Continual Learning for Fake Audio Detection
Open accessThe DOI matches its Crossref record; 30 of the 30 references it cites were found in a registry or an index.
Abstract
Fake audio attack becomes a major threat to the speaker verification system. Although current detection approaches have achieved promising results on dataset-specific scenarios, they encounter difficulties on unseen spoofing data. Fine-tuning and retraining from scratch have been applied to incorporate new data. However, fine-tuning leads to performance degradation on previous data. Retraining takes a lot of time and computation resources. Besides, previous data are unavailable due to privacy in some situations. To solve the above problems, this paper proposes detecting fake without forgetting, a continual-learning-based method, to make the model learn new spoofing attacks incrementally. A knowledge distillation loss is introduced to loss function to preserve the memory of original model. Supposing the distribution of genuine voice is consistent among different scenarios, an extra embedding similarity loss is used as another constraint to further do a positive sample alignment. Experiments are conducted on the ASVspoof2019 dataset. The results show that our proposed method outperforms fine-tuning by the relative reduction of average equal error rate up to 81.62%.
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
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 doesn’t support the claim as written.
In the citing paper
“The evaluation set of the FAD dataset contains 14,000 utterances generated by four unknown deepfake methods that were not included in the training and validation sets, which can better detect the robustness and generalization of the model in the face of unknown attacks.”
From Voice Deepfake Detection Using the Self-Supervised Pre-Training Model HuBERT (2023) · cites this paper as [26]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“DFWF achieved an Average Equal Error Rate (AvgEER) reduction of 81.63% compared to fine-tuning, maintaining robust performance on both new and old tasks of ASVspoof 2019.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [66]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Ma et al. [66] introduced the replay based method known as Detecting Fake Without Forgetting (DFWF) method, a robust framework designed to mitigate catastrophic forgetting and enable efficient audio deepfake detection. DFWF integrates Learning without Forgetting (LwF) and Positive Sample Alignment (PSA) to preserve past knowledge while incrementally adapting to new spoofing types.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [66]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“The FAD dataset consists of 12 types mainstream voice deepfake techniques such as STRAIGHT, LPCNet, and HifiGAN to generate fake audios, and real audios from six different corpora such as AISHELL1, AISHELL3, and THCHS-30 [26].”
From Voice Deepfake Detection Using the Self-Supervised Pre-Training Model HuBERT (2023) · cites this paper as [26]
Experimental — accuracy not yet measured
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
(2008) · Journal of Machine Learning Research · Cited by 33,051
Distilling the Knowledge in a Neural Network
Hinton, G., Vinyals, O., & Dean, J. (2015) · arXiv · doi:10.48550/arxiv.1503.02531 · Cited by 14,204
Overcoming catastrophic forgetting in neural networks
Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A. A., Milan, K., Quan, J., Ramalho, T., Grabska-Barwinska, A., et al. (2017) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1611835114 · Cited by 7,899
Li, Z., & Hoiem, D. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2773081 · Cited by 4,202
iCaRL: Incremental Classifier and Representation Learning
Rebuffi, S.-A., Kolesnikov, A., Sperl, G., & Lampert, C. H. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.587 · Cited by 3,988 · cited 2×
“(2) Replay approaches [14, 21] record some examples in the buffer and replay experience when training new task.”
BackgroundContinual lifelong learning with neural networks: A review
Parisi, G. I., Kemker, R., Part, J. L., Kanan, C., & Wermter, S. (2019) · Neural Networks · doi:10.1016/j.neunet.2019.01.012 · Cited by 3,117
Continual Learning with Deep Generative Replay
Shin, H., Lee, J. K., Kim, J., & Kim, J. (2017) · arXiv · doi:10.48550/arxiv.1705.08690 · Cited by 1,385
Rusu, A. A., Rabinowitz, N. C., Desjardins, G., Soyer, H., Kirkpatrick, J., Kavukcuoglu, K., Pascanu, R., & Hadsell, R. (2016) · arXiv · doi:10.48550/arxiv.1606.04671 · Cited by 1,307
End-to-End Incremental Learning
Castro, F. M., Marín-Jiménez, M. J., Guil, N., Schmid, C., & Alahari, K. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01258-8_15 · Cited by 1,184
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 · cited 1×
“Fake audio detection has attracted increasing attention since the organization of a series of automatic speaker verification spoofing and countermeasures challenge (ASVspoof) [1, 2, 3].”
BackgroundASVspoof 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
Li, Z., & Hoiem, D. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_37 · Cited by 659
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
ASVspoof 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 · cited 3×
“It can be explained by the distribution of LA attacks and PA attacks provided by the publisher of ASVspoof2019 dataset [27], different LA attacks can form relatively clear clustering while different PA attacks”
BackgroundMethodInfluentialASVspoof 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 1×
“Fake audio detection has attracted increasing attention since the organization of a series of automatic speaker verification spoofing and countermeasures challenge (ASVspoof) [1, 2, 3].”
BackgroundA comparison of features for synthetic speech detection
Sahidullah, M., Kinnunen, T., & Hanilçi, C. (2015) · Interspeech 2015 · doi:10.21437/interspeech.2015-472 · Cited by 382 · cited 2×
“We extract 60-dimensional linear frequency cepstral coefficients (LFCC) [28] of speech as low-level input feature to the neural network.”
