Spoofing countermeasures for the protection of automatic speaker recognition systems against attacks with artificial signals
Open accessThe DOI matches its Crossref record; 13 of the 14 references it cites were found in a registry or an index.
Abstract
The vulnerability of automatic speaker recognition systems to imposture or spoofing is widely acknowledged. This paper shows that extremely high false alarm rates can be provoked by simple spoofing attacks with artificial, non-speech-like signals and highlights the need for spoofing countermeasures. We show that two new, but trivial countermeasures based on higher-level, dynamic features and voice quality assessment offer varying degrees of protection and that further work is needed to develop more robust spoofing countermeasure mechanisms. Finally, we show that certain classifiers are inherently more robust to such attacks than others which strengthens the case for fused-system approaches to automatic speaker recognition.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The text we found for this paper looks like another document, so the self-check wasn’t run.
Citations
Cites
P.563—The ITU-T Standard for Single-Ended Speech Quality Assessment
Malfait, L., Berger, J., & Kastner, M. (2006) · IEEE Transactions on Audio, Speech and Language Processing · doi:10.1109/tasl.2006.883177 · Cited by 214
Kinnunen, T., Wu, Z.-Z., Lee, K. A., Sedlak, F., Chng, E. S., & Li, H. (2012) · 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2012.6288895 · Cited by 184
Vulnerability in speaker verification - a study of technical impostor techniques
Lindberg, J., & Blomberg, M. (1999) · European Conference on Speech Communication and Technology (Eurospeech) · doi:10.21437/eurospeech.1999-283 · Cited by 127
A straightforward and efficient implementation of the factor analysis model for speaker verification
Matrouf, D., Scheffer, N., Fauve, B., & Bonastre, J.-F. (2007) · Interspeech (USB) · doi:10.21437/interspeech.2007-156 · Cited by 124
On the security of HMM-based speaker verification systems against imposture using synthetic speech
Masuko, T., Hitotsumatsu, T., Tokuda, K., & Kobayashi, T. (1999) · European Conference on Speech Communication and Technology (Eurospeech) · doi:10.21437/eurospeech.1999-286 · Cited by 83
State-of-the-Art Performance in Text-Independent Speaker Verification Through Open-Source Software
Fauve, B., Matrouf, D., Scheffer, N., Bonastre, J.-F., & Mason, J. S. D. (2007) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/tasl.2007.902877 · Cited by 82
Artificial impostor voice transformation effects on false acceptance rates
Bonastre, J.-F., Matrouf, D., & Fredouille, C. (2007) · Interspeech (USB) · doi:10.21437/interspeech.2007-173 · Cited by 76
Voice Forgery Using ALISP: Indexation in a Client Memory
Perrot, P., Aversano, G., Blouet, R., Charbit, M., & Chollet, G. (2006) · Proceedings. (ICASSP '05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005. · doi:10.1109/icassp.2005.1415039 · Cited by 60
On The Vulnerability Of Automatic Speaker Recognition To Spoofing Attacks With Artificial Signals
Alegre, F., Evans, N., Favre, B., & Vipperla, R. (2012) · Graduate School and Research Center in Digital Science (EURECOM) · doi:10.5281/zenodo.43138 · Cited by 57
Self-citationRevisiting the security of speaker verification systems against imposture using synthetic speech
De Leon, P. L., Apsingekar, V. R., Pucher, M., & Yamagishi, J. (2010) · 2010 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2010.5495413 · Cited by 37
Cited by
Spoofing and countermeasures for speaker verification: A survey
Wu, Z., Evans, N., Kinnunen, T., Yamagishi, J., Alegre, F., & Li, H. (2015) · Speech Communication · doi:10.1016/j.specom.2014.10.005 · Cited by 601
Self-citationASVspoof 2015: the first automatic speaker verification spoofing and countermeasures challenge
Wu, Z., Kinnunen, T., Evans, N., Yamagishi, J., Hanilçi, C., Sahidullah, M., & Sizov, A. (2015) · Interspeech 2015 · doi:10.21437/interspeech.2015-462 · Cited by 484
“Countermeasures have been reported for replay attacks [2, 3, 4, 5], speech synthesis [6, 7, 8, 9], voice conversion [10, 11, 12, 13] and non-speech, artificial signals [14].”
BackgroundSelf-citationSpoofing and countermeasures for automatic speaker verification
Evans, N., Kinnunen, T., & Yamagishi, J. (2013) · Interspeech 2013 · doi:10.21437/interspeech.2013-288 · Cited by 160
“Two approaches to artificial signal detection are reported in [42].”
