Deep Neural Network Embeddings with Gating Mechanisms for Text-Independent Speaker Verification

You, Lanhua; Guo, Wenbin; Dai, Li-Rong; Du, Jun

doi:10.21437/interspeech.2019-1746

Cited by 18 publications

(11 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, [15] proposed the usage of Gated Convolutional Neural Networks (GCNN) for speaker recognition. Matched with a gated-attention pooling method for frame-level feature aggregation, they evaluate the performance of GCNN in an x-vector [16] system on SRE16 and SRE18 datasets.…”

Section: Related Workmentioning

confidence: 99%

Frequency and Temporal Convolutional Attention for Text-Independent Speaker Recognition

Yadav¹,

Rai²

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Majority of the recent approaches for text-independent speaker recognition apply attention or similar techniques for aggregation of frame-level feature descriptors generated by a deep neural network (DNN) front-end. In this paper, we propose methods of convolutional attention for independently modelling temporal and frequency information in a convolutional neural network (CNN) based front-end. Our system utilizes convolutional block attention modules (CBAMs) [1] appropriately modified to accommodate spectrogram inputs. The proposed CNN front-end fitted with the proposed convolutional attention modules outperform the no-attention and spatial-CBAM baselines by a significant margin on the Vox-Celeb [2, 3] speaker verification benchmark. Our best model achieves an equal error rate of 2.031% on the VoxCeleb1 test set, which is a considerable improvement over comparable state of the art results. For a more thorough assessment of the effects of frequency and temporal attention in real-world conditions, we conduct ablation experiments by randomly dropping frequency bins and temporal frames from the input spectrograms, concluding that instead of modelling either of the entities, simultaneously modelling temporal and frequency attention translates to better real-world performance.

show abstract

Section: Related Workmentioning

confidence: 99%

Frequency and Temporal Convolutional Attention for Text-Independent Speaker Recognition

Yadav¹,

Rai²

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…In [14] a Gaussian constrained training approach is proposed that impose on x-vectors to have a Gaussian distribution. Using gated convolutional layers instead of time delay layers and gated pooling layer has improved the performance of x-vector system [15]. In [16], a hybrid LSTM and CNN network used for frame level layers and by using multi-level pooling strategy and applying a regularization scheme on embedding layer, the performance of x-vector baseline system was improved.…”

Section: Introductionmentioning

confidence: 99%

Data augmentation versus noise compensation for x-vector speaker recognition systems in noisy environments

Amini

Matrouf

2021

2020 28th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

The explosion of available speech data and new speaker modeling methods based on deep neural networks (DNN) have given the ability to develop more robust speaker recognition systems. Among DNN speaker modelling techniques, x-vector system has shown a degree of robustness in noisy environments. Previous studies suggest that by increasing the number of speakers in the training data and using data augmentation more robust speaker recognition systems are achievable in noisy environments. In this work, we want to know if explicit noise compensation techniques continue to be effective despite the general noise robustness of these systems. For this study, we will use two different x-vector networks: the first one is trained on Voxceleb1 (Protocol1), and the second one is trained on Voxceleb1+Voxveleb2 (Protocol2).We propose to add a denoising x-vector subsystem before scoring. Experimental results show that, the x-vector system used in Protocol2 is more robust than the other one used Protocol1. Despite this observation we will show that explicit noise compensation gives almost the same EER relative gain in both protocols. For example, in the Protocol2 we have 21% to 66% improvement of EER with denoising techniques.

show abstract

“…By introducing deep learning into anti-spoofing, deep neural networks (DNN) have achieved promising results in anti-spoofing of ASVspoof 2017 [9]- [11] and ASVspoof 2019 [12]. ASV as a standalone task has also gained great improvement from deep learning [13], [14]. Given those achievements and in order to make ASV and anti-spoofing a step forward to practical usage, some early studies have proposed that a separately designed antispoofing system is implement before ASV, only the utterances which have passed spoofing detection are verified again by a ASV system [4], [14], which is in fact a cascaded structure as is illustrated in FIGURE 1.…”

Section: Introductionmentioning

confidence: 99%

Joint Decision of Anti-Spoofing and Automatic Speaker Verification by Multi-Task Learning With Contrastive Loss

Sun

Zhang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Automatic speaker verification (ASV) is an emerging biometric verification technique with more and more applications. However, both verification accuracy and anti-spoofing should be considered carefully before putting ASV into practice, where anti-spoofing is also called replay detection in which voice is recorded, stored and replayed to deceive ASV systems. Cascaded decision of anti-spoofing and ASV is a straightforward solution to tackle the two issues. In this paper, joint decision of anti-spoofing and ASV was investigated in a multi-task learning framework with contrastive loss in order to improve the cascaded decision approach. A modified triplet loss was firstly constructed to supervise deep neural networks to extract embedding vectors containing information of both speaker identity and spoofing. The embedding vectors were subsequently taken as input features by back-end classifiers towards speaker and spoofing classification. The experimental results on both ASVspoof 2017 and ASVspoof 2019 showed that the proposed joint decision approach with triplet loss outperformed the corresponding baselines, a recent work on joint decision with Gaussian back-end fusion and our previous joint decision approach with cross-entropy loss.

show abstract

Deep Neural Network Embeddings with Gating Mechanisms for Text-Independent Speaker Verification

Cited by 18 publications

References 19 publications

Frequency and Temporal Convolutional Attention for Text-Independent Speaker Recognition

Frequency and Temporal Convolutional Attention for Text-Independent Speaker Recognition

Data augmentation versus noise compensation for x-vector speaker recognition systems in noisy environments

Joint Decision of Anti-Spoofing and Automatic Speaker Verification by Multi-Task Learning With Contrastive Loss

Contact Info

Product

Resources

About