Pitch Detection Algorithm for Continuous Speech Signals Using Viterbi Traceback with Temporal Forgetting

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F11%3A00183475" target="_blank" >RIV/68407700:21230/11:00183475 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Pitch Detection Algorithm for Continuous Speech Signals Using Viterbi Traceback with Temporal Forgetting

Popis výsledku v původním jazyce

This paper presents a pitch-detection algorithm aimed to be applied on signals containing continuous speech. The core of the method is based on merged normalized forward-backward correlation (MNFBC) working in a time domain with ability to voicing decision. In addition to this, Viterbi traceback is used for post-processing the MNFBC output considering three best F0 candidates in each step. This should make final pitch contour smoother and should also prevent octave errors. In transition probabilities between these F0 candidates two major improvements were made over existing post-processing method. Firstly, we compare pitch distance in musical cents, and secondly a temporal forgetting is being applied not to penalize pitch jumps after prosodic pauses ofone speaker or changes in pitch connected with turn-taking in dialogs. Results computed on a pitch-reference database definitely show the benefit of first one, but did not prove benefits of the second yet.

Název v anglickém jazyce

Pitch Detection Algorithm for Continuous Speech Signals Using Viterbi Traceback with Temporal Forgetting

Popis výsledku anglicky

This paper presents a pitch-detection algorithm aimed to be applied on signals containing continuous speech. The core of the method is based on merged normalized forward-backward correlation (MNFBC) working in a time domain with ability to voicing decision. In addition to this, Viterbi traceback is used for post-processing the MNFBC output considering three best F0 candidates in each step. This should make final pitch contour smoother and should also prevent octave errors. In transition probabilities between these F0 candidates two major improvements were made over existing post-processing method. Firstly, we compare pitch distance in musical cents, and secondly a temporal forgetting is being applied not to penalize pitch jumps after prosodic pauses ofone speaker or changes in pitch connected with turn-taking in dialogs. Results computed on a pitch-reference database definitely show the benefit of first one, but did not prove benefits of the second yet.

Druh

D - Stať ve sborníku

CEP obor

JA - Elektronika a optoelektronika, elektrotechnika

OECD FORD obor

—

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2011

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

POSTER 2011 - 15th International Student Conference on Electrical Engineering

ISBN

978-80-01-04806-1

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

1-5

Název nakladatele

ČVUT, Fakulta elektrotechnická

Místo vydání

Praha

Místo konání akce

Prague

Datum konání akce

12. 5. 2011

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

—