Can Large-Scale Vocoded Spoofed Data Improve Speech Spoofing Countermeasure with a Self-Supervised Front End?
Open accessAbstract
A speech spoofing countermeasure (CM) that discriminates between unseen spoofed and bona fide data requires diverse training data. While many datasets use spoofed data generated by speech synthesis systems, it was recently found that data vocoded by neural vocoders were also effective as the spoofed training data. Since many neural vocoders are fast in building and generation, this study used multiple neural vocoders and created more than 9,000 hours of vocoded data on the basis of the VoxCeleb2 corpus. This study investigates how this large-scale vocoded data can improve spoofing countermeasures that use data-hungry self-supervised learning (SSL) models. Experiments demonstrated that the overall CM performance on multiple test sets improved when using features extracted by an SSL model continually trained on the vocoded data. Further improvement was observed when using a new SSL distilled from the two SSLs before and after the continual training. The CM with the distilled SSL outperformed the previous best model on challenging unseen test sets, including the ASVspoof 2019 logical access, WaveFake, and In-the-Wild.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
[22] showed state-of-the-art performance when using large-scale vocoded spoofed data for continuous SSL training.
“Recently, [22] showed state-of-the-art (SOTA) performance when using large-scale vocoded spoofed data for continuous SSL training.”
From Exploring Self-supervised Embeddings and Synthetic Data Augmentation for Robust Audio Deepfake Detection (2024) · cites this paper as [22]
Section 3.3 reports that continual SSL training on large-scale vocoded data improved pooled error rates. The Conclusion (p. unavailable) says a countermeasure using an additional distillation method outperformed the previously best model on challenging test sets such as In-the-Wild. The unqualified claim of state-of-the-art performance from continual training alone overstates that narrower result. (LLM, conf=0.86)
Experimental — accuracy not yet measuredThe proposed approach outperforms the continuous SSL training strategy of [22] by extending the original training set with vocoded data without requiring an additional large corpus for pre-training the upstream.
“Our proposed approach outperforms the continuous SSL training strategy proposed in [22] by extending the original training set with the vocoded data without requiring an additional large corpus for pre-training the upstream.”
From Exploring Self-supervised Embeddings and Synthetic Data Augmentation for Robust Audio Deepfake Detection (2024) · cites this paper as [22]
Even when only the original ASVspoof 2019 training data is used (without additional vocoded data), the best system achieves a competitive overall performance (NN-ACP with 8.51% EER) compared with systems proposed in [22].
“Moreover, even when only the original training data for ASVspoof 2019 is used (i.e., without additional vocoded data), our corresponding best system still achieves a competitive overall performance (NN-ACP with 8.51% EER) compared with the systems proposed in [22].”
From Exploring Self-supervised Embeddings and Synthetic Data Augmentation for Robust Audio Deepfake Detection (2024) · cites this paper as [22]
The P3 system yields the best results on WaveFake, but our models achieve better results on the remaining test sets and the best overall performance (5.13% vs 9.98% EER).
“The P3 system still yields the best results on WaveFake, but our models achieve better results on the remaining test sets and the best overall performance (5.13% vs 9.98% EER).”
From Exploring Self-supervised Embeddings and Synthetic Data Augmentation for Robust Audio Deepfake Detection (2024) · cites this paper as [22]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · see citing works
fairseq: A Fast, Extensible Toolkit for Sequence Modeling
Ott, M., Edunov, S., Baevski, A., Fan, A., Gross, S., Ng, N., Grangier, D., & Auli, M. (2019) · North American Chapter of the Association for Computational Linguistics · doi:10.18653/v1/n19-4009 · Cited by 3,538 · see citing works
VoxCeleb2: Deep Speaker Recognition
Chung, J. S., Nagrani, A., & Zisserman, A. (2018) · Interspeech 2018 · doi:10.21437/interspeech.2018-1929 · Cited by 2,256 · see citing works
HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech SynthesisPreprint
Kong, J., Kim, J., & Bae, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.05646 · Cited by 1,188 · see citing works
“Waveglow: a flow-based generative network for speech synthesis,”
Prenger, R., Valle, R., & Catanzaro, B. (2019) · ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2019.8683143 · Cited by 1,004 · see citing works
Unsupervised Cross-Lingual Representation Learning for Speech Recognition
Conneau, A., Baevski, A., Collobert, R., Mohamed, A., & Auli, M. (2021) · Interspeech 2021 · doi:10.21437/interspeech.2021-329 · Cited by 637 · see citing works
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 · see citing works
Self-citationASVspoof 2019: A large-scale public database of synthesized, converted and replayed speech
Wang, X., Yamagishi, J., Todisco, M., Delgado, H., Nautsch, A., Evans, N., Sahidullah, M., Vestman, V., Kinnunen, T., Lee, K. A., et al. (2020) · Computer Speech & Language · doi:10.1016/j.csl.2020.101114 · Cited by 505 · see citing works
Self-citationSelf-supervised speech representation learning: A review
Mohamed, A., Lee, H.-Y., Borgholt, L., Havtorn, J. D., Edin, J., Igel, C., Kirchhoff, K., Li, S.-W., Livescu, K., Maaløe, L., et al. (2022) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2022.3207050 · Cited by 372 · cited 2× · see citing works
“A well-performed SSL is typically trained using a self-supervised training criterion [13] on an extensive amount of bona fide speech data and computational resources (e.g., 128 Tesla V100 GPUs [14]).”
