다음을 통해 공유


IntentRecognizer class

의도 인식기.

Extends

생성자

IntentRecognizer(SpeechConfig, AudioConfig)

IntentRecognizer의 인스턴스를 초기화합니다.

속성

authorizationToken

서비스와 통신하는 데 사용되는 권한 부여 토큰을 가져옵니다.

canceled

이벤트가 취소됨은 인식 중에 오류가 발생했음을 나타냅니다.

internalData
properties

이 IntentRecognizer에 대해 정의된 속성 및 해당 값의 컬렉션입니다.

recognized

이벤트가 인식되면 최종 인식 결과가 수신됨을 알 수 있습니다.

recognizing

이벤트를 인식하면 중간 인식 결과가 수신됨을 알 수 있습니다.

speechRecognitionLanguage

인식의 음성 언어를 가져옵니다.

telemetryEnabled

이 메서드는 원격 분석 설정의 현재 상태를 반환합니다.

상속된 속성

sessionStarted

세션 시작 이벤트에 대한 이벤트 처리기를 정의합니다.

sessionStopped

세션 중지 이벤트에 대한 이벤트 처리기를 정의합니다.

speechEndDetected

음성 중지 이벤트에 대한 이벤트 처리기를 정의합니다.

speechStartDetected

음성 시작 이벤트에 대한 이벤트 처리기를 정의합니다.

메서드

addAllIntents(LanguageUnderstandingModel, string)
addIntent(string, string)

의도로 인식되어야 하는 구를 추가합니다.

addIntentWithLanguageModel(string, LanguageUnderstandingModel, string)

인식을 위해 Language Understanding 서비스의 의도를 추가합니다.

close(() => void, (error: string) => void)

는 이 클래스의 인스턴스가 보유한 모든 외부 리소스를 닫습니다.

recognizeOnceAsync((e: IntentRecognitionResult) => void, (e: string) => void)

의도 인식을 시작하고 첫 번째 발화가 인식된 후 중지됩니다. 작업은 인식 텍스트와 의도를 결과로 반환합니다. 참고: RecognizeOnceAsync()는 첫 번째 발화가 인식되면 를 반환하므로 명령 또는 쿼리와 같은 단일 샷 인식에만 적합합니다. 장기 실행 인식의 경우 대신 StartContinuousRecognitionAsync()를 사용합니다.

startContinuousRecognitionAsync(() => void, (e: string) => void)

stopContinuousRecognitionAsync()가 호출될 때까지 음성 인식을 시작합니다. 사용자는 인식 결과를 받으려면 이벤트를 구독해야 합니다.

startKeywordRecognitionAsync(KeywordRecognitionModel, () => void, (e: string) => void)

stopKeywordRecognitionAsync()가 호출될 때까지 키워드 스포팅을 사용하여 음성 인식을 시작합니다. 사용자는 인식 결과를 받으려면 이벤트를 구독해야 합니다. 참고: 핵심 단어 스포팅 기능은 Speech Devices SDK에서만 사용할 수 있습니다. 이 기능은 현재 SDK 자체에 포함되지 않습니다.

stopContinuousRecognitionAsync(() => void, (e: string) => void)

연속 의도 인식을 중지합니다.

stopKeywordRecognitionAsync(() => void, (e: string) => void)

연속 음성 인식을 중지합니다. 참고: 핵심 단어 스포팅 기능은 Speech Devices SDK에서만 사용할 수 있습니다. 이 기능은 현재 SDK 자체에 포함되지 않습니다.

상속된 메서드

enableTelemetry(boolean)

이 메서드는 원격 분석을 전역적으로 사용하거나 사용하지 않도록 설정합니다.

