Dataset Artefacts in Anti-Spoofing Systems: A Case Study on the ASVspoof 2017 Benchmark
Open accessThe DOI matches its Crossref record; 65 of the 65 references it cites were found in a registry or an index.
Abstract
The Automatic Speaker Verification Spoofing and Countermeasures Challenges motivate research in protecting speech biometric systems against a variety of different access attacks. The 2017 edition focused on replay spoofing attacks, and involved participants building and training systems on a provided dataset (ASVspoof 2017). More than 60 research papers have so far been published with this dataset, but none have sought to answer why countermeasures appear successful in detecting spoofing attacks. This article shows how artefacts inherent to the dataset may be contributing to the apparent success of published systems. We first inspect the ASVspoof 2017 dataset and summarize various artefacts present in the dataset. Second, we demonstrate how countermeasure models can exploit these artefacts to appear successful in this dataset. Third, for reliable and robust performance estimates on this dataset we propose discarding nonspeech segments and silence before and after the speech utterance during training and inference. We create speech start and endpoint annotations in the dataset and demonstrate how using them helps countermeasure models become less vulnerable from being manipulated using artefacts found in the dataset. Finally, we provide several new benchmark results for both frame-level and utterance-level models that can serve as new baselines on this dataset.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures








Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Last, we have been alerted to the presence of shortcut artefacts (Geirhos et al., 2020) which plagues some earlier ASVspoof databases (Chettri et al., 2020; Müller et al., 2021; Shim et al., 2023; Zhang et al., 2023; Liu et al., 2023), namely artefacts which are semantically unrelated to the spoof/deepfake problem but which can nonetheless be utilised for detection.”
From ASVspoof 5: Design, collection and validation of resources for spoofing, deepfake, and adversarial attack detection using crowdsourced speech (2026) · cites this paper as Chettri et al., 2020
The cited source says what the claim attributes to it.
In the citing paper
“Our analysis revealed five potential shortcut artefacts: the peak waveform amplitude; the durations of the leading and trailing non-speech segments (Chettri et al., 2020; Müller et al., 2021); the duration and energy of the whole utterance.”
From ASVspoof 5: Design, collection and validation of resources for spoofing, deepfake, and adversarial attack detection using crowdsourced speech (2026) · cites this paper as Chettri et al., 2020
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · cited 2×
“Some of the well-known deep models from computer vision including ResNet [24] and light CNN [25] were adapted for replay spoofing detection in [20], [21] and [16], respectively demonstrating good performance on the ASVspoof 2017 dataset.”
MethodScikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016 · cited 2×
“We use the Scikit-learn [55] library with default parameters for training and testing the SVM model.”
MethodPattern Recognition and Machine Learning
(2006) · doi:10.1007/978-0-387-45528-0 · Cited by 9,824
Intriguing properties of neural networks
Szegedy, C., Zaremba, W., Sutskever, I., Bruna, J., Erhan, D., Goodfellow, I., & Fergus, R. (2013) · arXiv · doi:10.48550/arxiv.1312.6199 · Cited by 8,143 · cited 1×
“Studies on fooling ML models using adversarial machine learning is an active research topic [59]–[61].”
Background(2006) · Springer eBooks · Cited by 8,123
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1512.03385 · Cited by 4,196 · cited 2×
“Some of the well-known deep models from computer vision including ResNet [24] and light CNN [25] were adapted for replay spoofing detection in [20], [21] and [16], respectively demonstrating good performance on the ASVspoof 2017 dataset.”
MethodFront-End Factor Analysis for Speaker Verification
Dehak, N., Kenny, P. J., Dehak, R., Dumouchel, P., & Ouellet, P. (2011) · IEEE Transactions on Audio, Speech, and Language Processing · doi:10.1109/tasl.2010.2064307 · Cited by 3,549 · cited 1×
“We study five different types of classifiers: GMMs trained on CQCCs [15]; Cosine Distance and Support Vector Machines (SVMs) trained on CQCC-based i-vectors [52]; and two CNNs”
MethodDeep neural networks are easily fooled: High confidence predictions for unrecognizable images
Nguyen, A., Yosinski, J., & Clune, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298640 · Cited by 3,258 · cited 1×
“models using adversarial machine learning is an active research topic [59]–[61].”
