A pitch synchronous feature extraction method for speaker recognition

Kim, Sun Ah; Eriksson, Thomas; Kang, Hong-Goo; Youn, Dae Hee

doi:10.1109/icassp.2004.1326008

Cited by 26 publications

(13 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Especially, it is crucial in speaker recognition systems since feature characteristics vary dynamically due to the difference of articulation on each phoneme. In addition, since the relative ratio and distinctiveness of each phoneme to recognition is not equivalent, a method to select relevant features among many candidates becomes an important research topic recently [3,4].…”

Section: Introductionmentioning

confidence: 99%

Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Jung

Kim

Kang

2009

2009 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

In this paper, an information theoretical approach to select features for speaker recognition systems is proposed. Conventional approaches having a fixed interval of analysis frames are not appropriate to represent dynamically varying characteristics of speech signals. To maximize the speakerrelated information varied by the characteristics of speech signals, we propose an information theory based feature selection method where features are selected to have minimum-redundancy with in selected features but maximumrelevancy to training speaker models.Experimental results verify that the proposed method reduces the error rates of speaker verification systems by 27.37 % in NIST 2002 database.

show abstract

Section: Introductionmentioning

confidence: 99%

Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Jung

Kim

Kang

2009

2009 IEEE International Conference on Acoustics, Speech and Signal Processing

View full text Add to dashboard Cite

show abstract

“…The order of static feature analysis is set to 12-15 plus various differential feature forms [4], [20] - [24], [3] and [5]. These values of analysis parameters are used without any argumentation or questioning, though it is evident that in some cases revised analysis parameters will have influence on speech recognition rate.…”

Section: Feature Extractionmentioning

confidence: 99%

“…The overlap is selected equal to half or one third of the analysis window length usually [20] - [22]. However, some original ideas for setting the frame shift (or the frame rate) are proposed.…”

Section: Speech Analysis Parametersmentioning

confidence: 99%

See 1 more Smart Citation

Improving Speech Recognition Rate through Analysis Parameters

Eringis

Tamulevičius

2014

Electrical, Control and Communication Engineering

View full text Add to dashboard Cite

-Speech signal is redundant and non-stationary by nature. Because of vocal tract inertness these variations are not very rapid and the signal can be considered as stationary in short segments. It is presumed that in short-time magnitude spectrum the most distinct information of speech is contained. This is the main reason for speech signal analysis in frame-by-frame manner. The analyzed speech signal is segmented into overlapping segments (so-called frames) for this purpose. Segments of 15-25 ms with the overlap of 10-15 ms are used usually.In this paper we present results of our investigation of analysis window length and frame shift influence on speech recognition rate. We have analyzed three different cepstral analysis approaches for this purpose: mel frequency cepstral analysis (MFCC), linear prediction cepstral analysis (LPCC) and perceptual linear prediction cepstral analysis (PLPC). The highest speech recognition rate was obtained using 10 ms length analysis window with the frame shift varying from 7.5 to 10 ms (regardless of analysis type). The highest increase of recognition rate was 2.5 %.

show abstract

“…The use of pitch-synchronous features has also been investigated for speaker recognition. Voice source information, as manifest in the pitch, is a speaker-specific characteristic, and source features derived from a pitch-synchronous analysis have been proposed as features for speaker recognition [16], [17]. Zilca et al [18] proposed a pitch-adaptive analysis, referred to as "depitching", which attempts to filter out pitch information from the speech signal.…”

Section: Pitch-synchronous Analysismentioning

confidence: 99%

Combining Spectral Representations for Large-Vocabulary Continuous Speech Recognition

Garau

Renals

2008

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Abstract-In this paper we investigate the combination of complementary acoustic feature streams in large vocabulary continuous speech recognition (LVCSR). We have explored the use of acoustic features obtained using a pitch-synchronous analysis, STRAIGHT, in combination with conventional features such as mel frequency cepstral coefficients. Pitch-synchronous acoustic features are of particular interest when used with vocal tract length normalisation (VTLN) which is known to be affected by the fundamental frequency. We have combined these spectral representations directly at the acoustic feature level using heteroscedastic linear discriminant analysis (HLDA) and at the system level using ROVER.We evaluated this approach on three LVCSR tasks: dictated newspaper text (WSJCAM0), conversational telephone speech (CTS), and multiparty meeting transcription. The CTS and meeting transcription experiments were both evaluated using standard NIST test sets and evaluation protocols. Our results indicate that combining conventional and pitch-synchronous acoustic feature sets using HLDA results in a consistent, significant decrease in word error rate across all three tasks. Combining at the system level using ROVER resulted in a further significant decrease in word error rate.

show abstract

A pitch synchronous feature extraction method for speaker recognition

Cited by 26 publications

References 5 publications

Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Normalized minimum-redundancy and maximum-relevancy based feature selection for speaker verification systems

Improving Speech Recognition Rate through Analysis Parameters

Combining Spectral Representations for Large-Vocabulary Continuous Speech Recognition

Contact Info

Product

Resources

About