Deep Normalization for Speaker Vectors

Cai, Yunqi; Li, Lantian; Abel, Andrew; Zhu, Xiaoyan; Wang, Dong

doi:10.1109/taslp.2020.3039573

Cited by 26 publications

(13 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This contradiction between experimental observations and theoretical expectation deserves thoughtful investigations on PLDA. In [7][8][9], Cai et al argued that the problem should have arise from the neural speaker embeddings. It is noted that embeddings extracted from neural networks tend to be non-Gaussian for individual speakers and the distributions across different speakers are non-homogeneous.…”

Section: Introductionmentioning

confidence: 99%

“…These irregular distributions cause the performance degradation of verification systems with the PLDA back-end. In relation to this perspective, a series of regularization approaches have been proposed to force the neural embeddings to be homogeneously Gaussian distributed, e.g., Gaussian-constrained loss [7], variational auto-encoder [8] and discriminative normalization flow [9,10].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unifying Cosine and PLDA Back-ends for Speaker Verification

Peng¹,

He²,

Ding³

et al. 2022

Preprint

View full text Add to dashboard Cite

State-of-art speaker verification (SV) systems use a backend model to score the similarity of speaker embeddings extracted from a neural network model. The commonly used back-end models are the cosine scoring and the probabilistic linear discriminant analysis (PLDA) scoring. With the recently developed neural embeddings, the theoretically more appealing PLDA approach is found to have no advantage against or even be inferior the simple cosine scoring in terms of SV system performance. This paper presents an investigation on the relation between the two scoring approaches, aiming to explain the above counter-intuitive observation. It is shown that the cosine scoring is essentially a special case of PLDA scoring. In other words, by properly setting the parameters of PLDA, the two back-ends become equivalent. As a consequence, the cosine scoring not only inherits the basic assumptions for the PLDA but also introduces additional assumptions on the properties of input embeddings. Experiments show that the dimensional independence assumption required by the cosine scoring contributes most to the performance gap between the two methods under the domain-matched condition. When there is severe domain mismatch and the dimensional independence assumption does not hold, the PLDA would perform better than the cosine for domain adaptation.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Unifying Cosine and PLDA Back-ends for Speaker Verification

Peng¹,

He²,

Ding³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Probabilistic Linear Discriminant Analysis (PLDA) [3]- [6] is commonly used as a back-end scoring model. To satisfy the PLDA Gaussian assumptions for training data [7], extracted features generally require pre-processing, including Linear Discriminant Analysis (LDA) and length regularization [8], before being used to train a PLDA model. Both LDA and PLDA are trained in a supervised manner that requires training data with corresponding speaker labels.…”

Section: Introductionmentioning

confidence: 99%

When Speaker Recognition Meets Noisy Labels: Optimizations for Front-ends and Back-ends

Li¹,

Tong²,

Hong³

2021

Preprint

View full text Add to dashboard Cite

A typical speaker recognition system often involves two modules: a feature extractor front-end and a speaker identity back-end. Despite the superior performance that deep neural networks have achieved for the front-end, their success benefits from the availability of large-scale and correctly labeled datasets. While label noise is unavoidable in speaker recognition datasets, both the front-end and back-end are affected by label noise, which degrades the speaker recognition performance. In this paper, we first conduct comprehensive experiments to help improve the understanding of the effects of label noise on both the front-end and back-end. Then, we propose a simple yet effective training paradigm and loss correction method to handle label noise for the front-end. We combine our proposed method with the recently proposed Bayesian estimation of PLDA for noisy labels, and the whole system shows strong robustness to label noise. Furthermore, we show two practical applications of the improved system: one application corrects noisy labels based on an utterance’s chunk-level predictions, and the other algorithmically filters out high-confidence noisy samples within a dataset. By applying the second application to the NIST SRE0410 dataset and verifying filtered utterances by human validation, we identify that approximately 1% of the SRE04-10 dataset is made up of label errors.<br>

show abstract

“…This can be conducted either globally or class-dependently. The former is represented by the famous length normalization [16] and the latter is represented by the deep normalization model based on normalization flows [17].…”

Section: Introductionmentioning

confidence: 99%

Real Additive Margin Softmax for Speaker Verification

Li¹,

Nai²,

Wang³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

The additive margin softmax (AM-Softmax) loss has delivered remarkable performance in speaker verification. A supposed behavior of AM-Softmax is that it can shrink within-class variation by putting emphasis on target logits, which in turn improves margin between target and non-target classes. In this paper, we conduct a careful analysis on the behavior of AM-Softmax loss, and show that this loss does not implement real max-margin training. Based on this observation, we present a Real AM-Softmax loss which involves a true margin function in the softmax training. Experiments conducted on VoxCeleb1, SITW and CNCeleb demonstrated that the corrected AM-Softmax loss consistently outperforms the original one. The code has been released at https://gitlab.com/csltstu/sunine.

show abstract

Deep Normalization for Speaker Vectors

Cited by 26 publications

References 47 publications

Unifying Cosine and PLDA Back-ends for Speaker Verification

Unifying Cosine and PLDA Back-ends for Speaker Verification

When Speaker Recognition Meets Noisy Labels: Optimizations for Front-ends and Back-ends

Real Additive Margin Softmax for Speaker Verification

Contact Info

Product

Resources

About