BackgroundMethodASVspoof 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 · see citing works
Self-citationDoes Audio Deepfake Detection Generalize?
Müller, N., Czempin, P., Diekmann, F., Froghyar, A., & Böttinger, K. (2022) · Interspeech 2022 · doi:10.21437/interspeech.2022-108 · Cited by 258 · see citing works
Tak, H., Todisco, M., Wang, X., Jung, J.-W., Yamagishi, J., & Evans, N. (2022) · The Speaker and Language Recognition Workshop (Odyssey 2022) · doi:10.21437/odyssey.2022-16 · Cited by 237 · see citing works
Self-citationInvestigating Self-Supervised Front Ends for Speech Spoofing Countermeasures
Wang, X., & Yamagishi, J. (2022) · The Speaker and Language Recognition Workshop (Odyssey 2022) · doi:10.21437/odyssey.2022-14 · Cited by 138 · see citing works
Self-citationNeural Source-Filter Waveform Models for Statistical Parametric Speech Synthesis
Wang, X., Takaki, S., & Yamagishi, J. (2019) · IEEE/ACM Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2019.2956145 · Cited by 133 · see citing works
Self-citationTak, H., Kamble, M. R., Patino, J., Todisco, M., & Evans, N. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746213 · Cited by 133 · see citing works
The Vicomtech Audio Deepfake Detection System Based on Wav2vec2 for the 2022 ADD Challenge
Martin-Donas, J. M., & Alvarez, A. (2022) · ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp43922.2022.9747768 · Cited by 100 · see citing works
AI-Synthesized Voice Detection Using Neural Vocoder Artifacts
Sun, C., Jia, S., Hou, S., & Lyu, S. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw59228.2023.00097 · Cited by 81 · see citing works
The Effect of Silence and Dual-Band Fusion in Anti-Spoofing System
Zhang, Y., Wang, W., & Zhang, P. (2021) · Interspeech 2021 · doi:10.21437/interspeech.2021-1281 · Cited by 80 · see citing works
WaveFake: A Data Set to Facilitate Audio Deepfake Detection
Frank, J., & Schönherr, L. (2021) · arXiv · doi:10.48550/arxiv.2111.02813 · Cited by 76 · see citing works
Deepfake audio detection by speaker verification
Pianese, A., Cozzolino, D., Poggi, G., & Verdoliva, L. (2022) · 2022 IEEE International Workshop on Information Forensics and Security (WIFS) · doi:10.1109/wifs55849.2022.9975428 · Cited by 74 · see citing works
Joint Speaker Verification and Antispoofing in the $i$ -Vector Space
Сизов, А. С., 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 · see citing works
Speech is Silver, Silence is Golden: What do ASVspoof-trained Models Really Learn?
Müller, N., Dieckmann, F., Czempin, P., Canals, R., Böttinger, K., & Williams, J. (2021) · 2021 Edition of the Automatic Speaker Verification and Spoofing Countermeasures Challenge · doi:10.21437/asvspoof.2021-9 · Cited by 69 · see citing works
Wang, X., & Yamagishi, J. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10094779 · Cited by 41 · see citing works
Self-citationInvestigating Active-Learning-Based Training Data Selection for Speech Spoofing Countermeasure
Wang, X., & Yamagishi, J. (2023) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt54892.2023.10023350 · Cited by 16 · see citing works
Self-citationA cross-vocoder study of speaker independent synthetic speech detection using phase information
Sánchez, J., Saratxaga, I., Hernáez, I., Navas, E., & Erro, D. (2014) · Interspeech (USB) · doi:10.21437/interspeech.2014-393 · Cited by 16 · see citing works
Towards single integrated spoofing-aware speaker verification embeddings
Mun, S. H., Shim, H.-J., Tak, H., Wang, X., Liu, X., Sahidullah, M., Jeong, M., Han, M. H., Todisco, M., Lee, K. A., et al. (2023) · INTERSPEECH 2023 · doi:10.21437/interspeech.2023-1402 · Cited by 9 · see citing works
Self-citationZeng, C., Wang, X., Miao, X., Cooper, E., & Yamagishi, J. (2023) · Interspeech (USB) · doi:10.21437/interspeech.2023-125 · Cited by 6 · see citing works
Self-citation
Flagged first, then by citations
1 of the 28 references couldn’t be matched to a known paper.
