Towards Robust Speaker Verification with Target Speaker Enhancement

Zhang, Chunlei; Yu, Meng; Weng, Chao; Yu, Dong

doi:10.1109/icassp39728.2021.9414017

Cited by 18 publications

(12 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, this is beyond the scope of this study and is left for future work. Figures 4,5,6, and 7 illustrate the EERs of the PLDA and GLASSO-PLDA on the evaluation trials, where the enrollment utterances were clean, whereas the test utterances had one of six conditions (i.e., two noise types: bus and cafe × three SNRs: 0, 5, and 10 dB). The black dotted vertical In Part 1, the average relative EER reductions in the evaluation trials were 4.1450% (0.74% to 9.16%) with the d-vector (see the r-vector (see Table 2).…”

Section: Results and Discussion 1) Demonstration Of The Conditional I...mentioning

confidence: 99%

See 1 more Smart Citation

Reflection of Conditional Independence Structure to Noise Variability for Noise Robust Text Dependent Speaker Verification

Yoon

2022

IEEE Access

View full text Add to dashboard Cite

In the field of speaker verification (SV), the development of noise-robust systems is a challenge for their deployment in real-world environments. Noise variability compensation is a common strategy for increasing the robustness to noise variations. The performance of noise compensation depends on how well the noise variability, which is inherent in within-class variability, is estimated. However, to date, there is no information about the true noise variability that could reduce the gap between the empirical and true statistics. Most studies assume that true noise covariates are independent. This study aims to demonstrate the assumption that true noise variability has a conditional independence structure rather than an independence structure. This assumption was motivated by our previous findings, which revealed that optimal withinclass variability has a conditional independence structure in text-dependent speaker verification (TD-SV) in clean environments. This indicates that the optima of all the variabilities in within-class variability, except noise variability, has a conditional independence structure; however, it is unknown whether this is also true for optimal noise variability. Our assumption was supported by the experimental results obtained under noisy TD-SV trials using systems built with graphical least absolute shrinking and selection operator-based probabilistic linear discriminant analysis, which achieved up to 10% relative equal error rate improvements.INDEX TERMS Background noise, conditional independence, graphical least absolute shrinking and selection operator (GLASSO), probabilistic linear discriminant analysis (PLDA), text-dependent speaker verification.

show abstract

Section: Results and Discussion 1) Demonstration Of The Conditional I...mentioning

confidence: 99%

“…Recently, many studies have been conducted to develop various types of noise-robust ASV systems [4], [5], [6], [7], [8]. We address noise robustness based on probabilistic linear discriminant analysis (PLDA) [9], [10].…”

Section: Introductionmentioning

confidence: 99%

Reflection of Conditional Independence Structure to Noise Variability for Noise Robust Text Dependent Speaker Verification

Yoon

2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…For instance, visual voice activity detection [5] might alleviate this issue. However, it is more challenging with audio clues [57], and further research may be required. 21 The results of the CHiME 6 challenge can be found at: https:// chimechallenge.github.io/chime6/results.html.…”

Section: A Deployment Of Tse Systemsmentioning

confidence: 99%

Neural Target Speech Extraction: An Overview

Žmolíková¹,

Delcroix²,

Ochiai³

et al. 2023

Preprint

View full text Add to dashboard Cite

Humans can listen to a target speaker even in challenging acoustic conditions that have noise, reverberation, and interfering speakers. This phenomenon is known as the cocktail-party effect. For decades, researchers have focused on approaching the listening ability of humans. One critical issue is handling interfering speakers because the target and non-target speech signals share similar characteristics, complicating their discrimination. Target speech/speaker extraction (TSE) isolates the speech signal of a target speaker from a mixture of several speakers with or without noises and reverberations using clues that identify the speaker in the mixture. Such clues might be a spatial clue indicating the direction of the target speaker, a video of the speaker's lips, or a pre-recorded enrollment utterance from which their voice characteristics can be derived. TSE is an emerging field of research that has received increased attention in recent years because it offers a practical approach to the cocktailparty problem and involves such aspects of signal processing as audio, visual, array processing, and deep learning. This paper focuses on recent neural-based approaches and presents an in-depth overview of TSE. We guide readers through the different major approaches, emphasizing the similarities among frameworks and discussing potential future directions.

show abstract

“…The first one is the Voxceleb corpus, which is our primary benchmark dataset to evaluate the proposed systems. The second one is our internal SV dataset, the test set is collected under the unconstrained vehicle environments in Mandarin (noted as "vehicle-spk") [7], the details of the corpora are listed below. Voxceleb: we use Voxceleb1 and Voxceleb2 training data for training the models.…”

Section: C3-dino Speaker Embedding Systemmentioning

confidence: 99%

C3-DINO: Joint Contrastive and Non-Contrastive Self-Supervised Learning for Speaker Verification

Zhang

2022

IEEE J. Sel. Top. Signal Process.

Self Cite

View full text Add to dashboard Cite

Self-supervised learning (SSL) has drawn an increased attention in the field of speech processing. Recent studies have demonstrated that contrastive learning is able to learn discriminative speaker embeddings in a self-supervised manner. However, base contrastive self-supervised learning (CSSL) assumes that the pairs generated from a view of anchor instance and any view of other instances are all negative, which introduces many false negative pairs in constructing the loss function. The problem is referred as class-collision, which remains as one major issue that impedes the CSSL based speaker verification (SV) systems from achieving better performances. In the meanwhile, studies reveal that negative sample free SSL frameworks perform well in learning speaker or image representations. In this study, we investigate SSL techniques that lead to an improved SV performance. We first analyse the impact of false negative pairs in the CSSL systems. Then, a multi-stage Class-Collision Correction (C3) method is proposed (e.g., false negative pair filtering, InfoNCE re-weighting and ProtoNCE), which leads to the stateof-the-art (SOTA) CSSL based speaker embedding system. On the basis of the pretrained CSSL model, we further propose to employ a negative sample free SSL objective (i.e., DINO) to fine-tune the speaker embedding network. The resulting speaker embedding system (C3-DINO) achieves 2.5% Equal Error Rate (EER) with a simple Cosine Distance Scoring (CDS) method on Voxceleb1 test set, which outperforms the previous SOTA SSL system (4.86%) by a significant +45% relative improvement. With speaker clustering and pseudo labeling on Voxceleb2 training set, a LDA/CDS back-end applying on the C3-DINO speaker embeddings is able to further push the EER to 2.2%. Comprehensive experimental investigations of the Voxceleb benchmarks and our internal dataset demonstrate the effectiveness of our proposed methods, and the performance gap between the SSL SV and the supervised counterpart narrows further.

show abstract

Towards Robust Speaker Verification with Target Speaker Enhancement

Cited by 18 publications

References 28 publications

Reflection of Conditional Independence Structure to Noise Variability for Noise Robust Text Dependent Speaker Verification

Reflection of Conditional Independence Structure to Noise Variability for Noise Robust Text Dependent Speaker Verification

Neural Target Speech Extraction: An Overview

C3-DINO: Joint Contrastive and Non-Contrastive Self-Supervised Learning for Speaker Verification

Contact Info

Product

Resources

About