[asterisk-speech-rec] How does one tell the difference betweennumbers and letters - Lumenvox

[David Merel]

ok thanks, unfortunately that wont go well with callers at least not in my
situation however I think I can come up with a logic which will check
against alternate possibilities.

Btw is there a resource anywhere on the web that you know of which has .gram
files one can use?



On Fri, Oct 10, 2008 at 6:55 PM, Stephen Keller
<StephenKeller at lumenvox.com>wrote:

>  This is a pretty common problem, and one without a great solution. The
> fact is that "Eight" and "A" sound very similar in English -- even humans
> have great difficulty distinguishing them over the phone. If it's hard for
> humans, it's harder for a machine.
>
>
>
> The best way to deal with it, if possible, is to ask people to say things
> like "Alpha" or "Alex" for "A" and add those to your grammar, which I think
> you've started doing with "Y."
>
> Thanks,
>
> *Stephen Keller
> *Project Manager
> *LumenVox**
> **P**: *1-877-977-0707 – just say, "Support"
> *P**: *+1-858-707-7700
> *F**: *+1-858-707-7072
> *support at LumenVox.com* <support at LumenVox.com>
> *www.LumenVox.com* <http://www.lumenvox.com/>
>
>
>
>
>
> *From:* asterisk-speech-rec-bounces at lists.digium.com [mailto:
> asterisk-speech-rec-bounces at lists.digium.com] *On Behalf Of *David Merel
> *Sent:* Friday, October 10, 2008 3:31 PM
> *To:* asterisk-speech-rec at lists.digium.com
> *Subject:* [asterisk-speech-rec] How does one tell the difference
> betweennumbers and letters - Lumenvox
>
>
>
> So I have a Grammar file where I am trying to combine letters and numbers
> so that people can speak a model number of a MAC address of a network device
> into the phone.
>
> Problem is 8 and A are commonly being mixed up, even with the speech tuner
> it doesnt seem to be too accurate, is there any way in the grammar file to
> further enhance a Letters recognition?
>
> Here is a small example, btw for example you will see below I added "Ay"
> not sure if that even valid, but with or without I am getting the same
> accuracy issues.
>
> $Digit = (ONE:"1"        |
>           TWO:"2"        |
>           THREE:"3"      |
>           FOUR:"4"       |
>           FIVE:"5"       |
>           SIX:"6"        |
>           SEVEN:"7"      |
>           EIGHT:"8"      |
>           NINE:"9"       |
>           (ZERO | O):"0" );
>
> $A = (A|Ay):"A";
> $B = (B|bee):"B";
> $C = (C|see):"C";
> $D = (D|dee):"D";
>
> $Y = (Yankee $GARBAGE|
> Young|Why):"Y";
>
> $Digits = {$=''} (($Digit|$A|$B|$C|$D|$Y){$+=$$})<1->;
>
> Any help would be great
>
> Fyi, I am working with Lumenvox Speech Engine
>
> _______________________________________________
> --Bandwidth and Colocation Provided by http://www.api-digital.com--
>
> asterisk-speech-rec mailing list
> To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-speech-rec
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-speech-rec/attachments/20081010/7c3e8da7/attachment.htm