Self-citationAn Investigation of Deep-Learning Frameworks for Speaker Verification Antispoofing
Zhang, C., Yu, C., & Hansen, J. H. L. (2017) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2016.2647199 · Cited by 122
“Previous studies have shown that the false acceptance rate of state-of-the-art speaker verification systems has been significantly increased with replayed speech, impersonation, synthesized speech, voice conversion and artificial signals as spoofing attacks [2], [6], [11]–[17].”
BackgroundAlegre, F., Amehraye, A., & Evans, N. (2013) · 2013 IEEE Sixth International Conference on Biometrics: Theory, Applications and Systems (BTAS) · doi:10.1109/btas.2013.6712706 · Cited by 107
Self-citationRe-assessing the threat of replay spoofing attacks against automatic speaker verification
Alegre, F., Janicki, A., & Evans, N. (2014) · Graduate School and Research Center in Digital Science (EURECOM) · Cited by 80
Self-citationSpoofing countermeasures to protect automatic speaker verification from voice conversion
Alegre, F., Amehraye, A., & Evans, N. (2013) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2013.6638222 · Cited by 78
Self-citationToward Robust Audio Spoofing Detection: A Detailed Comparison of Traditional and Learned Features
Balamurali, B. T., Lin, K. W. E., Lui, S., Chen, J., & Herremans, D. (2019) · IEEE Access · doi:10.1109/access.2019.2923806 · Cited by 83
Joint Speaker Verification and Antispoofing in the $i$ -Vector Space
Sizov, A., Khoury, E., Kinnunen, T., Wu, Z., & Marcel, S. (2015) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2015.2407362 · Cited by 74
“Personal use is permitted, but republication/redistribution requires IEEE permission.”
BackgroundSAS: A speaker verification spoofing database containing diverse attacks
Wu, Z., Khodabakhsh, A., Demiroğlu, C., Yamagishi, J., Saito, D., Toda, T., & King, S. (2015) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2015.7178810 · Cited by 83
“Recently, an increasing number of studies have assessed the vulnerability of ASV systems to various forms of spoofing attacks [2, 3, 4], including impersonation [5, 6], replay [7], speech synthesis [8, 9] and voice conversion [10, 11, 12, 13, 14].”
BackgroundIntroduction to Voice Presentation Attack Detection and Recent Advances
Sahidullah, M., Delgado, H., Todisco, M., Kinnunen, T., Evans, N., Yamagishi, J., & Lee, K. A. (2019) · Advances in computer vision and pattern recognition · doi:10.1007/978-3-319-92627-8_15 · Cited by 84
Self-citationDeep features for automatic spoofing detection
Qian, Y., Chen, N., & Yu, K. (2016) · Speech Communication · doi:10.1016/j.specom.2016.10.007 · Cited by 78
“Other work has focused on modeling, including supervector-based SVM (Alegre et al., 2012; Liu et al., 2015b), i-vector (Novoselov et al., 2015; Weng et al., 2015; Khoury et al., 2014), and have been reported to be robust against artificial signal attacks.”
BackgroundMethodInfluentialAudio Deepfake Detection: What Has Been Achieved and What Lies Ahead
Zhang, B., Cui, H., Nguyen, V., & Whitty, M. (2025) · Sensors · doi:10.3390/s25071989 · Cited by 72
Voice conversion and spoofing attack on speaker verification systems
Wu, Z., & Li, H. (2013) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipa.2013.6694344 · Cited by 63
“In [79], high level features, which is extracted over a long speech context, is employed to capture the change of speech”
BackgroundSynthetic speech detection using fundamental frequency variation and spectral features
Pal, M., Paul, D., & Saha, G. (2017) · Computer Speech & Language · doi:10.1016/j.csl.2017.10.001 · Cited by 59
“Countermeasure using high-level dynamic features and voice qual24 ity assessment was employed in Alegre et al. (2012).”
MethodHuang, L. F., & Pun, C. (2020) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2020.2998870 · Cited by 58
“Support vector machine (SVM) [23], [24], linear discriminant analysis (LDA) [25], [26], and deep neural networks (DNN) [27] are also used as a more efficient classifier.”
MethodDua, M., Jain, C., & Kumar, S. (2021) · Journal of Ambient Intelligence and Humanized Computing · doi:10.1007/s12652-021-02960-0 · Cited by 58
Speaker Recognition Anti-spoofing
Evans, N., Kinnunen, T., Yamagishi, J., Wu, Z., Alegre, F., & Leon, P. L. D. (2014) · Advances in computer vision and pattern recognition · doi:10.1007/978-1-4471-6524-8_7 · Cited by 51
“Two approaches to artificial signal detection are reported in [101].”
