change projectid per individual mrcp recognize (004 grammar-load-failure)

129 views
Skip to first unread message

assanta

unread,
Dec 27, 2018, 6:36:24 AM12/27/18
to UniMRCP
Hi
all works ok if i use the projectid in the configuration file umsgdf.xml
but it's possible to leave empty this projectid in the umsgdf and then try to use it in a individual recognize? in that case i receive "004 grammar-load-failure"

builtin:speech/transcribe?projectid=XXXXXXX
[Dec 27 11:33:33] NOTICE[2000]: src/mrcp_client_connection.c:633 : Receive MRCPv2 Data 127.0.0.1:54748 <-> 127.0.0.1:1544 [125 bytes]
MRCP/2.0 125 1 406 COMPLETE
Channel-Identifier: 2d5d4c8d90754663@speechrecog
Completion-Cause: 004 grammar-load-failure


[Dec 27 11:33:33] NOTICE[1998]: src/mrcp_client_session.c:498 : Raise App MRCP Response ASR-133 <2d5d4c8d90754663>
[Dec 27 11:33:33] ERROR[20818][C-0000001e]: app_synthandrecog.c:1626 app_synthandrecog_exec: (ASR-133) Unable to start recognition

Arsen Chaloyan

unread,
Jan 3, 2019, 7:44:50 PM1/3/19
to UniMRCP
Hi Asier,

The described method should work flawlessly with any recent version of the GDF plugin.

Server logs would be required to identify the root cause of your problem.

--
You received this message because you are subscribed to the Google Groups "UniMRCP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unimrcp+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


--
Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org

assanta

unread,
Jan 7, 2019, 11:25:54 AM1/7/19
to UniMRCP
yes, it works fine, the problem was because of a bug in my code

another question, i'm trying to use both the GSR and the GDF
      <engine id="GSR-1" name="umsgsr" enable="true"/>
      <engine id="GDF-1" name="umsgdf" enable="true"/>
but i see that always is using the GDF (2019-01-07 14:44:44:468272 [NOTICE] GDF Usage: 0/1/2)
why? it's possible to use both alternatively? how?

Arsen Chaloyan

unread,
Jan 7, 2019, 9:14:26 PM1/7/19
to UniMRCP
Well, glad everything works as intended.

As for using both GSR and GDF, you may define two different profiles in unimrcpserver.xml and assign a particular speech recognition plugin to each profile. There have been several similar discussions lately. Here is a link, where you can find a sample configuration example.

Alternatively, starting with UniMRCP 1.6.0 release, media feature tags can optionally be specified in SIP INVITE, indicating a specific engine to use.

Here is the relevant configuration of a UniMRCP client profile:

    <sip-settings id="UniMRCP-SIP-Settings">

      <feature-tags>speechrecog.engine="GSR-1";speechsynth.engine="GSS-1"</feature-tags>
    </sip-settings>

which would result in a SIP INVITE having the Accept-Contact header field set accordingly as follows:

Accept-Contact: *;speechrecog.engine="GSR-1";speechsynth.engine="GSS-1"
Reply all
Reply to author
Forward
0 new messages