Spectro-temporal modulation energy based mask for robust speaker identification

Chi, Tai-Shih; Lin, Ting; Hsu, Chung Chien

doi:10.1121/1.3697534

Cited by 10 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [44], an algorithm which distinguishes speech from non-speech based on spectro-temporal modulation energies is proposed and evaluated in robust text-independent closed-set speaker identification simulations. An exemplar-based representation and sparse discrimination was proposed in [45] that outperformed the baseline GMM-universal background model (UBM) and HMM-based systems with a large margin.…”

Section: Performance Improvementmentioning

confidence: 99%

Speaker identification using multimodal neural networks and wavelet analysis

2015

View full text Add to dashboard Cite

The rapid momentum of the technology progress in the recent years has led to a tremendous rise in the use of biometric authentication systems. The objective of this research is to investigate the problem of identifying a speaker from its voice regardless of the content. In this study, the authors designed and implemented a novel text-independent multimodal speaker identification system based on wavelet analysis and neural networks. Wavelet analysis comprises discrete wavelet transform, wavelet packet transform, wavelet sub-band coding and Mel-frequency cepstral coefficients (MFCCs). The learning module comprises general regressive, probabilistic and radial basis function neural networks, forming decisions through a majority voting scheme. The system was found to be competitive and it improved the identification rate by 15% as compared with the classical MFCC. In addition, it reduced the identification time by 40% as compared with the back-propagation neural network, Gaussian mixture model and principal component analysis. Performance tests conducted using the GRID database corpora have shown that this approach has faster identification time and greater accuracy compared with traditional approaches, and it is applicable to real-time, text-independent speaker identification systems

show abstract

Section: Performance Improvementmentioning

confidence: 99%

Speaker identification using multimodal neural networks and wavelet analysis

2015

View full text Add to dashboard Cite

show abstract

“…The digitized speech samples are converted into the Mal-Frequency Cepstrum Coefficient (MFCC) to produce the features of the wave vectors, which can be then used in learning algorithms such as SVM, and ANN. Even though most of the known acoustic-property-based algorithms have achieved an almost 100% level of accuracy in an uncluttered environment [14], [15], the performance of these algorithms is degraded with channel variations due to the microphones, used or environmental or background distortions [14].…”

Section: Introductionmentioning

confidence: 99%

A Holy Quran Reader/Reciter Identification System Using Support Vector Machine

Nahar¹,

Al-Shannaq²,

Manasrah³

et al. 2019

IJMLC

View full text Add to dashboard Cite

Holy Quran Reader Identification is the process of identifying the reader or reciter of the Holy Quran based on several features in the corresponding acoustic wave. In this research, we build our own corpus, which contains 15 known readers of the Holy Quran. The Mel-Frequency Cepstrum Coefficients (MFCC) are used for the extraction of these features from the input acoustic signal. These MFCCs are the reader's features matrix, which is used for recognition via Support Vector Machine (SVM) and Artificial Neural Networks (ANN). According to our experimental results, the Holy Quran Reader Identification System identifies the reader with 96.59% accuracy when using SVM, in contrast to accuracy of 86.1% when using ANN.

show abstract

“…In general, the features from the speech signal for speaker recognition are extracted by modeling the human voice production system (such as linear prediction cepstral coefficient, LPCC [ 5 ]) or from the responses of human auditory system. Human listeners are capable of recognizing speakers in noisy environments, while most of the traditional speaker recognition systems do not perform well in the presence of noise [ 6 ]. Unlike traditional methods in which features are extracted from the properties of the acoustic signal, this study proposes a speaker identification technique using neural responses from a physiologically-based computational model of the auditory periphery.…”

Section: Introductionmentioning

confidence: 99%

“…The traditional MFCC-based system achieved almost 100% classification accuracy in clean condition [ 7 , 8 ]. However, the performance of these acoustic-property-based methods degrades substantially for speech signals under channel variations induced by the handset or microphones as well as for environmental or background distortions [ 6 ]. In recent years, efforts have been made aiming to extract features by removing the noise from the speaker characteristic information directly such as cepstral mean normalization [ 9 ], RASTA processing [ 10 ], warping methods [ 11 ], and robust parameterizations [ 12 ].…”

Section: Introductionmentioning

confidence: 99%

A Robust Speaker Identification System Using the Responses from a Model of the Auditory Periphery

et al. 2016

View full text Add to dashboard Cite

Speaker identification under noisy conditions is one of the challenging topics in the field of speech processing applications. Motivated by the fact that the neural responses are robust against noise, this paper proposes a new speaker identification system using 2-D neurograms constructed from the responses of a physiologically-based computational model of the auditory periphery. The responses of auditory-nerve fibers for a wide range of characteristic frequency were simulated to speech signals to construct neurograms. The neurogram coefficients were trained using the well-known Gaussian mixture model-universal background model classification technique to generate an identity model for each speaker. In this study, three text-independent and one text-dependent speaker databases were employed to test the identification performance of the proposed method. Also, the robustness of the proposed method was investigated using speech signals distorted by three types of noise such as the white Gaussian, pink, and street noises with different signal-to-noise ratios. The identification results of the proposed neural-response-based method were compared to the performances of the traditional speaker identification methods using features such as the Mel-frequency cepstral coefficients, Gamma-tone frequency cepstral coefficients and frequency domain linear prediction. Although the classification accuracy achieved by the proposed method was comparable to the performance of those traditional techniques in quiet, the new feature was found to provide lower error rates of classification under noisy environments.

show abstract

Spectro-temporal modulation energy based mask for robust speaker identification

Cited by 10 publications

References 18 publications

Speaker identification using multimodal neural networks and wavelet analysis

Speaker identification using multimodal neural networks and wavelet analysis

A Holy Quran Reader/Reciter Identification System Using Support Vector Machine

A Robust Speaker Identification System Using the Responses from a Model of the Auditory Periphery

Contact Info

Product

Resources

About