Autoencoder-Based Semi-Supervised Curriculum Learning for Out-of-Domain Speaker Verification

Zheng, Siqi; Liu, Gang; Suo, Hongbin; Lei, Yun

doi:10.21437/interspeech.2019-1440

Cited by 13 publications

(8 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The demographic structure of the data from online smart speakers provides a possible explanation for this improvement. A preliminary investigation on a random sample of 100 online devices suggests at least 70% of the data has high speaker labeling quality [12].…”

Section: Experiments and Resultsmentioning

confidence: 99%

Towards a Fault-Tolerant Speaker Verification System: A Regularization Approach to Reduce the Condition Number

Zheng¹,

Liu²,

Suo³

et al. 2019

Interspeech 2019

Self Cite

View full text Add to dashboard Cite

Large-scale deployment of speech interaction devices makes it possible to harvest tremendous data quickly, which also introduces the problem of wrong labeling during data mining. Mislabeled training data has a substantial negative effect on the performance of speaker verification system. This study aims to enhance the generalization ability and robustness of the model when the training data is contaminated by wrong labels. Several regularization approaches are proposed to reduce the condition number of the speaker verification problem, making the model less sensitive to errors in the inputs. They are validated on both NIST SRE corpus and far-field smart speaker data. The results suggest that the performance deterioration caused by mislabeled training data can be significantly ameliorated by proper regularization.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Towards a Fault-Tolerant Speaker Verification System: A Regularization Approach to Reduce the Condition Number

Zheng¹,

Liu²,

Suo³

et al. 2019

Interspeech 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are some approaches that takes advantage of large scale unlabelled training data. Curriculum learning is one of them (Marchi et al, 2018;Ranjan and Hansen, 2018;Zheng et al, 2019). It starts by learning certain DNN model using a labeled corpus and continuously introduces unlabeled, out-of-domain text independent speaker samples.…”

Section: Unlabeled Datamentioning

confidence: 99%

“…It starts by learning certain DNN model using a labeled corpus and continuously introduces unlabeled, out-of-domain text independent speaker samples. Both LSTM (Marchi et al, 2018) and TDNN (Zheng et al, 2019) based systems are proposed that outperform baseline methods.…”

Section: Unlabeled Datamentioning

confidence: 99%

Deep Learning Methods in Speaker Recognition: A Review

Schiferl

Szaszák

Beke

2021

Period. Polytech. Elec. Eng. Comp. Sci.

View full text Add to dashboard Cite

This paper reviews the applied Deep Learning (DL) practices in the field of Speaker Recognition (SR), both in verification and identification. Speaker Recognition has been a widely used topic of speech technology. Many research works have been carried out and little progress has been achieved in the past 5–6 years. However, as Deep Learning techniques do advance in most machine learning fields, the former state-of-the-art methods are getting replaced by them in Speaker Recognition too. It seems that Deep Learning becomes the now state-of-the-art solution for both Speaker Verification (SV) and identification. The standard x-vectors, additional to i-vectors, are used as baseline in most of the novel works. The increasing amount of gathered data opens up the territory to Deep Learning, where they are the most effective.

show abstract

“…Semi-supervised learning utilizes limited labeled data in tandem with abundant unlabeled data to obtain higher performances. Stronger performance has been observed when unlabeled data is incorporated [6] [7] [8].…”

Section: Introductionmentioning

confidence: 99%

Graph Convolutional Network Based Semi-Supervised Learning on Multi-Speaker Meeting Data

Tong

Zheng

Zhang

et al. 2022

ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Unsupervised clustering on speakers is becoming increasingly important for its potential uses in semi-supervised learning. In reality, we are often presented with enormous amounts of unlabeled data from multi-party meetings and discussions. An effective unsupervised clustering approach would allow us to significantly increase the amount of training data without additional costs for annotations. Recently, methods based on graph convolutional networks (GCN) have received growing attention for unsupervised clustering, as these methods exploit the connectivity patterns between nodes to improve learning performance. In this work, we present a GCNbased approach for semi-supervised learning. Given a pre-trained embedding extractor, a graph convolutional network is trained on the labeled data and clusters unlabeled data with "pseudo-labels". We present a self-correcting training mechanism that iteratively runs the cluster-train-correct process on pseudo-labels. We show that this proposed approach effectively uses unlabeled data and improves speaker recognition accuracy.

show abstract

Autoencoder-Based Semi-Supervised Curriculum Learning for Out-of-Domain Speaker Verification

Cited by 13 publications

References 0 publications

Towards a Fault-Tolerant Speaker Verification System: A Regularization Approach to Reduce the Condition Number

Towards a Fault-Tolerant Speaker Verification System: A Regularization Approach to Reduce the Condition Number

Deep Learning Methods in Speaker Recognition: A Review

Graph Convolutional Network Based Semi-Supervised Learning on Multi-Speaker Meeting Data

Contact Info

Product

Resources

About