Scores Calibration in Speaker Recognition Systems

Shulipa, Andrey; Novoselov, Sergey; Matveev, Yuri

doi:10.1007/978-3-319-43958-7_72

Cited by 16 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Observing that adverse acoustic conditions and duration variability in utterances could have detrimental effect on PLDA scores, a number of score calibration methods have been proposed to compensated for the effect by modeling it as a shift in the PLDA scores. While some of these methods only compensate for the duration mismatch between the i-vector pair during PLDA scoring [17,18,19], there are techniques also taking the SNR mismatch into account [20,21]. In [22], the shift is assumed to follow a Gaussian distribution with mean and variance dependent on the speech quality.…”

Section: Introductionmentioning

confidence: 99%

i-Vector DNN Scoring and Calibration for Noise Robust Speaker Verification

Tan

Mak

2017

Interspeech 2017

View full text Add to dashboard Cite

This paper proposes applying multi-task learning to train deep neural networks (DNNs) for calibrating the PLDA scores of speaker verification systems under noisy environments. To facilitate the DNNs to learn the main task (calibration), several auxiliary tasks were introduced, including the prediction of SNR and duration from i-vectors and classifying whether an i-vector pair belongs to the same speaker or not. The possibility of replacing the PLDA model by a DNN during the scoring stage is also explored. Evaluations on noise contaminated speech suggest that the auxiliary tasks are important for the DNNs to learn the main calibration task and that the uncalibrated PLDA scores are an essential input to the DNNs. Without this input, the DNNs can only predict the score shifts accurately, suggesting that the PLDA model is indispensable.

show abstract

Section: Introductionmentioning

confidence: 99%

i-Vector DNN Scoring and Calibration for Noise Robust Speaker Verification

Tan

Mak

2017

Interspeech 2017

View full text Add to dashboard Cite

show abstract

“…First of all, we propose that adding information about the duration of the utterances as an input to the deep neural network should help to improve the calibration. According to [9] we used the logarithm of duration.…”

Section: Durationmentioning

confidence: 99%

“…We evaluate speaker recognition performance in terms of Equal Error Rates (EER) and minimum detection cost functions with Ptar = 0.05: C 0.05 min [9,22]. To estimate calibration performance of the systems at this point we use conventional actual detection cost function C 0.05 act [22].…”

Section: Evaluation Metricsmentioning

confidence: 99%

See 1 more Smart Citation

Investigation of Different Calibration Methods for Deep Speaker Embedding based Verification Systems

Lavrentyeva¹,

Novoselov²,

Shulipa³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Deep speaker embedding extractors have already become new state-of-the-art systems in the speaker verification field. However, the problem of verification score calibration for such systems often remains out of focus. An irrelevant score calibration leads to serious issues, especially in the case of unknown acoustic conditions, even if we use a strong speaker verification system in terms of threshold-free metrics.This paper presents an investigation over several methods of score calibration: a classical approach based on the logistic regression model; the recently presented magnitude estimation network MagnetO that uses activations from the pooling layer of the trained deep speaker extractor and generalization of such approach based on separate scale and offset prediction neural networks.An additional focus of this research is to estimate the impact of score normalization on the calibration performance of the system.The obtained results demonstrate that there are no serious problems if in-domain development data are used for calibration tuning. Otherwise, a trade-off between good calibration performance and threshold-free system quality arises. In most cases using adaptive s-norm helps to stabilize score distributions and to improve system performance.Meanwhile, some experiments demonstrate that novel approaches have their limits in score stabilization on several datasets.

show abstract

“…Observing that adverse acoustic conditions and duration variability in utterances could have detrimental effect on PLDA scores, researchers explored the potential of other backends to replace the PLDA models, e.g., support vector machines (SVMs) [15] or even end-to-end learning [16]. Besides, a number of score calibration methods [17]- [19] have been proposed to compensate for the detrimental effect on the PLDA scores. While many of these methods can compensate for the duration mismatch only, there are techniques also take the SNR mismatch into account [20]- [23].…”

Section: Introductionmentioning

confidence: 99%

DNN-Based Score Calibration With Multitask Learning for Noise Robust Speaker Verification

Tan

Mak

2018

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

This paper proposes and investigates several deep neural network (DNN)-based score compensation, transformation and calibration algorithms for enhancing the noise robustness of i-vector speaker verification systems. Unlike conventional calibration methods where the required score shift is a linear function of SNR or log-duration, the DNN approach learns the complex relationship between the score shifts and the combination of i-vector pairs and uncalibrated scores. Furthermore, with the flexibility of DNNs, it is possible to explicitly train a DNN to recover the clean scores without having to estimate the score shifts. To alleviate the overfitting problem, multi-task learning is applied to incorporate auxiliary information such as SNRs and speaker ID of training utterances into the DNN. Experiments on NIST 2012 SRE show that score calibration derived from multi-task DNNs can improve the performance of the conventional score-shift approch significantly, especially under noisy conditions.

show abstract

Scores Calibration in Speaker Recognition Systems

Cited by 16 publications

References 9 publications

i-Vector DNN Scoring and Calibration for Noise Robust Speaker Verification

i-Vector DNN Scoring and Calibration for Noise Robust Speaker Verification

Investigation of Different Calibration Methods for Deep Speaker Embedding based Verification Systems

DNN-Based Score Calibration With Multitask Learning for Noise Robust Speaker Verification

Contact Info

Product

Resources

About