MethodA New Feature for Automatic Speaker Verification Anti-Spoofing: Constant Q Cepstral Coefficients
Todisco, M., Delgado, H., & Evans, N. (2016) · The Speaker and Language Recognition Workshop (Odyssey 2016) · doi:10.21437/odyssey.2016-41 · Cited by 335
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
STC Antispoofing Systems for the ASVspoof2019 Challenge
Lavrentyeva, G., Novoselov, S., Tseren, A., Volkova, M., Gorlanov, A., & Kozlov, A. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-1768 · Cited by 245
Distilling Knowledge from Ensembles of Neural Networks for Speech Recognition
Chebotar, Y., & Waters, A. (2016) · Interspeech 2016 · doi:10.21437/interspeech.2016-1190 · Cited by 167
Patel, T. B., & Patil, H. A. (2015) · Interspeech 2015 · doi:10.21437/interspeech.2015-467 · Cited by 164
Lifelong learning of human actions with deep neural network self-organization
Parisi, G. I., Tani, J., Weber, C., & Wermter, S. (2017) · Neural Networks · doi:10.1016/j.neunet.2017.09.001 · Cited by 129
Lei, Z., Yang, Y., Liu, C., & Ye, J. (2020) · Interspeech 2020 · doi:10.21437/interspeech.2020-2723 · Cited by 49
Assessing the Scope of Generalized Countermeasures for Anti-Spoofing
Das, R. K., Yang, J., & Li, H. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9053086 · Cited by 45
Continual Learning in Automatic Speech Recognition
Sadhu, S., & Heřmanský, H. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-2962 · Cited by 41
Knowledge Distillation for Recurrent Neural Network Language Modeling with Trust Regularization
Shi, Y., Hwang, M.-Y., Lei, X., & Sheng, H. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8683533 · Cited by 30
Lee, J., Joo, D., Hong, H. G., & Kim, J. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i04.5884 · Cited by 27
Monteiro, J., Alam, J., & Falk, T. H. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9054558 · Cited by 17
Multi-Task Siamese Neural Network for Improving Replay Attack Detection
von Platen, P., Tao, F., & Tür, G. (2020) · Interspeech (USB) · doi:10.21437/interspeech.2020-86 · Cited by 10
Multi-Task Siamese Neural Network for Improving Replay Attack DetectionPreprint
von Platen, P., Tao, F., & Tür, G. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.07629 · Cited by 4
Research works printed but not linked: 4
- [12][12] H. Wang, H. Dinkel, S. Wang, Y. Qian, and K. Yu, 'Dualadversarial domain adaptation for generalized replay attack detection,' Proc. Interspeech 2020 , pp. 1086-1090, 2020. -
- [15][15] G. M. van de Ven, H. T. Siegelmann, and A. S. Tolias, 'Braininspired replay for continual learning with artificial neural networks,' Nature communications , vol. 11, no. 1, pp. 1-14, 2020. -
- [19][19] B. Houston and K. Kirchhoff, 'Continual learning for multidialect acoustic models,' Proc. Interspeech 2020 , pp. 576-580, 2020. -
- [29][29] G. Lavrentyeva, S. Novoselov, A. Tseren, M. Volkova, A. Gorlanov, and A. Kozlov, 'Stc antispoofing systems for the asvspoof2019 challenge,' arXiv preprint arXiv:1904.05576 , 2019. - · arXiv
Cited by
Masood, M., Nawaz, M., Malik, K. M., Javed, A., Irtaza, A., & Malik, H. (2023) · Applied Intelligence · doi:10.1007/s10489-022-03766-z · Cited by 543
ADD 2022: the first Audio Deep Synthesis Detection Challenge
Yi, J., Fu, R., Tao, J., Nie, S., Ma, H., Wang, C., Wang, T., Tian, Z., Bai, Y., Fan, C., et al. (2022) · ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp43922.2022.9746939 · Cited by 165
Self-citationDeepfake Media Forensics: Status and Future Challenges
Amerini, I., Barni, M., Battiato, S., Bestagini, P., Boato, G., Bruni, V., Caldelli, R., De Natale, F., De Nicola, R., Guarnera, L., et al. (2025) · Journal of Imaging · doi:10.3390/jimaging11030073 · Cited by 113
Shaaban, O. A., Yıldırım, R., & Alguttar, A. A. (2023) · IEEE Access · doi:10.1109/access.2023.3333866 · Cited by 107
Recent Advances of Multimodal Continual Learning: A Comprehensive Survey
Yu, D., Zhang, X., Chen, Y., Liu, A., Zhang, Y., Yu, P. S., & King, I. (2026) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2026.3658485 · Cited by 64
Metadata
- Authors
- Ma, H., Yi, J., Tao, J., Bai, Y., Tian, Z., & Wang, C.
- Affiliations
- Chinese Academy of Sciences (CN)
- Center for Excellence in Education (US)
- University of Science and Technology of China (CN)
- Venue
- Interspeech 2021
- Publisher
- ISCA
- Published
- 2021-08-30
- Pages
- 886-890
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W3156192236
- Cited by
- 39
- References
- 27
- Citations by year
- 2026: 4 · 2025: 9 · 2024: 14 · 2023: 3 · 2022: 9
- FWCI
- 3.00
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Interspeech (USB) (publishedVersion)
- arXiv (Cornell University) (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
- continual learning, knowledge distillation, ASVspoof 2019
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.