Pre-processing for automatic speech recognition

Pre-processing for automatic speech recognition

A noise analyser 131 and processor 141 receives speech signals from a mic array 111-115 and identifies defined types of noise (eg. wind, handling, motor, structural or actuator) based on energy and correlation estimates and supplies the filtered signal for Automatic Speech Recognition 161. Mic signa...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hannu Pulakka, Matti Sakari Hämäläinen, Jorma Mäkinen
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:A noise analyser 131 and processor 141 receives speech signals from a mic array 111-115 and identifies defined types of noise (eg. wind, handling, motor, structural or actuator) based on energy and correlation estimates and supplies the filtered signal for Automatic Speech Recognition 161. Mic signals may be selected based on whether the maximum frequency of the noise lies above a threshold (fig. 3). The signals may be weighted, time-aligned or filtered based on direction of arrival.