생성자 세부 정보

IntentRecognizer(SpeechConfig, AudioConfig)

IntentRecognizer의 인스턴스를 초기화합니다.

new IntentRecognizer(speechConfig: SpeechConfig, audioConfig?: AudioConfig)

매개 변수

speechConfig
SpeechConfig

구성 속성 집합입니다.

audioConfig
AudioConfig

인식기와 연결된 선택적 오디오 입력 구성

속성 세부 정보

authorizationToken

서비스와 통신하는 데 사용되는 권한 부여 토큰을 가져옵니다.

string authorizationToken

속성 값

string

권한 부여 토큰입니다.

canceled

이벤트가 취소됨은 인식 중에 오류가 발생했음을 나타냅니다.

public canceled: (sender: IntentRecognizer, event: IntentRecognitionCanceledEventArgs) => void

속성 값

(sender: IntentRecognizer, event: IntentRecognitionCanceledEventArgs) => void

internalData

object internalData

속성 값

object

properties

이 IntentRecognizer에 대해 정의된 속성 및 해당 값의 컬렉션입니다.

PropertyCollection properties

속성 값

이 IntentRecognizer에 대해 정의된 속성 및 해당 값의 컬렉션입니다.

recognized

이벤트가 인식되면 최종 인식 결과가 수신됨을 알 수 있습니다.

public recognized: (sender: IntentRecognizer, event: IntentRecognitionEventArgs) => void

속성 값

(sender: IntentRecognizer, event: IntentRecognitionEventArgs) => void

recognizing

이벤트를 인식하면 중간 인식 결과가 수신됨을 알 수 있습니다.

public recognizing: (sender: IntentRecognizer, event: IntentRecognitionEventArgs) => void

속성 값

(sender: IntentRecognizer, event: IntentRecognitionEventArgs) => void

speechRecognitionLanguage

인식의 음성 언어를 가져옵니다.

string speechRecognitionLanguage

속성 값

string

인식의 음성 언어입니다.

telemetryEnabled

이 메서드는 원격 분석 설정의 현재 상태를 반환합니다.

static boolean telemetryEnabled

속성 값

boolean

원격 분석을 사용하도록 설정하면 true이고, 그렇지 않으면 false입니다.

상속된 속성 세부 정보

sessionStarted

세션 시작 이벤트에 대한 이벤트 처리기를 정의합니다.

public sessionStarted: (sender: Recognizer, event: SessionEventArgs) => void

속성 값

(sender: Recognizer, event: SessionEventArgs) => void

Recognizer.sessionStarted에서 상속

sessionStopped

세션 중지 이벤트에 대한 이벤트 처리기를 정의합니다.

public sessionStopped: (sender: Recognizer, event: SessionEventArgs) => void

속성 값

(sender: Recognizer, event: SessionEventArgs) => void

Recognizer.sessionStopped에서 상속됨

speechEndDetected

음성 중지 이벤트에 대한 이벤트 처리기를 정의합니다.

public speechEndDetected: (sender: Recognizer, event: RecognitionEventArgs) => void

속성 값

(sender: Recognizer, event: RecognitionEventArgs) => void

Recognizer.speechEndDetected에서 상속

speechStartDetected

음성 시작 이벤트에 대한 이벤트 처리기를 정의합니다.

public speechStartDetected: (sender: Recognizer, event: RecognitionEventArgs) => void

속성 값

(sender: Recognizer, event: RecognitionEventArgs) => void

Recognizer.speechStartDetected에서 상속

메서드 세부 정보

addAllIntents(LanguageUnderstandingModel, string)

function addAllIntents(model: LanguageUnderstandingModel, intentId?: string)

매개 변수

model
LanguageUnderstandingModel

의도를 포함하는 언어 이해 모델입니다.

intentId

string

IntentRecognitionResult의 getIntentId() 메서드에서 반환할 사용자 지정 ID String입니다.

addIntent(string, string)

의도로 인식되어야 하는 구를 추가합니다.

function addIntent(simplePhrase: string, intentId?: string)

매개 변수

simplePhrase

string

intentId

string

인식할 의도의 식별자를 나타내는 String입니다.

addIntentWithLanguageModel(string, LanguageUnderstandingModel, string)

인식을 위해 Language Understanding 서비스의 의도를 추가합니다.

function addIntentWithLanguageModel(intentId: string, model: LanguageUnderstandingModel, intentName?: string)

매개 변수

intentId

string

인식할 의도의 식별자를 나타내는 String입니다. intentName이 비어 있으면 무시됩니다.

model
LanguageUnderstandingModel

Language Understanding 서비스의 의도 모델입니다.

