Interface ASRWord

Word-level recognition result with timing and optional phone details.

interface ASRWord {
    text: string;
    startTime: number;
    duration: number;
    isTag: boolean;
    phones?: ASRPhone[];
}

Properties

text: string

Recognized word text. May include #tag suffix if an alternative pronunciation was matched.

startTime: number

Start time of this word in seconds from the beginning of the audio.

duration: number

Duration of this word in seconds.

isTag: boolean

Whether this word is a special token (e.g., <SPOKEN_NOISE>).

phones?: ASRPhone[]

Array of phonemes that comprise this word.