The Biometric Vox System for the ASVspoof 2021 Challenge

Caceres, Joaquin; Font, Roberto; Grau, Teresa; Molina, J.

doi:10.21437/asvspoof.2021-11

Cited by 11 publications

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As test sets of ADD 2022 include unseen genuine and fake utterances which are not present in train and development data, it is essential to develop CMs that are robust to out-of-domain data. Data augmentation strategy is efficient to improve the performance of anti-spoofing systems on cross-dataset in previous works [23][24][25][26][27][28][29][30][31][32][33]. Thus, we design low-quality data augmentation strategy to address the unseen genuine and fake utterances.…”

Section: Data Augmentationmentioning

confidence: 99%

Deep Spectro-temporal Artifacts for Detecting Synthesized Speech

Liu

Zhang

et al. 2022

Proceedings of the 1st International Workshop on Deepfake Detection for Audio Multimedia

View full text Add to dashboard Cite

The Audio Deep Synthesis Detection (ADD) Challenge has been held to detect generated human-like speech. With our submitted system, this paper provides an overall assessment of track 1 (Lowquality Fake Audio Detection) and track 2 (Partially Fake Audio Detection). In this paper, spectro-temporal artifacts were detected using raw temporal signals, spectral features, as well as deep embedding features. To address track 1, low-quality data augmentation, domain adaptation via finetuning, and various complementary feature information fusion were aggregated in our system. Furthermore, we analyzed the clustering characteristics of subsystems with different features by visualization method and explained the effectiveness of our proposed greedy fusion strategy. As for track 2, frame transition and smoothing were detected using self-supervised learning structure to capture the manipulation of PF attacks in the time domain. We ranked 4th and 5th in track 1 and track 2, respectively.

show abstract

Section: Data Augmentationmentioning

confidence: 99%

Deep Spectro-temporal Artifacts for Detecting Synthesized Speech

Liu

Zhang

et al. 2022

Proceedings of the 1st International Workshop on Deepfake Detection for Audio Multimedia

View full text Add to dashboard Cite

show abstract

“…To verify that this improvement is not due to random factors in neural network training (e.g., different, random initial network weights), we conducted a statistical analysis of the results following [69]. The results 6 suggest that the improvement is statistically significant and is hence unlikely to be caused by factors other than DA. Its effect is more pronounced when using the wav2vec 2.0 front-end for which the EER decreases from 4.48% to 0.82%.…”

Section: Data Augmentationmentioning

confidence: 99%

“…Data augmentation (DA) is already known to reduce overfitting and hence to improve generalisation [3,4,6,7,27] and is particularly effective in the case of LA scenarios in which there is substantial variability stemming from, e.g., encoding, transmission and acquisition devices [2]. We are interested to determine whether self-supervised learning is complementary to DA.…”

Section: Data Augmentationmentioning

confidence: 99%

“…For the most recent ASVspoof 2021 Logical Access (LA) evaluation [2], the equal error rate (EER) of the best performing baseline solution increased from 0.55% for the development set to 9.26% for the evaluation set [2]. Submission results show better performance [3][4][5][6][7][8][9], but the fundamental gap between performance for development and evaluation data remains, indicating a persisting lack of generalisation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation

Tak,

Todisco,

Wang

et al. 2022

Preprint

View full text Add to dashboard Cite

The performance of spoofing countermeasure systems depends fundamentally upon the use of sufficiently representative training data. With this usually being limited, current solutions typically lack generalisation to attacks encountered in the wild. Strategies to improve reliability in the face of uncontrolled, unpredictable attacks are hence needed. We report in this paper our efforts to use self-supervised learning in the form of a wav2vec 2.0 front-end with fine tuning. Despite initial base representations being learned using only bona fide data and no spoofed data, we obtain the lowest equal error rates reported in the literature for both the ASVspoof 2021 Logical Access and Deepfake databases. When combined with data augmentation, these results correspond to an improvement of almost 90% relative to our baseline system.

show abstract