METHOD FOR DETECTING VOICE END POINT USING LINE SPECTRAL PAIRS

METHOD FOR DETECTING VOICE END POINT USING LINE SPECTRAL PAIRS

PURPOSE: A method for detecting a voice end point using line spectral pairs is provided to decodes the line spectral pairs from an input voice packet which was vocoded. CONSTITUTION: A method for detecting a voice end point using line spectral pairs includes following steps. At the first step(212),...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: KIM, JAE WON, KANG, MYEONG SU, MIN, BYUNG JUN, KIM, BYUNG MU
Format: Patent
Sprache:eng
Schlagworte:
ACOUSTICS
MUSICAL INSTRUMENTS
PHYSICS
SPEECH ANALYSIS OR SYNTHESIS
SPEECH OR AUDIO CODING OR DECODING
SPEECH OR VOICE PROCESSING
SPEECH RECOGNITION
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:PURPOSE: A method for detecting a voice end point using line spectral pairs is provided to decodes the line spectral pairs from an input voice packet which was vocoded. CONSTITUTION: A method for detecting a voice end point using line spectral pairs includes following steps. At the first step(212), a position at which the distance between line spectral pairs vector decoded in a voice packet and the line spectral pairs is longer than a threshold value is determined as a beginning position of the voice interval. At the second step, the position at which the line spectral pairs distance is to be shorter than the threshold value(223) is determined as an end position of the voice interval.