MethodDeep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Images
Nguyen, A., Yosinski, J., & Clune, J. (2014) · arXiv · doi:10.48550/arxiv.1412.1897 · Cited by 3,258 · cited 1×
“models using adversarial machine learning is an active research topic [59]–[61].”
MethodX-Vectors: Robust DNN Embeddings for Speaker Recognition
Snyder, D., Garcia-Romero, D., Sell, G., Povey, D., & Khudanpur, S. (2018) · 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2018.8461375 · Cited by 2,620
Torralba, A., & Efros, A. A. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995347 · Cited by 2,440 · cited 1×
“Therefore, it is essential to perform an in-depth dataset analysis [46], [47], detect presence of artefacts or confounders [43], ensuring models do not exploit irrelevant”
BackgroundSpeaker identification and verification using Gaussian mixture speaker models
Reynolds, D. A. (1995) · Speech Communication · doi:10.1016/0167-6393(95)00009-d · Cited by 1,215 · cited 1×
“AUTOMATIC speaker verification (ASV) [1] aims at verifying the claimed identity of a person using their voice”
BackgroundA Light CNN for Deep Face Representation With Noisy Labels
Wu, X., He, R., Sun, Z., & Tan, T. (2018) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2018.2833032 · Cited by 1,186 · cited 1×
“models from computer vision including ResNet [24] and light CNN [25] were adapted for replay spoofing detection in [20], [21] and [16], respectively demonstrating good performance on”
MethodMachine Learning and Prediction in Medicine — Beyond the Peak of Inflated Expectations
Chen, J. H., & Asch, S. M. (2017) · New England Journal of Medicine · doi:10.1056/nejmp1702071 · Cited by 1,171 · cited 1×
“Therefore, it is essential to perform an in-depth dataset analysis [46], [47], detect presence of artefacts or confounders [43], ensuring models do not exploit irrelevant”
BackgroundSpeaker Recognition from Raw Waveform with SincNet
Ravanelli, M., & Bengio, Y. (2018) · 2018 IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt.2018.8639585 · Cited by 896
ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection
Todisco, M., Wang, X., Vestman, V., Sahidullah, M., Delgado, H., Nautsch, A., Yamagishi, J., Evans, N., Kinnunen, T., & Lee, K. A. (2019) · arXiv · doi:10.48550/arxiv.1904.05441 · Cited by 688
ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection
Todisco, M., Wang, X., Vestman, V., Sahidullah, M., Delgado, H., Nautsch, A., Yamagishi, J., Evans, N., Kinnunen, T. H., & Lee, K. A. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2249 · Cited by 687
The ASVspoof 2017 challenge: assessing the limits of replay spoofing attack detection
Kinnunen, T., Sahidullah, M., Delgado, H., Todisco, M., Evans, N., Yamagishi, J., & Lee, K. A. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-1111 · Cited by 535 · cited 2×
“See [9] [10] and [11] for an overview of the ASVspoof challenges held in 2015, 2017 and 2019 respectively.”
BackgroundMethodASVspoof 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 · cited 2×
“See [9] [10] and [11] for an overview of the ASVspoof challenges held in 2015, 2017 and 2019 respectively.”
BackgroundConstant Q Cepstral Coefficients: A Spoofing Countermeasure for Automatic Speaker Verification
Todisco, M., Delgado, H., & Evans, N. (2017) · Computer Speech & Language · doi:10.1016/j.csl.2017.01.001 · Cited by 424
Ethics Guidelines for Trustworthy AI
Cannarsa, M. (2021) · The Cambridge Handbook of Lawyering in the Digital Age · doi:10.1017/9781108936040.022 · Cited by 404 · cited 1×
“To that end, following the guidelines from [33], this paper considers technical robustness, fairness and accountability as key criteria for any machine learning (ML) model to be called trustworthy .”
BackgroundWhy deep-learning AIs are so easy to fool
Heaven, D. (2019) · Nature · doi:10.1038/d41586-019-03013-5 · Cited by 342
Audio replay attack detection with deep learning frameworks
Lavrentyeva, G., Novoselov, S., Malykh, E., Kozlov, A., Kudashev, O., & Shchemelinin, V. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-360 · Cited by 292 · cited 7×
“Spoofing countermeasures are mostly developed either in a frame-level [15] or a fixed-duration utterance-level [16] setting.”
