Text-independent speaker recognition based on the Hurst parameter and the multidimensional fractional Brownian motion model

Santana, Ricardo; Coelho, R.; Alcaim, Abraham

doi:10.1109/tsa.2005.858054

Cited by 46 publications

(48 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Two different vector-based distances were employed in the feature space: the classic Euclidean distance and the Minkowsky distance. We let the parameter δ vary from 2 to 25, whereas the dimensionality of the feature vector was analyzed in the range [2][3][4][5][6][7][8][9][10]. The results are graphically shown in Fig.…”

Section: Resultsmentioning

confidence: 99%

“…The Hurst exponent H has a clear interpretation: if H = 1/2, then the fBm represents the standard Brownian motion (namely the process has no memory); for H > 1/2, the process is characterized by a persistence (positive or negative) -namely the process has a clear trend; finally, for H < 1/2 the process is characterized by an antipersistence. Even if this method has been largely applied in the finance community (mainly to understand some characteristics of the financial series), some works applying the fBm function have also been presented in the pattern recognition context, like in medical imaging [8] or in speech recognition [9]. It is important to note that fBm is related to the fractal theory [10] -a theory largely employed in image analysis [11] or even object classification [12,13] -since it is modeling self-similarity.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

2D Shape Classification Using Multifractional Brownian Motion

Bicego

Trudda

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In this paper a novel approach to contour-based 2D shape recognition is proposed. The main idea is to characterize the contour of an object using the multifractional Brownian motion (mBm), a mathematical method able to capture the local self similarity and long-range dependence of a signal. The mBm estimation results in a sequence of Hurst coefficients, which we used to derive a fixed size feature vector. Preliminary experimental evaluations using simple classifiers with these feature vectors produce encouraging results, also in comparison with the state of the art.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

2D Shape Classification Using Multifractional Brownian Motion

Bicego

Trudda

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Daubechies [25]), which was computed by applying a wavelet-based multidimensional transformation of the short-time input speech in the work by Sant'Ana [124]. Thanks to its statistical definition, pH is robust to channel distortions as it models the stochastic behavior of input speech signal (see Zao et al [125], or Palo et al [126]).…”

Section: Wavelet-based Physical Featuresmentioning

confidence: 99%

“…Thanks to its statistical definition, pH is robust to channel distortions as it models the stochastic behavior of input speech signal (see Zao et al [125], or Palo et al [126]). pH was originally applied as a means to improve speech-related problems, such as text-independent speaker recognition [124], speech emotion classification [125,126], or speech enhancement [127]. However, it has also been applied to sound source localization in noisy environments recently, as in the work by Dranka and Coelho [128].…”

Section: Wavelet-based Physical Featuresmentioning

confidence: 99%

A Review of Physical and Perceptual Feature Extraction Techniques for Speech, Music and Environmental Sounds

2016

View full text Add to dashboard Cite

Endowing machines with sensing capabilities similar to those of humans is a prevalent quest in engineering and computer science. In the pursuit of making computers sense their surroundings, a huge effort has been conducted to allow machines and computers to acquire, process, analyze and understand their environment in a human-like way. Focusing on the sense of hearing, the ability of computers to sense their acoustic environment as humans do goes by the name of machine hearing. To achieve this ambitious aim, the representation of the audio signal is of paramount importance. In this paper, we present an up-to-date review of the most relevant audio feature extraction techniques developed to analyze the most usual audio signals: speech, music and environmental sounds. Besides revisiting classic approaches for completeness, we include the latest advances in the field based on new domains of analysis together with novel bio-inspired proposals. These approaches are described following a taxonomy that organizes them according to their physical or perceptual basis, being subsequently divided depending on the domain of computation (time, frequency, wavelet, image-based, cepstral, or other domains). The description of the approaches is accompanied with recent examples of their application to machine hearing related problems.

show abstract

“…The goal of the multi-style training is to reduce the mismatch between training and test features by corrupting the speech signals. Features based on discrete wavelet transform, and pH features [94] are fused with the MFCCs. Speech enhancement techniques are also applied.…”

Section: Robust Features Against Additive Noisementioning

confidence: 99%

A Review on Feature Extraction for Speaker Recognition under Degraded Conditions

DİŞKEN

Tüfekçi

Sarıbulut

et al. 2016

IETE Technical Review

View full text Add to dashboard Cite

Speech is a signal that includes speaker's emotion, characteristic specification, phonemeinformation etc. Various methods have been proposed for speaker recognition by extracting specifications of a given utterance. Among them, short-term cepstral features are used excessively in speech, and speaker recognition areas because of their low complexity, and high performance in controlled environments. On the other hand, their performances decrease dramatically under degraded conditions such as channel mismatch, additive noise, emotional variability, etc. In this paper, a literature review on speaker-specific information extraction from speech is presented by considering the latest studies offering solutions to the aforementioned problem. The studies are categorized in three groups considering their robustness against channel mismatch, additive noise, and other degradations such as vocal effort, emotion mismatch, etc. For a more understandable representation, they are also classified into two tables by utilizing their classification methods, and used data-sets.

show abstract

Text-independent speaker recognition based on the Hurst parameter and the multidimensional fractional Brownian motion model

Cited by 46 publications

References 24 publications

2D Shape Classification Using Multifractional Brownian Motion

2D Shape Classification Using Multifractional Brownian Motion

A Review of Physical and Perceptual Feature Extraction Techniques for Speech, Music and Environmental Sounds

A Review on Feature Extraction for Speaker Recognition under Degraded Conditions

Contact Info

Product

Resources

About