WordLevelTiming interface

Package:

@azure/ai-language-conversations

Word level timing information generated by the speech to text API. The words in this object should have 1:1 correspondence with the 'lexical' input to allow for audio redaction.

Extends

AudioTiming

Properties

word	The word recognized.

Inherited Properties

duration	Duration of word articulation, in ticks. 1 tick = 100 ns.
offset	Offset from start of speech audio, in ticks. 1 tick = 100 ns.

Property Details

word

The word recognized.

word?: string

Property Value

string

Inherited Property Details

duration

Duration of word articulation, in ticks. 1 tick = 100 ns.

duration?: number

Property Value

number

Inherited From AudioTiming.duration

offset

Offset from start of speech audio, in ticks. 1 tick = 100 ns.

offset?: number

Property Value

number

Inherited From AudioTiming.offset