BackgroundMethodInfluentialASVspoof 2017 Version 2.0: meta-data analysis and baseline enhancements
Delgado, H., Todisco, M., Sahidullah, M., Evans, N., Kinnunen, T., Lee, K. A., & Yamagishi, J. (2018) · The Speaker and Language Recognition Workshop (Odyssey 2018) · doi:10.21437/odyssey.2018-42 · Cited by 178
A Simple Method to Determine if a Music Information Retrieval System is a “Horse”
Sturm, B. L. (2014) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2014.2330697 · Cited by 173 · cited 1×
“The trustworthiness of models trained on data containing artefacts is therefore called into question; some can behave much like a “horse” [32], [44], i.”
BackgroundSelf-citationResNet and Model Fusion for Automatic Spoofing Detection
Chen, Z., Xie, Z., Zhang, W., & Xu, X. (2017) · Interspeech (USB) · doi:10.21437/interspeech.2017-1085 · Cited by 171 · cited 2×
“models from computer vision including ResNet [24] and light CNN [25] were adapted for replay spoofing detection in [20], [21] and [16], respectively demonstrating good performance on”
BackgroundMethodrVAD: An unsupervised segment-based robust voice activity detection method
Tan, Z., Sarkar, A. K., & Dehak, N. (2019) · Computer Speech & Language · doi:10.1016/j.csl.2019.06.005 · Cited by 163
The GTZAN dataset: Its contents, its faults, their effects on evaluation, and its future usePreprint
Sturm, B. L. T. (2013) · VBN Forskningsportal (Aalborg Universitet) · doi:10.48550/arxiv.1306.1461 · Cited by 156 · cited 3×
“As demonstrated in [35]–[37], models can learn irrelevant cues, artefacts or confounders during training.”
BackgroundSelf-citationAn analysis of the GTZAN music genre dataset
Sturm, B. L. T. (2012) · International ACM Workshop on Music Information Retrieval with User-Centered and Multimodal Strategies (MIRUM) · doi:10.1145/2390848.2390851 · Cited by 155 · cited 3×
“As demonstrated in [35]–[37], models can learn irrelevant cues, artefacts or confounders during training.”
BackgroundSelf-citationA Light CNN for Deep Face Representation with Noisy LabelsPreprint
Wu, X., He, R., Sun, Z., et al. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.02683 · Cited by 151 · cited 1×
“models from computer vision including ResNet [24] and light CNN [25] were adapted for replay spoofing detection in [20], [21] and [16], respectively demonstrating good performance on”
MethodA study on replay attack and anti-spoofing for text-dependent speaker verification
Wu, Z., Gao, S., Cling, E. S., & Li, H. (2014) · Signal and Information Processing Association Annual Summit and Conference (APSIPA), 2014 Asia-Pacific · doi:10.1109/apsipa.2014.7041636 · Cited by 147 · cited 1×
“Impersonation [4] and replay [5] are examples of such attacks.”
BackgroundAudio Replay Attack Detection Using High-Frequency Features
Witkowski, M., Kacprzak, S., Żelasko, P., Kowalczyk, K., & Gałka, J. (2017) · Interspeech 2017 · doi:10.21437/interspeech.2017-776 · Cited by 140 · cited 1×
“with GMMs (or deep neural networks) have been studied extensively on the ASVspoof 2017 dataset [12], [13], [53], [54]; (2) CQCC-based i-vectors with Cosine backend classifiers have”
MethodLeakage in data mining: formulation, detection, and avoidance
Kaufman, S., Rosset, S., & Perlich, C. (2011) · Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/2020408.2020496 · Cited by 137 · cited 1×
“Data leakage can often lead to overestimation of model performance, producing too-good-to-be-true results [41].”
BackgroundVulnerability of speaker verification to voice mimicking
Lau, Y. W., Wagner, M., & Tran, D. (2005) · Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing, 2004. · doi:10.1109/isimp.2004.1434021 · Cited by 128 · cited 1×
“Impersonation [4] and replay [5] are examples of such attacks.”
