CoChat
Sign Up
Cited by 10Speech Recognition and SynthesisBiometric Identification and SecuritySpeech and Audio Processing

Towards single integrated spoofing-aware speaker verification embeddings

Open access
Mun, Sung Hwan, Shim, Hye-jin, Tak, Hemlata, Wang, Xin, Liu, Xuechen, Sahidullah, Md, Jeong, Myeonghun, Han, Min Hyun +7 more·2023·INTERSPEECH 2023
DOI 10.21437/interspeech.2023-1402
Open this paper in CoChat — free accountOpen PDF

Abstract

This study aims to develop a single integrated spoofing-aware speaker verification (SASV) embeddings that satisfy two aspects. First, rejecting non-target speakers' input as well as target speakers' spoofed inputs should be addressed. Second, competitive performance should be demonstrated compared to the fusion of automatic speaker verification (ASV) and countermeasure (CM) embeddings, which outperformed single embedding solutions by a large margin in the SASV2022 challenge. We analyze that the inferior performance of single SASV embeddings comes from insufficient amount of training data and distinct nature of ASV and CM tasks. To this end, we propose a novel framework that includes multi-stage training and a combination of loss functions. Copy synthesis, combined with several vocoders, is also exploited to address the lack of spoofed data. Experimental results show dramatic improvements, achieving a SASV-EER of 1.06% on the evaluation protocol of the SASV2022 challenge.

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.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.21437/interspeech.2023-1402

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.

A few free checks a day · papers already verified open instantly · results are public

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

Still being checked.

