speech/io.github.maniramezan.kommon.speech/SpeechTranscript

SpeechTranscript

data class SpeechTranscript(val finalizedUtterances: List<RecognizedUtterance> = emptyList(), val volatileText: String = "")

Everything recognized so far in one recognition run.

finalizedUtterances won't change again. volatileText is the recognizer's current guess at speech it hasn't finalized yet, and is replaced by each update, so show it as a preview.

Constructors

SpeechTranscript

constructor(finalizedUtterances: List<RecognizedUtterance> = emptyList(), volatileText: String = "")

Properties

finalizedText

val finalizedText: String

The finalized text, without the volatile preview.

finalizedUtterances

val finalizedUtterances: List<RecognizedUtterance>

Utterances the recognizer has finalized, in order.

isFinal

val isFinal: Boolean

Whether nothing is pending finalization.

text

val text: String

The finalized text followed by the volatile preview, parts trimmed and joined by single spaces.

volatileText

val volatileText: String

The recognizer's current guess at not-yet-final speech; empty when nothing is pending.

words

val words: List<RecognizedWord>

Every finalized word, in order.