Improved RawNet with Feature Map Scaling for Text-Independent Speaker Verification Using Raw Waveforms

Jung, Jee-weon; Kim, Seung-bin; Shim, Hye-jin; Kim, Ju-Ho; Yu, Ha-Jin

doi:10.21437/interspeech.2020-1011

Cited by 68 publications

(50 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This section provides a brief review of past work that led to the development of RawNet2 [7]. The treatment is upon automatic speaker verification.…”

Section: Previous Workmentioning

confidence: 99%

“…RawNet2 [7] combines the merits of the original RawNet approach (RawNet1) with those of SincNet. The first layer of RawNet2 is essentially the same as that of SincNet, whereas the upper layers consist of the same residual blocks and GRU layer as RawNet1.…”

Section: Previous Workmentioning

confidence: 99%

“…This section describes modifications made to the original RawNet2 architecture [7] so that it can be applied successfully to anti-spoofing. Modifications to parameters or architecture components are highlighted in Table 1 in bold text.…”

Section: Application To Anti-spoofingmentioning

confidence: 99%

“…Attack-specific artefacts are likely to be insufficient for the detection of unseen attacks and a higher-level, more generalisable representation is needed to ensure robust performance. This paper reports what is, to the best of our knowledge, the first application of RawNet2 [7] deep neural network classifiers to anti-spoofing. RawNet2 operates directly upon the raw speech waveform, circumventing almost entirely any dependence upon hand-crafted features.…”

Section: Introductionmentioning

confidence: 98%

See 3 more Smart Citations

End-to-End anti-spoofing with RawNet2

Tak

Patino

Todisco

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

164

View full text Add to dashboard Cite

Spoofing countermeasures aim to protect automatic speaker verification systems from being manipulated by spoofed speech signals. While results from the most recent ASVspoof 2019 evaluation show great potential to detect most forms of attack, some continue to evade detection. This paper reports the first application of RawNet2 to anti-spoofing. RawNet2 ingests raw audio and has potential to learn cues that are not detectable using more traditional countermeasure solutions. We describe modifications made to the original RawNet2 architecture so that it can be applied to anti-spoofing. For A17 attacks, our RawNet2 systems results are the second-best reported, while the fusion of RawNet2 and baseline countermeasures gives the second-best results reported for the full ASVspoof 2019 logical access condition. Our results are reproducible with open source software.

show abstract

“…This section provides a brief review of past work that led to the development of RawNet2 [7]. The treatment is upon automatic speaker verification.…”

Section: Previous Workmentioning

confidence: 99%

Section: Previous Workmentioning

confidence: 99%

Section: Application To Anti-spoofingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

See 2 more Smart Citations

End-to-End anti-spoofing with RawNet2

Tak

Patino

Todisco

et al. 2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

164

View full text Add to dashboard Cite

show abstract

“…While MOS is a desired measure for audio naturalness/fidelity in singing voice synthesis tasks, voice similarity is more diffi-1 Audio demo: https://nobody996.github.io/FastSVC/. Therefore, we adopt a pre-trained end-to-end speaker recognition model named RawNet2 [33] to objectively measure the voice similarity of the converted singing samples. We measure cosine similarities between embedding vectors of audio samples and the desired target speaker embedding vectors before and after conversion, where all embedding vectors are computed by the RawNet2 and singer/speaker embedding vectors are obtained by averaging his/her training audio samples.…”

Section: Objective Evaluationmentioning

confidence: 99%

Fastsvc: Fast Cross-Domain Singing Voice Conversion With Feature-Wise Linear Modulation

Liu

Cao

et al. 2021

2021 IEEE International Conference on Multimedia and Expo (ICME)

View full text Add to dashboard Cite

This paper presents FastSVC, a light-weight cross-domain singing voice conversion (SVC) system, which can achieve high conversion performance, with inference speed 4x faster than real-time on CPUs. FastSVC uses Conformer-based phoneme recognizer to extract singer-agnostic linguistic features from singing signals. A feature-wise linear modulation based generator is used to synthesize waveform directly from linguistic features, leveraging information from sineexcitation signals and loudness features. The waveform generator can be trained conveniently using a multi-resolution spectral loss and an adversarial loss. Experimental results show that the proposed FastSVC system, compared with a computationally heavy baseline system, can achieve comparable conversion performance in some scenarios and significantly better conversion performance in other scenarios. Moreover, the proposed FastSVC system achieves desirable cross-lingual singing conversion performance. The inference speed of the FastSVC system is 3x and 70x faster than the baseline system on GPUs and CPUs, respectively.

show abstract

Whisper+AASIST for DeepFake Audio Detection

Luo,

Vinayagam Sivasundari

2024

HCI for Cybersecurity, Privacy and Trust

View full text Add to dashboard Cite

Improved RawNet with Feature Map Scaling for Text-Independent Speaker Verification Using Raw Waveforms

Cited by 68 publications

References 0 publications

End-to-End anti-spoofing with RawNet2

End-to-End anti-spoofing with RawNet2

Fastsvc: Fast Cross-Domain Singing Voice Conversion With Feature-Wise Linear Modulation

Whisper+AASIST for DeepFake Audio Detection

Contact Info

Product

Resources

About