION (CCBR 2019), 2019, 11818 : 466 - 473
  • [2] On the Use of Ensemble X-Vector Embeddings for Improved Sleepiness Detection
    Egas-Lopez, Jose Vicente
    Busa-Fekete, Robert
    Gosztolya, Gabor
    SPEECH AND COMPUTER, SPECOM 2022, 2022, 13721 : 178 - 187
  • [3] Improving replay attack detection by combination of spatial and spectral features
    Ryoya, Yaguchi
    Shiota, Sayaka
    Ono, Nobutaka
    Kiya, Hitoshi
    2019 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA ASC), 2019, : 833 - 837
  • [4] A Comparison of Features for Replay Attack Detection
    Xie, Zhifeng
    Zhang, Weibin
    Chen, Zhuxin
    Xu, Xiangmin
    2019 3RD INTERNATIONAL CONFERENCE ON MACHINE VISION AND INFORMATION TECHNOLOGY (CMVIT 2019), 2019, 1229
  • [5] A Robust Method for Speech Replay Attack Detection
    Lin, Lang
    Wang, Rangding
    Yan, Diqun
    Dong, Li
    KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS, 2020, 14 (01): : 168 - 182
  • [6] VOICE QUALITY FEATURES FOR REPLAY ATTACK DETECTION
    Woubie, Abraham
    Backstrom, Tom
    2022 30TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO 2022), 2022, : 384 - 388
  • [7] Features and Classifiers for Replay Spoofing Attack Detection
    Hanilci, Cemal
    2017 10TH INTERNATIONAL CONFERENCE ON ELECTRICAL AND ELECTRONICS ENGINEERING (ELECO), 2017, : 1187 - 1191
  • [8] Exploration of Compressed ILPR Features for Replay Attack Detection
    Jelil, Sarfaraz
    Kalita, Sishir
    Prasanna, S. R. Mahadeva
    Sinha, Rohit
    19TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2018), VOLS 1-6: SPEECH RESEARCH FOR EMERGING MARKETS IN MULTILINGUAL SOCIETIES, 2018, : 631 - 635
  • [9] X-vector anonymization using autoencoders and adversarial training for preserving speech privacy
    Perero-Codosero, Juan M.
    Espinoza-Cuadros, Fernando M.
    Hernandez-Gomez, Luis A.
    COMPUTER SPEECH AND LANGUAGE, 2022, 74
  • [10] New Acoustic Features for Synthetic and Replay Spoofing Attack Detection
    Wei, Linqiang
    Long, Yanhua
    Wei, Haoran
    Li, Yijie
    SYMMETRY-BASEL, 2022, 14 (02):