The stt
module converts normalized speech frontend frame data into a text transcription.
Mode | Input | Output | Description |
---|---|---|---|
nf
|
w
|
Accepts normalized feature frames and performs speech-to-text. Produces a word label stream. |
Example:
w <- stt (_, nf)
Diag | LatFile |
DiagFile | LatScale |
DiagFileLabel | LatWordFile |
DnnScale | Mode |
EnableConfidence | ModeValue |
FrameDupl | SpeedBiasLevel |
Lang* |
*You must set the value of this configuration parameter by setting the corresponding action parameter.
|