Self-citationVoice Spoofing Countermeasure for Logical Access Attacks Detection
Arif, T., Javed, A., Alhameed, M., Jeribi, F., & Tahir, A. (2021) · IEEE Access · doi:10.1109/access.2021.3133134 · Cited by 48
“In [31], authors investigated an utterance level feature termed as longer contexts or high level feature (HLF) and voice assessment tool (p563) which calculates Mean Opinion Score to detect artificial signals.”
BackgroundVoice conversion versus speaker verification: an overview
Wu, Z., & Li, H. (2014) · APSIPA Transactions on Signal and Information Processing · doi:10.1017/atsip.2014.17 · Cited by 45
“In [98], a long-term dynamic feature, which was extracted at the utterance-level, was proposed to capture the utterance-level speech variation for detecting converted speech.”
A new speaker verification spoofing countermeasure based on local binary patterns
Alegre, F., Vipperla, R., Amehraye, A., & Evans, N. (2013) · Interspeech (USB) · doi:10.21437/interspeech.2013-291 · Cited by 45
Self-citationDeep Feature Engineering for Noise Robust Spoofing Detection
Qian, Y., Chen, N., Dinkel, H., & Wu, Z. (2017) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2017.2732162 · Cited by 46
“[33] J. Villalba, A. Miguel, A. Ortega, and E. Lleida, “Spoofing detection with DNN and one-class SVM for the ASVspoof 2015 challenge,” in Proc.”
BackgroundOn the study of replay and voice conversion attacks to text-dependent speaker verification
Wu, Z., & Li, H. (2015) · Multimedia Tools and Applications · doi:10.1007/s11042-015-3080-9 · Cited by 48
“The past individual studies [3,33,4,15,2,18,48,20] have shown that voice conversion techniques are able to increase the error rates of state-of-the-art classifiers to unacceptable levels.”
BackgroundAn assessment of automatic speaker verification vulnerabilities to replay spoofing attacks
Janicki, A., Alegre, F., & Evans, N. (2016) · Security and Communication Networks · doi:10.1002/sec.1499 · Cited by 54
Self-citationSafeEar: Content Privacy-Preserving Audio Deepfake Detection
Li, X., Li, K., Zheng, Y., Yan, C., Ji, X., & Xu, W. (2024) · Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3658644.3670285 · Cited by 39
“…especially spectral features like MFCC and LFCC [63, 86], and deep learning features like Wav2Vec2 [91]; (2) classification algorithms, e.g. , SVM [3], GMM [12], CNN [63], GNN [36], and Transformer [53]; (3) generalization methods, e.g. , investigating novel loss functions [17, 108] and using…”
Introducing i-vectors for joint anti-spoofing and speaker verification
Khoury, E., Kinnunen, T., Сизов, А. С., Wu, Z., & Marcel, S. (2014) · Interspeech (USB) · doi:10.21437/interspeech.2014-13 · Cited by 44
“Interestingly, the quality of the converted voice does not have to be particularly high; even artificial signal attacks [13, 14] involving unintelligible speech can spoof a recognizer.”
BackgroundOn robustness of speech based biometric systems against voice conversion attack
Pal, M., & Saha, G. (2015) · Applied Soft Computing · doi:10.1016/j.asoc.2015.01.036 · Cited by 32
Wu, Z., Larcher, A., Lee, K. A., Chng, E. S., Kinnunen, T., & Li, H. (2013) · Interspeech (USB) · doi:10.21437/interspeech.2013-293 · Cited by 31
“Several methods have been employed to simulate the spoofing attack, including replay attacks [3, 4], human voice mimicking [5] and artificial signal spoofing [6].”
MethodNovel speech features for improved detection of spoofing attacks
Paul, D., Pal, M., & Saha, G. (2015) · IEEE India Conference · doi:10.1109/indicon.2015.7443805 · Cited by 24
“Other countermeasure techniques developed for this purpose were based on high level dynamic features and voice quality assessment [16], pairwise distance between input and target speaker data in training [17] and local binary pattern analysis [18], [19].”
