How to Improve Your Speaker Embeddings Extractor in Generic Toolkits

Zeinali, Hossein; Burget, Lukáš; Rohdin, Johan; Stafylakis, Themos; Černocký, Jaň

doi:10.48550/arxiv.1811.02066

Cited by 5 publications

(8 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different with Kaldi, there is no dilation used. This performs better in our experiments and is also suggested in other works [24]. Statistics pooling and a 2-layer segment-level network is appended after the frame-level network.…”

Section: Training Detailssupporting

confidence: 56%

“…The loss converges after the learning rate goes down below 10 −5 , resulting to around 2.5M training steps. No dropout is applied in our networks as described in [24].…”

Section: Training Detailsmentioning

confidence: 99%

“…In this paper, we first build a baseline system using a generic toolkit similar with [24]. Several training strategies are used to improve the accuracy.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Large Margin Softmax Loss for Speaker Verification

Liu¹,

He²,

Lu³

2019

Interspeech 2019

157

100

View full text Add to dashboard Cite

In neural network based speaker verification, speaker embedding is expected to be discriminative between speakers while the intra-speaker distance should remain small. A variety of loss functions have been proposed to achieve this goal. In this paper, we investigate the large margin softmax loss with different configurations in speaker verification. Ring loss and minimum hyperspherical energy criterion are introduced to further improve the performance. Results on VoxCeleb show that our best system outperforms the baseline approach by 15% in EER, and by 13%, 33% in minDCF08 and minDCF10, respectively.

show abstract

Section: Training Detailssupporting

confidence: 56%

“…The loss converges after the learning rate goes down below 10 −5 , resulting to around 2.5M training steps. No dropout is applied in our networks as described in [24].…”

Section: Training Detailsmentioning

confidence: 99%

See 1 more Smart Citation

Large Margin Softmax Loss for Speaker Verification

Liu¹,

He²,

Lu³

2019

Interspeech 2019

157

100

View full text Add to dashboard Cite

show abstract

“…The kernel sizes of the first five layers are 5, 3, 3, 1 and 1, while the dilation rates are set to 1, 2, 3, 1 and 1 respectively. The same type of L2 weight decay and batch normalization as described in [29] are used in the baseline system to prevent overfitting. ACNN: In this system, the ACNN is only applied in the fourth frame-level layer.…”

Section: Model Configurationmentioning

confidence: 99%

An Adaptive X-Vector Model for Text-Independent Speaker Verification

Gu¹,

Guo²,

Ding³

et al. 2020

Interspeech 2020

View full text Add to dashboard Cite

In this paper, adaptive mechanisms are applied in deep neural network (DNN) training for x-vector-based text-independent speaker verification. First, adaptive convolutional neural networks (ACNNs) are employed in frame-level embedding layers, where the parameters of the convolution filters are adjusted based on the input features. Compared with conventional CNNs, ACNNs have more flexibility in capturing speaker information. Moreover, we replace conventional batch normalization (BN) with adaptive batch normalization (ABN). By dynamically generating the scaling and shifting parameters in BN, ABN adapts models to the acoustic variability arising from various factors such as channel and environmental noises. Finally, we incorporate these two methods to further improve performance. Experiments are carried out on the speaker in the wild (SITW) and VOiCES databases. The results demonstrate that the proposed methods significantly outperform the original xvector approach.

show abstract

“…However, [6] a recent work has shown that data augmentation, consisting of added noise and reverberation, can significantly improve the performance of these embeddings (xvectors as they referred to), while it is not so effective for ivectors [6]. There have also been some efforts to improve the quality and generalization powers of x-vectors by the modification applied to the network architecture [10] and the training procedure [11,12,13].…”

Section: Introductionmentioning

confidence: 99%

Self Multi-Head Attention for Speaker Recognition

India¹,

Safari²,

Hernando³

2019

Interspeech 2019

View full text Add to dashboard Cite

Most state-of-the-art Deep Learning (DL) approaches for speaker recognition work on a short utterance level. Given the speech signal, these algorithms extract a sequence of speaker embeddings from short segments and those are averaged to obtain an utterance level speaker representation. In this work we propose the use of an attention mechanism to obtain a discriminative speaker embedding given non fixed length speech utterances. Our system is based on a Convolutional Neural Network (CNN) that encodes short-term speaker features from the spectrogram and a self multi-head attention model that maps these representations into a long-term speaker embedding. The attention model that we propose produces multiple alignments from different subsegments of the CNN encoded states over the sequence. Hence this mechanism works as a pooling layer which decides the most discriminative features over the sequence to obtain an utterance level representation. We have tested this approach for the verification task for the VoxCeleb1 dataset. The results show that self multi-head attention outperforms both temporal and statistical pooling methods with a 18% of relative EER. Obtained results show a 58% relative improvement in EER compared to i-vector+PLDA.

show abstract

How to Improve Your Speaker Embeddings Extractor in Generic Toolkits

Cited by 5 publications

References 22 publications

Large Margin Softmax Loss for Speaker Verification

Large Margin Softmax Loss for Speaker Verification

An Adaptive X-Vector Model for Text-Independent Speaker Verification

Self Multi-Head Attention for Speaker Recognition

Contact Info

Product

Resources

About