Research works printed but not linked: 43

  1. T. Kinnunen and H. Li, "An overview of text-independent speaker recognition: From features to supervectors," Speech communica- tion, vol. 52, no. 1, 2010.
  2. Z. Bai and X.-L. Zhang, "Speaker recognition based on deep learning: An overview," Neural Networks, vol. 140, 2021.
  3. Y. Zhang, Z. Lv, H. Wu et al., "Mfa-conformer: Multi-scale fea- ture aggregation conformer for automatic speaker verification," in Proc. INTERSPEECH, 2022.
  4. Z. Chen, S. Chen, Y. Wu et al., "Large-scale self-supervised speech representation learning for automatic speaker verification," in Proc. ICASSP, 2022.
  5. T. Kinnunen, M. Sahidullah, H. Delgado et al., "The ASVspoof 2017 Challenge: Assessing the limits of replay spoofing attack detection," in Proc. INTERSPEECH, 2017.
  6. X. Wang, J. Yamagishi, M. Todisco et al., "ASVspoof 2019: A large-scale public database of synthesized, converted and replayed speech," Computer Speech & Language, vol. 64, 2020.
  7. J. Yamagishi, X. Wang, M. Todisco et al., "ASVspoof 2021: ac- celerating progress in spoofed and deepfake speech detection," in Proc. the Automatic Speaker Verification and Spoofing Counter- measures Challenge, 2021.
  8. J.-w. Jung, H. Tak, H.-j. Shim et al., "SASV 2022: The first spoofing-aware speaker verification challenge," in Proc. INTER- SPEECH, 2022.
  9. H.-j. Shim, J.-w. Jung, J.-h. Kim and H.-j. Yu, "Integrated replay spoofing-aware text-independent speaker verification," Applied Sciences, vol. 10, no. 18, 2020.
  10. J. Li, M. Sun, X. Zhang and Y. Wang, "Joint decision of anti- spoofing and automatic speaker verification by multi-task learning with contrastive loss," IEEE Access, vol. 8, 2020.
  11. M. Todisco, H. Delgado, K. A. Lee et al., "Integrated presenta- tion attack detection and automatic speaker verification: Common features and gaussian back-end fusion," in Proc. INTERSPEECH, 2018.
  12. Y. Zhang, G. Zhu and Z. Duan, "A probabilistic fusion frame- work for spoofing aware speaker verification," in Proc. Speaker Odyssey, 2022.
  13. J. Lin, T. Chen, J. Huang et al., "The CLIPS system for 2022 spoofing-aware speaker verification challenge," in Proc. INTER- SPEECH, 2022.
  14. X. Wang, X. Qin, Y. Wang et al., "The DKU-OPPO system for the 2022 spoofing-aware speaker verification challenge," in Proc. INTERSPEECH, 2022.
  15. J.-H. Choi, J.-Y. Yang, Y. R. Jeoung and C. J.-H., "HYU sub- mission for the SASV challenge 2022: Reforming speaker em- beddings with spoofing-aware conditioning," in Proc. INTER- SPEECH, 2022.
  16. W. H. Kang, J. Alam and A. Fathan, "End-to-end framework for spoof-aware speaker verification," in Proc. INTERSPEECH, 2022.
  17. Z. Teng, Q. Fu, J. White et al., "SA-SASV: An end-to-end spoof- aggregated spoofing-aware speaker verification system," in Proc. INTERSPEECH, 2022.
  18. W. Ge, H. Tak, M. Todisco and N. Evans, "On the potential of jointly-optimised solutions to spoofing attack detection and auto- matic speaker verification," in Proc. IberSPEECH, 2022.
  19. A. Alenin, N. Torgashov, A. Okhotnikov et al., "A subnetwork approach for spoofing aware speaker verification," in Proc. IN- TERSPEECH, 2022.
  20. P. Zhang, P. Hu and X. Zhang, "Norm-constrained score-level ensemble for spoofing aware speaker verification," in Proc. IN- TERSPEECH, 2022.
  21. L. Zhang, Y. Li, H. Zhao and L. Xie, "Backend ensemble for speaker verification and spoofing countermeasure," in Proc. IN- TERSPEECH, 2022.
  22. J. S. Chung, A. Nagrani and A. Zisserman, "Voxceleb2: Deep speaker recognition," in Proc. INTERSPEECH, 2018.
  23. X. Wang and J. Yamagishi, "Spoofed training data for speech spoofing countermeasure can be efficiently created using neural vocoders," in Proc. ICASSP, 2023.
  24. S. H. Mun, J.-w. Jung, M. H. Han and N. S. Kim, "Frequency and multi-scale selective kernel attention for speaker verification," in Proc. SLT, 2023.
  25. W. J. Holmes, "Copy synthesis of female speech using the JSRU parallel formant synthesiser.," in Proc. EUROSPEECH, 1989, pp. 2513-2516.
  26. J. Deng, J. Guo, N. Xue and S. Zafeiriou, "Arcface: Additive an- gular margin loss for deep face recognition," in Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition, 2019.
  27. H. S. Heo, B.-J. Lee, J. Huh and J. S. Chung, "Clova baseline sys- tem for the voxceleb speaker recognition challenge 2020," arXiv preprint arXiv:2009.14153, 2020. · arXiv
  28. Y. Kwon, H. S. Heo, B.-J. Lee and J. S. Chung, "The ins and outs of speaker recognition: lessons from VoxSRC 2020," in Proc. ICASSP, 2021.
  29. M. Pal, D. Paul and G. Saha, "Synthetic speech detection using fundamental frequency variation and spectral features," Computer Speech & Language, vol. 48, 2018.
  30. J. Frank and L. Schönherr, "Wavefake: A data set to facilitate au- dio deepfake detection," in Proc. Neural Information Processing Systems Datasets and Benchmarks, 2021.
  31. Z. Wu, X. Xiao, E. S. Chng and H. Li, "Synthetic speech detection using temporal modulation feature," in Proc. ICASSP, 2013.
  32. A. Sizov, E. Khoury, T. Kinnunen et al., "Joint speaker verification and antispoofing in the i-vector space," IEEE Transactions on Information Forensics and Security, vol. 10, no. 4, 2015.
  33. C. Sun, S. Jia, S. Hou et al., "Exposing ai-synthesized hu- man voices using neural vocoder artifacts," arXiv preprint arXiv:2302.09198, 2023. · arXiv
  34. J. Kong, J. Kim and J. Bae, "HiFi-GAN: Generative adversarial networks for efficient and high fidelity speech synthesis," in Proc. NIPS, 2020, vol. 33.
  35. R. Prenger, R. Valle and B. Catanzaro, "WaveGlow: A flow-based generative network for speech synthesis," in Proc. ICASSP, 2019.
  36. X. Wang, S. Takaki and J. Yamagishi, "Neural source-filter-based waveform model for statistical parametric speech synthesis," in Proc. ICASSP, 2019.
  37. J. Engel, L. Hantrakul, C. Gu and A. Roberts, "DDSP: Differen- tiable digital signal processing," in Proc. ICLR, 2020.
  38. J. S. Chung, J. Huh, S. Mun et al., "In defence of metric learning for speaker recognition," in Proc. INTERSPEECH, 2020.
  39. R. Caruana, Multitask learning, Springer, 1998.
  40. I. Loshchilov and F. Hutter, "Decoupled weight decay regulariza- tion," in Proc. ICLR, 2017.
  41. B. Desplanques, J. Thienpondt and K. Demuynck, "Ecapa-tdnn: Emphasized channel attention, propagation and aggregation in tdnn based speaker verification," in Proc. INTERSPEECH, 2020.
  42. J.-w. Jung, H.-S. Heo, H. Tak et al., "Aasist: Audio anti-spoofing using integrated spectro-temporal graph attention networks," in Proc. ICASSP, 2022.
  43. H.-j. Shim, H. Tak, X. Liu et al., "Baseline systems for the first spoofing-aware speaker verification challenge: score and embed- ding fusion," in Proc. Speaker Odyssey, 2022.

Cited by

Sorted by citations

Related

Authors
Mun, S. H., Shim, H.-J., Tak, H., Wang, X., Liu, X., Sahidullah, M., Jeong, M., Han, M. H., Todisco, M., Lee, K. A., Yamagishi, J., Evans, N., Kinnunen, T., Kim, N. S., & Jung, J.-W.
ORCID
Venue
INTERSPEECH 2023
Publisher
ISCA
Published
2023-08-20
Pages
3989-3993
ISSN
2308-457X (ISSN-L)
DOI
10.21437/interspeech.2023-1402
OpenAlex
W4385823275
Cited by
9
References
0
Citations by year
2026: 2 · 2025: 5 · 2024: 1 · 2023: 1
FWCI
1.60
Type
conference paper
Language
en
Open access
Locations
  • Interspeech (USB) (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Funders
Commercializations Promotion Agency for R and D Outcomes
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
speaker verification, 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