BackgroundStowell, D., Petrusková, T., Šálek, M., & Linhart, P. (2019) · Journal of The Royal Society Interface · doi:10.1098/rsif.2018.0940 · Cited by 107 · cited 3×
“Unless explicitly accounted for during training and inference such artefacts often contribute in achieving good results overestimating the actual performance on a testing set [35]–[43].”
BackgroundKinnunen, T., Sahidullah, M., Falcone, M., Costantini, L., Hautamaki, R. G., Thomsen, D., Sarkar, A., Tan, Z.-H., Delgado, H., Todisco, M., et al. (2017) · 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2017.7953187 · Cited by 105 · cited 1×
“The individual replayed utterance was then retrieved based on this marker [56].”
MethodEnsemble Models for Spoofing Detection in Automatic Speaker Verification
Chettri, B., Stoller, D., Morfi, V., Ramírez, M. A. M., Benetos, E., & Sturm, B. L. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2505 · Cited by 103
Self-citationEnd-to-end audio replay attack detection using deep convolutional networks with attention
Tom, F., Jain, M., & Dey, P. (2018) · Interspeech 2018 · doi:10.21437/interspeech.2018-2279 · Cited by 89 · cited 3×
“Furthermore, attention-based models for replay detection have also been studied in [26] and [22].”
BackgroundMethodIntroduction 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
On the security of HMM-based speaker verification systems against imposture using synthetic speech
Masuko, T., Hitotsumatsu, T., Tokuda, K., & Kobayashi, T. (1999) · 6th European Conference on Speech Communication and Technology (Eurospeech 1999) · doi:10.21437/eurospeech.1999-286 · Cited by 82 · cited 1×
“to-Speech [6] and Voice Conversion [7] techniques are often used to produce artificial speech to perform LA attacks.”
BackgroundPatil, H. A., Kamble, M. R., Patel, T. B., & Soni, M. H. (2017) · Interspeech (USB) · doi:10.21437/interspeech.2017-1362 · Cited by 77 · cited 2×
“Teager energy operator (TEO) based spoof detection features were studied in [17].”
BackgroundTommasi, T., Patricia, N., Caputo, B., & Tuytelaars, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24947-6_42 · Cited by 75
An experimental study of speaker verification sensitivity to computer voice-altered imposters
Pellom, B. L., & Hansen, J. H. L. (1999) · 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258) · doi:10.1109/icassp.1999.759801 · Cited by 70 · cited 2×
“Text-to-Speech [6] and Voice Conversion [7] techniques are often used to produce artificial speech to perform LA attacks.”
BackgroundAttentive Filtering Networks for Audio Replay Attack Detection
Lai, C.-I., Abad, A., Richmond, K., Yamagishi, J., Dehak, N., & King, S. (2019) · ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2019.8682640 · Cited by 68
Cai, W., Cai, D., Liu, W., Li, G., & Li, M. (2017) · Interspeech (USB) · doi:10.21437/interspeech.2017-906 · Cited by 67
Modulation Dynamic Features for the Detection of Replay Attacks
Suthokumar, G., Sethu, V., Wijenayake, C. A., & Eliathamby Ambikairajah, A. (2018) · Interspeech (USB) · doi:10.21437/interspeech.2018-1846 · Cited by 66 · cited 1×
“Section IV, some of them have a profound impact on model decisions raising concerns on the validity of results reported in the literature [16], [26], [50], [51].”
BackgroundDetecting Spoofing Attacks Using VGG and SincNet: BUT-Omilia Submission to ASVspoof 2019 Challenge
Zeinali, H., Stafylakis, T., Athanasopoulou, G., Rohdin, J., Gkinis, I., Burget, L., & Černocký, J. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2892 · Cited by 62
Effectiveness of Speech Demodulation-Based Features for Replay Detection
Kamble, M. R., Tak, H., & Patil, H. A. (2018) · Interspeech (USB) · doi:10.21437/interspeech.2018-1675 · Cited by 55 · cited 1×
“features using the TEO and the Hilbert transform have been studied in [23].”
