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:10.48550/arxiv.2004.00526

Cited by 2 publications

(2 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work, Conv-TasNet and RawNet2 [31] (a pretrained waveform speaker verification model) are used as the basic model and SI model, respectively. Note the choices are not unique -for example, speaker separation models such as Dualpath methods [32]- [34] could also be used as the basic model.…”

Section: B Proposed Speaker-conditioned Modelmentioning

confidence: 99%

Individualized Conditioning and Negative Distances for Speaker Separation

Sun¹,

Abuhajar²,

Gong³

et al. 2022

Preprint

View full text Add to dashboard Cite

Speaker separation aims to extract multiple voices from a mixed signal. In this paper, we propose two speaker-aware designs to improve the existing speaker separation solutions. The first model is a speaker conditioning network that integrates speech samples to generate individualized speaker conditions, which then provide informed guidance for a separation module to produce well-separated outputs.The second design aims to reduce non-target voices in the separated speech. To this end, we propose negative distances to penalize the appearance of any non-target voice in the channel outputs, and positive distances to drive the separated voices closer to the clean targets. We explore two different setups, weightedsum and triplet-like, to integrate these two distances to form a combined auxiliary loss for the separation networks. Experiments conducted on LibriMix demonstrate the effectiveness of our proposed models.

show abstract

Section: B Proposed Speaker-conditioned Modelmentioning

confidence: 99%

Individualized Conditioning and Negative Distances for Speaker Separation

Sun¹,

Abuhajar²,

Gong³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…X-vectors are also embeddings extracted with Time-Delay Neural Networks (TDNN) [17][18][19] at the segment level, and they became the standard method for speaker recognition applications. Some other deep architectures employed in speaker recognition are RawNet [20,21] and ResNet [22][23][24]. These types of embeddings are extracted in an end-to-end manner, meaning that the network is in charge of both finding adequate representations as well as determining the final decision related to the speaker-related task.…”

Section: Related Workmentioning

confidence: 99%

Residual Information in Deep Speaker Embedding Architectures

Stan¹

2022

Mathematics

View full text Add to dashboard Cite

Speaker embeddings represent a means to extract representative vectorial representations from a speech signal such that the representation pertains to the speaker identity alone. The embeddings are commonly used to classify and discriminate between different speakers. However, there is no objective measure to evaluate the ability of a speaker embedding to disentangle the speaker identity from the other speech characteristics. This means that the embeddings are far from ideal, highly dependent on the training corpus and still include a degree of residual information pertaining to factors such as linguistic content, recording conditions or speaking style of the utterance. This paper introduces an analysis over six sets of speaker embeddings extracted with some of the most recent and high-performing deep neural network (DNN) architectures, and in particular, the degree to which they are able to truly disentangle the speaker identity from the speech signal. To correctly evaluate the architectures, a large multi-speaker parallel speech dataset is used. The dataset includes 46 speakers uttering the same set of prompts, recorded in either a professional studio or their home environments. The analysis looks into the intra- and inter-speaker similarity measures computed over the different embedding sets, as well as if simple classification and regression methods are able to extract several residual information factors from the speaker embeddings. The results show that the discriminative power of the analyzed embeddings is very high, yet across all the analyzed architectures, residual information is still present in the representations in the form of a high correlation to the recording conditions, linguistic contents and utterance duration. However, we show that this correlation, although not ideal, could still be useful in downstream tasks. The low-dimensional projections of the speaker embeddings show similar behavior patterns across the embedding sets with respect to intra-speaker data clustering and utterance outlier detection.

show abstract

Improved RawNet with Feature Map Scaling for Text-independent Speaker Verification using Raw Waveforms

Cited by 2 publications

References 31 publications

Individualized Conditioning and Negative Distances for Speaker Separation

Individualized Conditioning and Negative Distances for Speaker Separation

Residual Information in Deep Speaker Embedding Architectures

Contact Info

Product

Resources

About