BackgroundOn the use of deep recurrent neural networks for detecting audio spoofing attacks
Scardapane, S., Stoffl, L., Röhrbein, F., & Uncini, A. (2017) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2017.7966294 · Cited by 25
“Other, slightly more challenging groups of attacks are not considered in the challenge, including impersonation attacks (where a user tries to mimic the voice of the target speaker); replay attacks [29], consisting in the use of pre-recorded speech samples, either surreptitiously or by concatenating multiple speech fr…”
BackgroundWeng, S., Chen, S., Yu, L., Wu, X., Cai, W., Liu, Z., Zhou, Y., & Li, M. (2015) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipa.2015.7415492 · Cited by 17
“66 human+spoof[1,2,3,4] human+spoof[5] 4.”
BackgroundDemiroğlu, C., Büyük, O., Khodabakhsh, A., & Maia, R. (2017) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2017.2673807 · Cited by 17
“In [14], voice transformation using Gaussian Mixture Model (GMM) based parallel acoustic models [15]—one GMM for the space of normalized features that are used in ASV and a second GMM, whose Gaussian components match one-to-one with the first GMM, for computing the parameter transformation function—is found to be effe…”
MethodIntroduction to Voice Presentation Attack Detection and Recent Advances
Sahidullah, M., Delgado, H., Todisco, M., Nautsch, A., Wang, X., Kinnunen, T., Evans, N., Yamagishi, J., & Lee, K. A. (2023) · Advances in computer vision and pattern recognition · doi:10.1007/978-981-19-5288-3_13 · Cited by 22
Self-citationSpoofing voice verification systems with statistical speech synthesis using limited adaptation data
Khodabakhsh, A., Mohammadi, A., & Demiroğlu, C. (2016) · Computer Speech & Language · doi:10.1016/j.csl.2016.08.004 · Cited by 14
“In Alegre et al. (2012), Gaussian Mixture Model (GMM) based voice transformation using parallel data is found to be effective at spoofing the voice verification systems.”
MethodAnti-spoofing: Voice Conversion
Evans, N., Alegre, F., Wu, Z., & Kinnunen, T. (2014) · Springer eBooks · doi:10.1007/978-3-642-27733-7_9111-2 · Cited by 10
Self-citationTowards robust audio spoofing detection: a detailed comparison of traditional and learned featuresPreprint
Balamurali, B. T., Lin, K. W. E., Lui, S., Chen, J., & Herremans, D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.12439 · Cited by 4
Speech-Forensics: Towards Comprehensive Synthetic Speech Dataset Establishment and Analysis
Ji, Z., Lin, C., Wang, H., & Shen, C. (2024) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2024/46 · Cited by 11
“In early deceptive audio detection, machine learning techniques, such as SVM, were prevalent, with Alegre [ 2012 ] proposing an SVM-based robust voice spoofing detection strategy.”
MethodSecurity in Voice Authentication
Yang, C. (2014) · Digital WPI · Cited by 3
Spectral mapping for voice conversion
Wu, Z. Z. (2015) · DR-NTU (Nanyang Technological University) · Cited by 3
Özbay, M. C., Khodabakhsh, A., Mohammadi, A., & Demiroğlu, C. (2016) · European Signal Processing Conference (EUSIPCO) · doi:10.1109/eusipco.2016.7760440 · Cited by 2
“Two countermeasures are also proposed in [5].”
MethodDesign of an enhanced speech authentication system over mobile devices
Anwar, M. U. (2018) · Cited by 1
Application-Oriented Robustness Issues
Zheng, T. F., & Li, L. (2017) · Springer briefs in electrical and computer engineering · doi:10.1007/978-981-10-3238-7_4
Sun, K., & Wang, S. (2023) · SpringerBriefs in computer science · doi:10.1007/978-3-031-44748-8_1
Metadata
- Authors
- Alegre, F., Vipperla, R., & Evans, N.
- ORCID
- Evans, Nicholas
- Venue
- Interspeech 2012
- Publisher
- ISCA
- Published
- 2012-09-09
- Pages
- 1688-1691
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W2270140468
- Cited by
- 43
- References
- 14
- Citations by year
- 2025: 1 · 2024: 2 · 2023: 2 · 2021: 2 · 2020: 1 · 2019: 3 · 2018: 1 · 2017: 6 · 2016: 4 · 2015: 8 · 2014: 7 · 2013: 6
- FWCI
- 5.15
- Type
- conference paper
- Language
- en
- Open access
- green
- hal.science
- Licence
- other-oa
- Locations
- Interspeech (USB) (publishedVersion)
- HAL (submittedVersion, other-oa)
- Graduate School and Research Center in Digital Science (EURECOM) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Signal Processing › Biometric Identification and Security
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Keywords
- spoofing countermeasures, speaker recognition, spoofing attacks, anti-spoofing, biometric security
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.