Integrated system for prosodic features detection from speech

Author

Zbancioc, M.D. ; Feraru, M.

Author_Institution

Inst. of Comput. Sci., Tech. Univ. “Gheorghe Asachi” of Iasi, Iasi, Romania

fYear

2014

fDate

16-18 Oct. 2014

Firstpage

114

Lastpage

117

Abstract

The paper presents the instruments implemented in SROL (Voiced Sound of Romanian Project) used for the extraction of the fundamental frequency F0 and of the formants F1-F4. In order to have a better detection of the prosodic features, we put together four methods: autocorrelation method, cepstral methods, AMDF (average magnitude difference function), HPS (harmonic product spectrum). The final value of pitch (F0) is established after an integration of the partial results of each method depending on their performance. For the formants detection we used a fuzzy method of spectrum concatenation.

Keywords

cepstral analysis; correlation methods; error correction; feature extraction; speech recognition; AMDF; HPS; SROL; autocorrelation method; average magnitude difference function; cepstral methods; formants detection; fundamental frequency extraction; fuzzy method; harmonic product spectrum; prosodic features detection; spectrum concatenation; voiced sound of Romanian project; Cepstral analysis; Correlation; Feature extraction; Frequency estimation; Instruments; Software; Speech; fundamental frequency detection; precision annotation; prosodic features;

fLanguage

English

Publisher

ieee

Conference_Titel

Electrical and Power Engineering (EPE), 2014 International Conference and Exposition on

Conference_Location

Iasi

Type

conf

DOI

10.1109/ICEPE.2014.6969879

Filename

6969879