Posted: Fri Mar 16, 2007 10:37 pm Post subject: [asterisk-speech-rec] DTMF Delay CRM:00170810
Coincidentally, I just had a customer contact me directly with the DTMF
issue, and I forwarded him this discussion.
He just responded with:
"Sweet - this was perfect - I went into app_speech_utils.c at line 668
and changed it to '2' instead of '5' recompiled and everything was
good!"
I haven't tried it on a test system here yet, but you may want to check
that you're setting the same value.
Stephen Keller
LumenVox Technical Support
P: 877-977-0707, just say "Stephen Keller" or "Support"
F: 858-707-7072
support@LumenVox.com www.LumenVox.com
-----Original Message-----
From: asterisk-speech-rec-bounces@lists.digium.com
[mailto:asterisk-speech-rec-bounces@lists.digium.com] On Behalf Of Scott
Higginbotham
Sent: Friday, March 16, 2007 3:09 PM
To: Use of speech recognition in Asterisk
Subject: RE: [asterisk-speech-rec] DTMF Delay
I'll give it a shot when calls die down on the platform and post my
findings. Thanks!
Scott Higginbotham
Systems / Network Operations Manager
215.259.2185 or 1.800.835.5710 ext 2185
scotth@voicenet.com
Hey-
The culprit is most likely somewhere in the precompiled (blackbox)
"connector"-- which cant be tinkered with.
Ive noticed that it SOMETIMES takes several seconds for SpeechDestroy()
to work.
Perhaps there is an issue within the connector with killing the "speech
object" if it has NOT ACTUALLY received a valid decode from the speech
engine. This would account for moving on in the dialplan rapidly after
ASR but slowly after DTMF.
If you are going to use the speech engine "later" in the dialplan-- Id
suggest unloading the active grammars AND skipping the call to
SpeechDestroy()-- until the end of the "session."
Please post results if you try above. Thanks.
Cheers-
jjq
Quote:
From: "Scott Higginbotham" <scotth@voicenet.com>
Reply-To: Use of speech recognition in
Asterisk<asterisk-speech-rec@lists.digium.com>
To: "Use of speech recognition in
Asterisk"<asterisk-speech-rec@lists.digium.com>
Subject: RE: [asterisk-speech-rec] DTMF Delay
Date: Fri, 16 Mar 2007 11:51:46 -0400
Interesting.
I modified the app_speech_utils.c file to reduce the DTMF timeout from
5 seconds to 1 second - the time it takes from the time a DTMF is
pressed till the next appropriate action is reduced, but still taking
about 6-7 seconds longer than if speech is utilized to make your
selection instead of DTMF.
Is there any other time-out settings that may be hidden with code
somewhere?
Scott Higginbotham
Systems / Network Operations Manager
215.259.2185 or 1.800.835.5710 ext 2185 scotth@voicenet.com
Hey-
The way dialplan command SpeechBackground(Sound File|Timeout) is
written there is ALWAYS a 5 second digit time-out-- AFTER a key is
pressed. You can see this in app_speech_utils.c at line 668.
The whole app_speech_utils.c DTMF capture is a bit wonky. The speech
engine does NOT interpret any DTMF-- app_speech_utils.c simply creates
Quote:
a string of valid DTMF, and then populates speech->results AFTER the 5
second digit time-out.
This method of DTMF capture BYPASSES the speech engine AND subsequently
Quote:
ALL semantic command interpretation from within the decoder-- this is
NOT in the docs.
Cheers-
jjq
>From: "Scott Higginbotham" <scotth@voicenet.com>
>Reply-To: Use of speech recognition in
>Asterisk<asterisk-speech-rec@lists.digium.com>
>To: "Use of speech recognition in
>Asterisk"<asterisk-speech-rec@lists.digium.com>
>Subject: RE: [asterisk-speech-rec] DTMF Delay
>Date: Thu, 15 Mar 2007 12:11:00 -0400
>
>Anyone have any thoughts on the delay I'm experiencing? Is there
>anything I'm overlooking in the dial plan from what anyone can see?
>
>Scott Higginbotham
>Systems / Network Operations Manager
>215.259.2185 or 1.800.835.5710 ext 2185 scotth@voicenet.com
>
>-----Original Message-----
>From: asterisk-speech-rec-bounces@lists.digium.com
>[mailto:asterisk-speech-rec-bounces@lists.digium.com] On Behalf Of
>Scott Higginbotham
>Sent: Sunday, March 11, 2007 8:40 PM
>To: asterisk-speech-rec@lists.digium.com
>Subject: [asterisk-speech-rec] DTMF Delay
>
>I've got an odd situation with DTMF selections being given a delay
>compared to the responsiveness of a voice selected menu option.
>
>I load both grammars, one for selecting options 1-9 via Voice
>selection
>and one for DTMF selection so that the user can simply press 1-9 if
>they don't like to speak the menu option.
>
>Both work; however selecting the menu option with DTMF causes a good
>5-10 second delay before you are taken to the next context after
>selecting your option, whereas speaking the command is almost
>instantaneously taking you to the next context without the 5-10
>second delay.
>
>A snippet of my code looks like:
>
>exten => s,1,Playback(custom/default/welcome_0)
>exten => s,2,SpeechCreate(lumenvox)
>exten => s,3,SpeechActivateGrammar(numbers)
>exten => s,4,SpeechActivateGrammar(choicetone)
>exten => s,5,NoOp(Make it past grammar load) exten => s,n,SpeechStart
_________________________________________________________________
Mortgage rates as low as 4.625% - Refinance $150,000 loan for $579 a
month.
Intro*Terms
https://www2.nextag.com/goto.jsp?product=100000035&url=%2fst.jsp&tm=y&s
e
arch=mortgage_text_links_88_h27f6&disc=y&vers=743&s=4056&p=5117
_______________________________________________
--Bandwidth and Colocation provided by Easynews.com --
You can post new topics in this forum You can reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum