io.spokestack.spokestack.microsoft

Class AzureSpeechRecognizer

java.lang.Object

io.spokestack.spokestack.microsoft.AzureSpeechRecognizer

All Implemented Interfaces:

SpeechProcessor, AutoCloseable

public class AzureSpeechRecognizer
extends Object
implements SpeechProcessor

microsoft azure speech service recognizer

This component implements the speech processor interface using the Azure Speech Service for speech recognition.

When the speech context is triggered, the recognizer begins streaming buffered frames to the API for recognition. Once the speech context becomes inactive, the recognizer raises a RECOGNIZE event along with the audio transcript. Unfortunately, the Azure Speech SDK currently doesn't return confidence values alongside transcripts, so confidence is always set to 1.0.

Use of the Azure Speech Service implies acceptance of Microsoft's license terms, which can be found here.

This pipeline component requires the following configuration properties:

• sample-rate (integer): audio sampling rate, in Hz
• frame-width (integer): speech frame width, in ms
• locale (string): language code for speech recognition
• azure-api-key (string): API key for the Azure Speech service
• azure-region (string): Azure Speech service region

Constructor Summary

Constructors

Constructor and Description
AzureSpeechRecognizer(SpeechConfig speechConfig) initializes a new recognizer instance.

Method Summary

Modifier and Type	Method and Description
void	close() releases the resources associated with the recognizer.
void	process(SpeechContext speechContext, ByteBuffer frame) processes a frame of audio.
void	reset() resets all state internal to the stage.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

AzureSpeechRecognizer

public AzureSpeechRecognizer(SpeechConfig speechConfig)

initializes a new recognizer instance.

Parameters:

speechConfig - Spokestack speech configuration

Method Detail

reset

public void reset()

Description copied from interface: SpeechProcessor

resets all state internal to the stage.

Specified by:

reset in interface SpeechProcessor

close

public void close()

releases the resources associated with the recognizer.

Specified by:

close in interface AutoCloseable

process

public void process(SpeechContext speechContext,
                    ByteBuffer frame)
             throws Exception

processes a frame of audio.

Specified by:

process in interface SpeechProcessor

Parameters:

speechContext - the current speech context

frame - the audio frame to detect

Throws:

Exception - if there is an error performing active recognition.