Research works printed but not linked: 1
- Alexei Baevski, Yuhao Zhou, Abdelrahman Mohamed, and Michael Auli, "Wav2vec 2.0: A Framework for Self- Supervised Learning of Speech Representations," in Proc. NIPS, 2020, vol. 33, pp. 12449-12460.
Cited by
Sorted by citations
A Survey on Speech Deepfake Detection
Li, M., Ahmadiadli, Y., & Zhang, X.-P. (2025) · ACM Computing Surveys · doi:10.1145/3714458 · Cited by 101 · see citing works
The Codecfake Dataset and Countermeasures for the Universally Detection of Deepfake Audio
Xie, Y., Lu, Y., Fu, R., Wen, Z., Wang, Z., Tao, J., Qi, X., Wang, X., Liu, Y., Cheng, H., et al. (2025) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslpro.2025.3525966 · Cited by 51 · see citing works
Self-citationGuragain, A., Liu, T., Pan, Z., Sailor, H. B., & Wang, Q. (2024) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt61566.2024.10832226 · Cited by 19 · see citing works
Generalizable Audio Deepfake Detection via Latent Space Refinement and Augmentation
Huang, W. C., Gu, Y., Wang, Z., Zhu, H., & Qian, Y. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10888328 · Cited by 19 · see citing works
A comprehensive survey with critical analysis for deepfake speech detection
Pham, L. D., Lam, P., Tran, D., Tang, H., Nguyen, T., Schindler, A., Skopik, F., Polonsky, A., & Vu, H. C. (2025) · Computer Science Review · doi:10.1016/j.cosrev.2025.100757 · Cited by 23 · see citing works
Martín-Doñas, J. M., Álvarez, A., Rosello, E., Gomez, A. M., & Peinado, A. M. (2024) · Interspeech 2024 · doi:10.21437/interspeech.2024-942 · Cited by 18 · see citing works
“Our proposed approach outperforms the continuous SSL training strategy proposed in [22] by extending the original training set with the vocoded data without requiring an additional large corpus for pre-training the up-stream.”
BackgroundMethodResultInfluentialTrident of Poseidon: A Generalized Approach for Detecting Deepfake Voices
Doan, T.-P., Dinh-Xuan, H., Ryu, T., Kim, I., Lee, W., Hong, K., & Jung, S. (2024) · ACM SIGSAC Conference on Computer and Communications Security (CCS) · doi:10.1145/3658644.3690311 · Cited by 10 · see citing works
Schäfer, K., Choi, J.-E., & Neu, M. (2024) · The Automatic Speaker Verification Spoofing Countermeasures Workshop (ASVspoof 2024) · doi:10.21437/asvspoof.2024-9 · Cited by 8 · see citing works
“For example, [2] developed the SOTA detector on the in-the-wild dataset achieving an EER of 4.07%.”
InfluentialLlamaPartialSpoof: An LLM-Driven Fake Speech Dataset Simulating Disinformation Generation
Luong, H.-T., Li, H., Zhang, L., Lee, K. A., & Chng, E. S. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10888070 · Cited by 10 · see citing works
Dynamic Ensemble Teacher-Student Distillation Framework for Light-Weight Fake Audio Detection
Xue, J., Fan, C., Yi, J., Zhou, J., & Lv, Z. (2024) · IEEE Signal Processing Letters · doi:10.1109/lsp.2024.3431936 · Cited by 7 · see citing works
What Does an Audio Deepfake Detector Focus on? A Study in the Time Domain
Grinberg, P., Kumar, A., Koppisetti, S., & Bharaj, G. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10887568 · Cited by 7 · see citing works
Emergent Capability in Audio Deepfake Detection
Deng, Y., Laprevotte, A., Dunn, R. B., Khorrami, P., Dagli, C. K., & Torres‐Carrasquillo, P. A. (2025) · International Workshop on Biometrics and Forensics (IWBF) · doi:10.1109/iwbf63717.2025.11113428 · Cited by 5 · see citing works
Kheir, Y. E., Das, A., Erdogan, E. E., Ritter-Guttierez, F., Polzehl, T., & Möller, S. (2025) · IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) · doi:10.1109/waspaa66052.2025.11230938 · Cited by 6 · see citing works
Meta-Learning Approaches For Improving Detection of Unseen Speech Deepfakes
Kukanov, I., Laakkonen, J., Kinnunen, T., & Hautamäki, V. (2024) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt61566.2024.10832350 · Cited by 4 · see citing works
Hao, Y., Chen, Y., Xu, M., Zhan, J., He, L., Fang, L., Fang, S., & Liu, L. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11228261 · Cited by 3 · see citing works
Leveraging distance information for generalized spoofing speech detection
