UNPKG

microsoft-cognitiveservices-speech-sdk

Version:
69 lines (68 loc) 1.92 kB
import { OnSuccess } from "./OnSuccess"; import { OnUnknown } from "./OnUnknown"; import { LanguageIdOutput } from "./LanguageIdOutput"; /** * The enum that represents which mode will language detection take place * There is only detectAtAudioStart mode for now as language detection models are not trained for different modes * This enum can be extended in future to support different modes */ export declare enum LanguageIdDetectionMode { DetectAtAudioStart = "DetectAtAudioStart", DetectContinuous = "DetectContinuous", DetectSegments = "DetectSegments" } /** * The language id detection mode, setting this will load the detection setting of MaxAudioDuration and MaxSpeechDuration * If the maxAudioDuration and maxSpeechDuration is set in the speech.context, then this detection mode will be ignored */ export declare enum LanguageIdDetectionPriority { /** * default, Service decides the best mode to use. */ Auto = "Auto", /** * Offers lower latency via a trade-off of accuracy. */ PrioritizeLatency = "PrioritizeLatency", /** * Offers higher accuracy via a trade-off of latency. */ PrioritizeAccuracy = "PrioritizeAccuracy" } /** * The language id context */ export interface LanguageIdContext { /** * The candidate languages for speaker language detection. */ languages: string[]; /** * The on success action. */ onSuccess?: OnSuccess; /** * The language detection mode. */ mode?: LanguageIdDetectionMode; /** * The fallback language. */ onUnknown?: OnUnknown; /** * The output */ output?: LanguageIdOutput; /** * The max audio duration */ maxAudioDuration?: number; /** * The max speech duration */ maxSpeechDuration?: number; /** * The priority. */ priority?: LanguageIdDetectionPriority; }