MethodExtended Constant-Q Cepstral Coefficients for Detection of Spoofing Attacks
Yang, J., Das, R. K., & Li, H. (2018) · 2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.23919/apsipa.2018.8659537 · Cited by 53 · cited 1×
“with GMMs (or deep neural networks) have been studied extensively on the ASVspoof 2017 dataset [12], [13], [53], [54]; (2) CQCC-based i-vectors with Cosine backend classifiers have”
MethodIncreasing the robustness of DNNs against image corruptions by playing the Game of NoisePreprint
Rusak, E., Schott, L., Zimmermann, R., Bitterwolf, J., Bringmann, O., Bethge, M., & Brendel, W. (2020) · arXiv (Cornell University) · Cited by 49
Medical data mining: insights from winning two competitions
Rosset, S., Perlich, C., Świrszcz, G., Melville, P., & Liu, Y. (2009) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-009-0158-x · Cited by 44 · cited 3×
“In [37], authors found the encoding of patient identity in a dataset correlates highly with cancer likelihood.”
BackgroundA data augmentation methodology for training machine/deep learning gait recognition algorithms
Charalambous, C., & Bharath, A. (2016) · Procedings of the British Machine Vision Conference 2016 · doi:10.5244/c.30.110 · Cited by 43
Decision-level Feature Switching as a Paradigm for Replay Attack Detection
S, S. M., & Murthy, H. A. (2018) · Interspeech (USB) · doi:10.21437/interspeech.2018-1494 · Cited by 36 · cited 1×
“As we demonstrate in Section IV, some of them have a profound impact on model decisions raising concerns on the validity of results reported in the literature [16], [26], [50], [51].”
BackgroundFrequency Domain Linear Prediction Features for Replay Spoofing Attack Detection
Wickramasinghe, B., Irtza, S., Eliathamby Ambikairajah, A., & Epps, J. (2018) · Interspeech (USB) · doi:10.21437/interspeech.2018-1574 · Cited by 35 · cited 1×
“Authors in [18] proposed features for spoofing detection by exploiting the long-term temporal envelopes of the”
BackgroundSelection bias in the reported performances of AD classification pipelines
Mendelson, A. F., Zuluaga, M. A., Lorenzi, M., Hutton, B. F., & Ourselin, S. (2017) · NeuroImage: Clinical · doi:10.1016/j.nicl.2016.12.018 · Cited by 34
Analysing The Predictions Of a CNN-Based Replay Spoofing Detection System
Chettri, B., Mishra, S., Sturm, B. L., & Benetos, E. (2018) · 2018 IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt.2018.8639666 · Cited by 34 · cited 7×
“Following our prior findings in [14] we hypothesize that the pattern difference favours the bonafide class.”
BackgroundMethodResultInfluentialSelf-citationGazing into Clever Hans machines
Hernández‐Orallo, J. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0032-5 · Cited by 34 · cited 2×
“EER), however their trustworthiness is called into question as they behave much like a “horse” in machine learning [32], [45], overestimating the actual performance.”
BackgroundA Deeper Look at Gaussian Mixture Model Based Anti-Spoofing Systems
Chettri, B., & Sturm, B. L. T. (2018) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2018.8461467 · Cited by 33 · cited 5×
“in [14] and [12] but this study is different in terms of the signature we used for interventions.”
BackgroundMethodResultInfluentialSelf-citationDetection of Replay-Spoofing Attacks Using Frequency Modulation Features
Gunendradasan, T., Wickramasinghe, B., Le, N. P., Eliathamby Ambikairajah, A., & Epps, J. (2018) · Interspeech (USB) · doi:10.21437/interspeech.2018-1473 · Cited by 33 · cited 2×
“Spectral centroid based frequency modulation features were proposed in [19].”
BackgroundThe “Horse” Inside: Seeking Causes Behind the Behaviors of Music Content Analysis Systems
Sturm, B. L. T. (2016) · Computers in entertainment · doi:10.1145/2967507 · Cited by 29 · cited 2×
“EER), however their trustworthiness is called into question as they behave much like a “horse” in machine learning [32], [45], overestimating the actual performance.”
BackgroundSelf-citationSpeech Replay Detection with x-Vector Attack Embeddings and Spectral Features
Williams, J. A., & Rownicka, J. (2019) · Interspeech (USB) · doi:10.21437/interspeech.2019-1760 · Cited by 23 · cited 1×
“Similarly, x-vectors [29] which were originally proposed for speaker recognition, have been studied for spoofing detection in [30].”