intentName

string

의도 모델에 정의된 의도 이름입니다. 비어 있으면 모델에 정의된 모든 의도 이름이 추가됩니다.

close(() => void, (error: string) => void)

는 이 클래스의 인스턴스가 보유한 모든 외부 리소스를 닫습니다.

function close(cb?: () => void, errorCb?: (error: string) => void)

매개 변수

cb

() => void

errorCb

(error: string) => void

recognizeOnceAsync((e: IntentRecognitionResult) => void, (e: string) => void)

의도 인식을 시작하고 첫 번째 발화가 인식된 후 중지됩니다. 작업은 인식 텍스트와 의도를 결과로 반환합니다. 참고: RecognizeOnceAsync()는 첫 번째 발화가 인식되면 를 반환하므로 명령 또는 쿼리와 같은 단일 샷 인식에만 적합합니다. 장기 실행 인식의 경우 대신 StartContinuousRecognitionAsync()를 사용합니다.

function recognizeOnceAsync(cb?: (e: IntentRecognitionResult) => void, err?: (e: string) => void)

매개 변수

cb

(e: IntentRecognitionResult) => void

인식을 받은 콜백이 IntentRecognitionResult로 완료되었습니다.

err

(e: string) => void

오류 발생 시 호출된 콜백입니다.

startContinuousRecognitionAsync(() => void, (e: string) => void)

stopContinuousRecognitionAsync()가 호출될 때까지 음성 인식을 시작합니다. 사용자는 인식 결과를 받으려면 이벤트를 구독해야 합니다.

function startContinuousRecognitionAsync(cb?: () => void, err?: (e: string) => void)

매개 변수

cb

() => void

인식이 시작되면 콜백이 호출됩니다.

err

(e: string) => void

오류 발생 시 호출된 콜백입니다.

startKeywordRecognitionAsync(KeywordRecognitionModel, () => void, (e: string) => void)

stopKeywordRecognitionAsync()가 호출될 때까지 키워드 스포팅을 사용하여 음성 인식을 시작합니다. 사용자는 인식 결과를 받으려면 이벤트를 구독해야 합니다. 참고: 핵심 단어 스포팅 기능은 Speech Devices SDK에서만 사용할 수 있습니다. 이 기능은 현재 SDK 자체에 포함되지 않습니다.

function startKeywordRecognitionAsync(model: KeywordRecognitionModel, cb?: () => void, err?: (e: string) => void)

매개 변수

model
KeywordRecognitionModel

인식할 키워드를 지정하는 키워드 인식 모델입니다.

cb

() => void

인식이 시작되면 콜백이 호출됩니다.

err

(e: string) => void

오류 발생 시 호출된 콜백입니다.

stopContinuousRecognitionAsync(() => void, (e: string) => void)

연속 의도 인식을 중지합니다.

function stopContinuousRecognitionAsync(cb?: () => void, err?: (e: string) => void)

매개 변수

cb

() => void

인식이 중지되면 콜백이 호출됩니다.

err

(e: string) => void

오류 발생 시 호출된 콜백입니다.

stopKeywordRecognitionAsync(() => void, (e: string) => void)

연속 음성 인식을 중지합니다. 참고: 핵심 단어 스포팅 기능은 Speech Devices SDK에서만 사용할 수 있습니다. 이 기능은 현재 SDK 자체에 포함되지 않습니다.

function stopKeywordRecognitionAsync(cb?: () => void, err?: (e: string) => void)

매개 변수

cb

() => void

인식이 중지되면 콜백이 호출됩니다.

err

(e: string) => void

오류 발생 시 호출된 콜백입니다.

상속된 메서드 세부 정보

enableTelemetry(boolean)

이 메서드는 원격 분석을 전역적으로 사용하거나 사용하지 않도록 설정합니다.

static function enableTelemetry(enabled: boolean)

매개 변수

enabled

boolean

원격 분석 컬렉션에 대한 전역 설정입니다. true로 설정하면 마이크 오류, 인식 오류와 같은 원격 분석 정보가 수집되어 Microsoft로 전송됩니다. false로 설정하면 원격 분석이 Microsoft로 전송되지 않습니다.

Recognizer.enableTelemetry에서 상속됨