An analysis and comparative evaluation of MFCC variants for speaker identification over VoIP networks

Bouziane, Ayoub; Kharroubi, Jamal; Zarghili, Arsalane

doi:10.1109/wcitca.2015.7367018

Cited by 8 publications

(8 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The emphasizing step is firstly performed using a simple first order digital filter with transfer function H(z)=1-0.95z. Next, the emphasized speech signal is blocked into Hamming-windowed frames of 25 ms (400 samples) in length with 10 ms (160 samples) overlap between any two adjacent frames [5,48,49]. As regards the GFCCs, the features are extracted using a filter bank of 64 Gammatone filters and a down sampling frequency of 100 Hz (yielding frame rate of 10 ms), as recommended by [36].…”

Section: Experiments Results and Discussion 41 The Experimental Protocolmentioning

confidence: 99%

“…Instead of being used by themselves, the linear prediction coefficients were transformed into a set of robust, less correlated features such as, the linear prediction cepstral coefficients "LPCCs" [9], the perceptual linear prediction coefficients "PLP" [10], the perceptual linear predictive cepstral coefficients "PLPCC" [11][12][13][14][15] and the line spectral frequencies "LSF" [16] etc. In early 1980s, the so-called Mel-frequency cepstral coefficients "MFCCs" were introduced and yielded the best results compared to contemporary used features for speaker recognition [5,11]. One year later, the concept of dynamic features has been introduced to incorporate some temporal information to the extracted features [16,17].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Typically, high-level features are often used just as complementary features to the low-level features. In this study, we focus only on the low-level features (physical features) [5].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards an objective comparison of feature extraction techniques for automatic speaker recognition systems

2021

View full text Add to dashboard Cite

A common limitation of the previous comparative studies on speaker-features extraction techniques lies in the fact that the comparison is done independently of the used speaker modeling technique and its parameters. The aim of the present paper is twofold. Firstly, it aims to review the most significant advancements in feature extraction techniques used for automatic speaker recognition. Secondly, it seeks to evaluate and compare the currently dominant ones using an objective comparison methodology that overcomes the various limitations and drawbacks of the previous comparative studies. The results of the carried out experiments underlines the importance of the proposed comparison methodology.

show abstract

Section: Experiments Results and Discussion 41 The Experimental Protocolmentioning

confidence: 99%

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Towards an objective comparison of feature extraction techniques for automatic speaker recognition systems

2021

View full text Add to dashboard Cite

show abstract

“…Some studies have realized speaker identification [20], [23], [30]- [44] by comparing the voice of a speaker with the voice of a pre-registered person. Speaker identification has been applied to video conferences [45], criminal investigations [46], and television programs [47].…”

Section: A Speaker Recognition Using Stationary Devicementioning

confidence: 99%

Speaker Identification for Business-Card-Type Sensors

Yamaguchi

Oshima

et al. 2021

IEEE Open J. Comput. Soc.

View full text Add to dashboard Cite

“…Next, the emphasized speech signal is blocked into Hamming-windowed frames of 25 ms (400 samples) in length with 10 ms (160 samples) overlap between any two adjacent frames.Finally, 19 MelFrequency Cepstral Coefficients were extracted from each frame [15]. During the training phase, a universal background model (UBM) of 1024 Gaussian components was trained on the overall training data (7.5 hours of speech) using the EM algorithm.…”

Section: The Experimental Protocolmentioning

confidence: 99%

Towards an Optimal Speaker Modeling in Speaker Verification Systems using Personalized Background Models

Bouziane¹,

Kharroubi²,

Zarghili³

2017

IJECE

View full text Add to dashboard Cite

This paper presents a novel speaker modeling approachfor speaker recognition systems. The basic idea of this approach consists of deriving the target speaker model from a personalized background model, composed only of the UBM Gaussian components which are really present in the speech of the target speaker. The motivation behind the derivation of speakers' models from personalized background models is to exploit the observeddifference insome acoustic-classes between speakers, in order to improve the performance of speaker recognition systems. The proposed approach was evaluatedfor speaker verification task using various amounts of training and testing speech data. The experimental results showed that the proposed approach is efficientin termsof both verification performance and computational cost during the testing phase of the system, compared to the traditional UBM based speaker recognition systems. Keyword:Speaker modeling Speaker verification Gaussian mixture models Universal background model Maximum a posteriori (MAP) Personalized background models

show abstract

An analysis and comparative evaluation of MFCC variants for speaker identification over VoIP networks

Cited by 8 publications

References 16 publications

Towards an objective comparison of feature extraction techniques for automatic speaker recognition systems

Towards an objective comparison of feature extraction techniques for automatic speaker recognition systems

Speaker Identification for Business-Card-Type Sensors

Towards an Optimal Speaker Modeling in Speaker Verification Systems using Personalized Background Models

Contact Info

Product

Resources

About