Lu, J., Zhang, Y., Li, Z., Shang, Z., Wang, W., & Zhang, P. (2025) · Computer Speech & Language · doi:10.1016/j.csl.2025.101804 · Cited by 3 · see citing works
Post-training for Deepfake Speech Detection
Ge, W., Wang, X., Liu, X., & Yamagishi, J. (2025) · IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) · doi:10.1109/asru65441.2025.11434709 · Cited by 8 · see citing works
Self-citationNguyen-Le, H.-H., Tran, V.-T., Nguyen, D.-T., & Le-Khac, N.-A. (2026) · Artificial Intelligence Review · doi:10.1007/s10462-026-11608-4 · Cited by 1 · see citing works
Multi-Sinkhorn Teacher Knowledge Aggregation Framework for Adaptive Audio Anti-Spoofing
Zhang, R., Wei, J., Lu, X., Zhang, L., Jin, D., Lu, W., & Xu, J. (2025) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslpro.2025.3606191 · Cited by 1 · see citing works
Text Vs. Speech? Detecting Audio Deepfakes on Instagram
Schäfer, K. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-21300-6_32 · Cited by 1 · see citing works
SHDA: Sinkhorn Domain Attention for Cross-Domain Audio Anti-Spoofing
Zhang, R., Wei, J., Lu, X., Zhang, L., Jin, D., Xu, J., & Lu, W. (2025) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2025.3576576 · Cited by 1 · see citing works
Bidirectional Label Distribution-Aware Margin Method for SpeechPreprint
Wang, R., & Fu, X. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5218867
Make full use of your data: On copy-based augmentation in speech anti-spoofing
Zhang, L., Niu, B., Lee, K. A., & Wang, L. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.130799
Bidirectional label distribution-aware margin method for speech spoofing detection
Wang, R., & Fu, X. (2025) · Digital Signal Processing · doi:10.1016/j.dsp.2025.105660
Impact of neural-vocoder–generated spoofed speech on spoofing-detection performance
Park, S., & Kwak, I. (2025) · Korean Journal of Applied Statistics · doi:10.5351/kjas.2025.38.5.679
Generalization in Audio Deepfake Detection: Evaluating ASR Encoder-Based Feature Extraction
Schäfer, K., & Steinebach, M. (2025) · IEEE International Conference on Data Science and Advanced Analytics (DSAA) · doi:10.1109/dsaa65442.2025.11247976
Cao-Phan, K.-D., & Phuc, D. T. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-4957-3_30
ASV-CM Score Fusion for Robust Speaker Verification in Spoofed Scenarios
Carvalho, L. A. R., Lima, L., Rinaldo, G., Silva, C. O., & Beraldo, V. C. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-15993-9_10
BEAT2AASIST Model with Layer Fusion for ESDD 2026 Challenge
Chung, S. W., Kim, E., Kim, D., Heo, G., You, J., Lee, N., Choi, S., Han, S., Oh, S., & Kwak, I. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11461593
A Parameter-Efficient Multi-Scale Convolutional Adapter for Synthetic Speech Detection
Kheir, Y. E., Ritter-Guttierez, F., Das, A., Polzehl, T., & Möller, S. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11462905
Understanding the Strengths and Weaknesses of SSL Models for Audio Deepfake Model Attribution
Pîrlogeanu, G., Stan, A., & Cucu, H. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11464412
Study of Deepfake Audio Detection in Telephone Speech with Pretrained Self-supervised Models
Sielecki, S., Kisielewicz, M., & Kędziora, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-36868-3_24
- Authors
- Wang, X., & Yamagishi, J.
- Affiliations
- National Institute of Informatics (JP)
- Venue
- ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- Publisher
- IEEE
- Published
- 2024-04-14
- Pages
- 10311-10315
- ISSN
- 1520-6149 (ISSN-L)
- OpenAlex
- W4392910562
- Cited by
- 29
- References
- 28
- Citations by year
- 2026: 6 · 2025: 19 · 2024: 4
- FWCI
- 8.56
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE International Conference on Acoustics Speech and Signal Processing (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Japan Society for the Promotion of Science (21H04906)
- 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
- self-supervised learning, continuous training, knowledge distillation, VoxCeleb2, ASVspoof 2019, speaker verification 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
