• Corpus ID: 212538813

Feature Extraction and Classification for Automatic Speaker Recognition System – A Review

@inproceedings{Kaur2015FeatureEA,
  title={Feature Extraction and Classification for Automatic Speaker Recognition System – A Review},
  author={Kirandeep Kaur and Neelu Jain},
  year={2015}
}
Automatic speaker recognition (ASR) has found immense applications in the industries like banking, security, forensics etc. for its advantages such as easy implementation, more secure, more user friendly. To have a good recognition rate is a pre-requisite for any ASR system which can be achieved by making an optimal choice among the available techniques for ASR. In this paper, different techniques for the system have been discussed such as MFCC, LPCC, LPC, Wavelet decomposition for feature… 

Figures and Tables from this paper

Comparative Analysis on Different Cepstral Features for Speaker Identification Recognition
  • R. Hanifa, I. K., M. S.
  • Computer Science
    2020 IEEE Student Conference on Research and Development (SCOReD)
  • 2020
TLDR
The comparative analysis of the speaker identification system revealed that a combination of GFCC and pitch as the feature vectors (Model 4) produced the highest accuracy rate of 86.1%.
A Brief Review: Voice Biometric For Speaker Verification in Attendance Systems
TLDR
Different steps which are required in voice dependent attendance system are briefed and various classification algorithms has been briefed which can be used for testing and training purposes.
A Review on Basic of Voice Recognition
TLDR
Voice sample is observed with MFCC for extracting features and then used to train Artificial Neural for classification, which presents the main paradigms for speaker recognition.
A Survey on Speech Recognition Algorithms
TLDR
An overview of different techniques that can be used in application of speaker recognition such as MFCC, LPC, and LPCC for feature extraction and VQ, SVM, HMM; GMM for feature classification is delivered.
FEATURE EXTRACTION AND CLASSIFICATION TECHNIQUES OF AUTOMATIC SPEECH RECOGNITION SYSTEM: A REVIEW
TLDR
This paper described about various ASR algorithms and feature extraction techniques and their comparison with each other in this paper and suggested techniques like LPC, MFCCs, RASTA filtering, PLDA, LDA, PCA and algorithms like DTW, VQ, SVM, GMM, HMM.
Performance Improvement of Speaker Recognition by MCE-based Score Combination of Multiple Feature Parameters
In this thesis, an enhanced method for the feature extraction of vocal source signals and score combination using an MCE-Based weight estimation of the score of multiple feature vectors are proposed
An Analysis of an Isolated Assamese Digit Recognition using MFCC and DTW
  • B. K. Deka, P. Das
  • Computer Science
    2019 6th International Conference on Computing for Sustainable Global Development (INDIACom)
  • 2019
TLDR
This paper presents an isolated Assamese digit recognition using MFC Coefficient and Dynamic Time Warping classifier and shows the average accuracy rate based on three-dimensional coefficients of MFC are compared.
Speaker Identification based on Hybrid Feature Extraction Techniques
TLDR
This work is concerned with studying the effect of appropriate extracted features from various levels of discrete wavelet transformation and the concatenation of two techniques (discrete wavelet and curvelet transform and principal component analysis (PCA) on speaker identification.
Different language recognition model
  • Sankirna D. Joge, A. Shirsat
  • Computer Science, Linguistics
    2016 International Conference on Automatic Control and Dynamic Optimization Techniques (ICACDOT)
  • 2016
TLDR
The main focus is recognition of language which is spoken by a person, and Pc as a hardware with Matlab software is used to extract features of sentence spoken so as to recognize a language.
...
1
2
...

References

SHOWING 1-10 OF 22 REFERENCES
Experimental evaluation of features for robust speaker identification
  • D. Reynolds
  • Computer Science
    IEEE Trans. Speech Audio Process.
  • 1994
TLDR
This correspondence presents an experimental evaluation of different features and channel compensation techniques for robust speaker identification, and it is shown that performance differences between the basic features is small, and the major gains are due to the channel Compensation techniques.
Wavelet transform based automatic speaker recognition
TLDR
An effective feature extraction technique for speaker recognition that uses multi resolution property of wavelet transform and Mel-Frequency Cepstral Coefficients (MFCCs) for analyzing the speech signal.
GMM-UBM for text-dependent speaker recognition
TLDR
The experimental results showed the performance of GMM-UBM exceeded that of DTW, and that of frame-based outperformed that of wave-based, and a new UBM initialization method, LBG-VQ-EM, will be proposed.
GMM Based Speaker Recognition on Readily Available Databases
TLDR
An overview of the most popular databases used in speaker recognition evaluation is given and the means of training and testing an ASR system using them are outlined.
Linear versus mel frequency cepstral coefficients for speaker recognition
TLDR
Comparing the performances between MFCC and LFCC in the NIST SRE (Speaker Recognition Evaluation) 2010 extended-core task shows that LFCC consistently outperforms MFCC, mainly due to its better performance in the female trials, and shows some advantage of LFCC over MFCC in reverberant speech.
Hybrid of wavelet and MFCC features for speaker verification
  • Pawan Kumar, M. Chandra
  • Computer Science
    2011 World Congress on Information and Communication Technologies
  • 2011
TLDR
Wavelet Based Mel Frequency Cepstral Coefficient features are proposed for speaker verification and have shown an increment of 1.18% in performance over MFCC features for text dependent speaker verification system.
A comparison of four techniques for automatic speaker recognition
TLDR
The results of this study indicate that LPC derived parameters perform better than do those derived from cepstral and spectral data.
A Novel Windowing Technique for Efficient Computation of MFCC for Speaker Recognition
TLDR
A novel family of windowing technique to compute mel frequency cepstral coefficient (MFCC) for automatic speaker recognition from speech based on fundamental property of discrete time Fourier transform related to differentiation in frequency domain is proposed.
Speaker Identification Wavelet Transform based method
TLDR
Two identification methods are presented based on Continuous Wavelet Transform CWT, the male and female identification method and the SIM, which showed a high recognition rate about 98%, which makes the speaker identification wavelet transform based method to be very important in digital security systems applications.
...
1
2
3
...