The audiopreproc
module can perform DTMF (Dual-Tone Multi-Frequency) dial tone detection and identification. In this mode, the module can accurately detect and identify tones corresponding to the numbers 0-9, the letters A-D, and the asterisk (*) and hash (#) keys.
To enable DTMF dial tone identification you can:
audiopreproc
mode to A
to enable the output classification stream. In this case, by default the DoToneClass
parameter is set to Auto
, and Speech Server identifies DTMF tones for 8kHz (telephony) audio, but not for 16kHz audio. You can also set DoToneClass
to True
to always identify DTMF tones. Set DoToneClass
to False
to never identify DTMF tones. audiopreproc
mode to T
to output detected tones to a separate stream. These options are not mutually exclusive.
For an example schema incorporating DTMF dial tone identification, see Identify DTMF Dial Tones. This schema produces a stream of DTMF labels (such as <DTMF-1>
, <DTMF-A>
or <DTMF-#>
).
|