BackgroundCharacterising Confounding Effects in Music Classification Experiments through Interventions
Rodríguez-Algarra, F., Sturm, B. L. T., & Dixon, S. (2019) · Transactions of the International Society for Music Information Retrieval · doi:10.5334/tismir.24 · Cited by 14 · cited 3×
“2Following [38] we define Intervention as a process that updates audio signal either by adding or removing raw samples.”
BackgroundSelf-citationSpeech endpoint annotations and artefact details for ASVspoof 2017 version 2.0 dataset
Chettri, B., Benetos, E., & Sturm, B. L. T. (2020) · Figshare · doi:10.5281/zenodo.3601188 · Cited by 2
Self-citationSpeech endpoint annotations and artefact details for ASVspoof 2017 version 2.0 dataset
Chettri, B., Benetos, E., & Sturm, B. L. T. (2020) · Figshare · doi:10.5281/zenodo.3601187 · Cited by 2
Self-citationThe "Horse'' Inside: Seeking Causes Behind the Behaviours of Music Content Analysis SystemsPreprint
Sturm, B. L. T. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.03044 · Cited by 1 · cited 2×
“EER), however their trustworthiness is called into question as they behave much like a “horse” in machine learning [32], [45], overestimating the actual performance.”
BackgroundSelf-citation
Research works printed but not linked: 2
- 8. K. A. Lee, B. Ma, and H. Li, "Speaker verification makes its debut in smartphone," IEEE Signal Processing Society SLTC Newsletter, 2013.
- 58. W. Cai, C. Danwei, W. Liu, G. Li, and M. Li, "Countermeasures for automatic speaker verification replay spoofing attack: On data augmentation, feature representation, classification and fusion," 2017, pp. 17-21. · DOI
Cited by
ASVspoof 2021: Towards Spoofed and Deepfake Speech Detection in the Wild
Liu, X., Wang, X., Sahidullah, M., Patino, J., Delgado, H., Kinnunen, T., Todisco, M., Yamagishi, J., Evans, N., Nautsch, A., et al. (2023) · IEEE/ACM Transactions on Audio, Speech, and Language Processing · doi:10.1109/taslp.2023.3285283 · Cited by 312
A Comparative Study on Recent Neural Spoofing Countermeasures for Synthetic Speech Detection
Wang, X., & Yamagishi, J. (2021) · Interspeech 2021 · doi:10.21437/interspeech.2021-702 · Cited by 137
A Survey on Speech Deepfake Detection
Li, M., Ahmadiadli, Y., & Zhang, X.-P. (2025) · ACM Computing Surveys · doi:10.1145/3714458 · Cited by 101
“Here, we categorize these models based on their primary structure.”
The Impact of Silence on Speech Anti-Spoofing
Zhang, Y., Li, Z., Lu, J., Hua, H., Wang, W., & Zhang, P. (2023) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2023.3306711 · Cited by 53
“Several studies have demonstrated the impact of silence in both LA and PA spoof speech detection [28], [29], [30], [31], [32], [33].”
BackgroundGe, W., Patino, J., Todisco, M., & Evans, N. (2022) · ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp43922.2022.9747476 · Cited by 40
“Additional motivation comes from the numerous reports in the literature which show that some solutions apply greater attention to non-speech intervals than to speech intervals [17–19].”
BackgroundResult
Metadata
- Authors
- Chettri, B., Benetos, E., & Sturm, B. L. T.
- Affiliations
- University of London (GB)
- KTH Royal Institute of Technology (SE)
- Venue
- IEEE/ACM Transactions on Audio, Speech, and Language Processing
- Published
- 2020
- Volume
- 28
- Pages
- 3018-3028
- ISSN
- 2329-9290 (ISSN-L)
- 2329-9304
- OpenAlex
- W3092742905
- Cited by
- 35
- References
- 61
- Citations by year
- 2026: 3 · 2025: 11 · 2024: 6 · 2023: 6 · 2022: 6 · 2021: 3
- FWCI
- 1.47
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE/ACM Transactions on Audio Speech and Language Processing (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by-nc-sa)
- arXiv (Cornell University) (submittedVersion)
- Queen Mary Research Online (Queen Mary University of London) (submittedVersion)
- arXiv (Cornell University)
- 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 › Signal Processing › Speech and Audio Processing
- Keywords
- ASVspoof 2017, anti-spoofing countermeasures, speaker verification, anti-spoofing, benchmark results
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.