SASV Challenge 2022: A Spoofing Aware Speaker Verification Challenge Evaluation Plan

Jung, Jee-weon; Tak, Hemlata; Shim, Hye-jin; Heo, Hee-Soo; Lee, Bong-Jin; Chung, Soo-Whan; Kang, Hong-Goo; Yu, Ha-Jin; Evans, Nicholas; Kinnunen, Tomi

doi:10.48550/arxiv.2201.10283

Cited by 6 publications

(13 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the SASV Challenge 2022 [29], participants are restricted to utilise ASVspoof 2019 [36] and VoxCeleb2 [37] datasets for training the anti-spoofing and ASV model, respectively.…”

Section: Methodsmentioning

confidence: 99%

“…However, previous works only take the stand-alone CM models into account. Recently, a spoofing aware speaker verification (SASV) challenge [29] was proposed as a special session in ISCA INTERSPEECH 2022. This challenge aims to facilitate the research of integrated CM and ASV models, arguing that jointly optimizing CM and ASV models will lead to better performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Spoofing-Aware Speaker Verification by Multi-Level Fusion

Wu¹,

Meng²,

Kang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Recently, many novel techniques have been introduced to deal with spoofing attacks, and achieve promising countermeasure (CM) performances. However, these works only take the standalone CM models into account. Nowadays, a spoofing aware speaker verification (SASV) challenge which aims to facilitate the research of integrated CM and ASV models, arguing that jointly optimizing CM and ASV models will lead to better performance, is taking place. In this paper, we propose a novel multi-model and multi-level fusion strategy to tackle the SASV task. Compared with purely scoring fusion and embedding fusion methods, this framework first utilizes embeddings from CM models, propagating CM embeddings into a CM block to obtain a CM score. In the second-level fusion, the CM score and ASV scores directly from ASV systems will be concatenated into a prediction block for the final decision. As a result, the best single fusion system has achieved the SASV-EER of 0.97% on the evaluation set. Then by ensembling the top-5 fusion systems, the final SASV-EER reached 0.89%.

show abstract

“…In the SASV Challenge 2022 [29], participants are restricted to utilise ASVspoof 2019 [36] and VoxCeleb2 [37] datasets for training the anti-spoofing and ASV model, respectively.…”

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Spoofing-Aware Speaker Verification by Multi-Level Fusion

Wu¹,

Meng²,

Kang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Then, gate operator g outputs self-distilled representation using the two embeddings f θ (ct) and et. Finally, we measure spoof-aware speaker similarity as equation (1).…”

Section: Spoof-aware Speaker Verificationmentioning

confidence: 99%

“…Automatic speaker verification (ASV) system aims to automatically verify the identity of individual speakers for given utterances. To verify that the input sample is directly uttered by the target speaker, it is essential that ASV systems intrinsically reject any kind of spurious attempts, e.g., non-target utterances, synthetic speeches, or converted voices [1]. Although recent ASV systems have shown prominent results, such as achieving an equal error rates (EERs) less than 1% in the in-the-wild scenarios [2,3], there is room for improvement in being prepared for spoofing attacks [4].…”

Section: Introductionmentioning

confidence: 99%

Representation Selective Self-distillation and wav2vec 2.0 Feature Exploration for Spoof-aware Speaker Verification

Lee,

Kim,

Koo

et al. 2022

Preprint

View full text Add to dashboard Cite

Text-to-speech and voice conversion studies are constantly improving to the extent where they can produce synthetic speech almost indistinguishable from bona fide human speech. In this regrad, the importance of countermeasures (CM) against synthetic voice attacks of the automatic speaker verification (ASV) systems emerges. Nonetheless, most end-to-end spoofing detection networks are black box systems, and the answer to what is an effective representation for finding artifacts still remains veiled. In this paper, we examine which feature space can effectively represent synthetic artifacts using wav2vec 2.0, and study which architecture can effectively utilize the space. Our study allows us to analyze which attribute of speech signals is advantageous for the CM systems. The proposed CM system achieved 0.31% equal error rate (EER) on ASVspoof 2019 LA evaluation set for the spoof detection task. We further propose a simple yet effective spoofing aware speaker verification (SASV) methodology, which takes advantage of the disentangled representations from our countermeasure system. Evaluation performed with the SASV Challenge 2022 database show 1.08% of SASV EER. Quantitative analysis shows that using the explored feature space of wav2vec 2.0 advantages both spoofing CM and SASV.

show abstract

“…To encourage research on integrated Spoofing-Aware Speaker Verification (SASV) systems, the SASV Challenge 2022 [15] was proposed using the ASVSpoof 2019 Logical Access Dataset with new metrics, SASV-EER. In the challenge, a single score determines if the input speech sample is the target speaker.…”

Section: Introductionmentioning

confidence: 99%

SA-SASV: An End-to-End Spoof-Aggregated Spoofing-Aware Speaker Verification System

Teng¹,

Fu²,

White³

et al. 2022

Preprint

View full text Add to dashboard Cite

Research in the past several years has boosted the performance of automatic speaker verification systems and countermeasure systems to deliver low Equal Error Rates (EERs) on each system. However, research on joint optimization of both systems is still limited. The Spoofing-Aware Speaker Verification (SASV) 2022 challenge was proposed to encourage the development of integrated SASV systems with new metrics to evaluate joint model performance. This paper proposes an ensemblefree end-to-end solution, known as Spoof-Aggregated-SASV (SA-SASV) to build a SASV system with multi-task classifiers, which are optimized by multiple losses and has more flexible requirements in training set. The proposed system is trained on the ASVSpoof 2019 LA dataset, a spoof verification dataset with small number of bonafide speakers. Results of SASV-EER indicate that the model performance can be further improved by training in complete automatic speaker verification and countermeasure datasets.

show abstract

SASV Challenge 2022: A Spoofing Aware Speaker Verification Challenge Evaluation Plan

Cited by 6 publications

References 23 publications

Spoofing-Aware Speaker Verification by Multi-Level Fusion

Spoofing-Aware Speaker Verification by Multi-Level Fusion

Representation Selective Self-distillation and wav2vec 2.0 Feature Exploration for Spoof-aware Speaker Verification

SA-SASV: An End-to-End Spoof-Aggregated Spoofing-Aware Speaker Verification System

Contact Info

Product

Resources

About