Powered by Universal Speech Solutions LLC

 MRCP

Azure Bot Service Plugin

Usage Guide

 

Revision: 2

Created: May 1, 2020

Last updated: January 4, 2021

Author: Arsen Chaloyan


 

Table of Contents

 

1  Overview.. 4

1.1         Installation. 4

1.2         Applicable Versions 4

2  Supported Features. 5

2.1         MRCP Methods. 5

2.2         MRCP Events. 5

2.3         MRCP Header Fields. 5

2.4         Grammars. 6

2.5         Results. 6

3  Configuration Format 7

3.1         Document 7

3.2         Streaming Recognition. 8

3.3         Results. 10

3.4         Speech Contexts. 11

3.5         Speech Context 12

3.6         Phrase. 13

3.7         Speech and DTMF Input Detector 13

3.8         Utterance Manager 15

3.9         RDR Manager 16

3.10       Monitoring Agent 17

3.11       Usage Change Handler 18

3.12       Usage Refresh Handler 19

3.13       License Server 19

4  Configuration Steps. 21

4.1         Using Default Configuration. 21

4.2         Specifying LUIS Model 21

4.3         Specifying Intents. 22

4.4         Specifying Recognition Language. 22

4.5         Specifying Sampling Rate. 23

4.6         Specifying Speech Input Parameters. 23

4.7         Specifying DTMF Input Parameters 24

4.8         Specifying No-Input and Recognition Timeouts. 24

4.9         Specifying Vendor-Specific Parameters. 24

4.10       Maintaining Utterances. 26

4.11       Maintaining Recognition Details Records. 26

5  Recognition Grammars and Results. 28

5.1         Using Built-in Speech Transcription. 28

5.2         Using Built-in DTMF Grammars 28

5.3         Retrieving Results. 29

6  Monitoring Usage Details 31

6.1         Log Usage. 31

6.2         Update Usage. 31

6.3         Dump Channels. 32

7  Usage Examples 33

7.1         Speech Transcription. 33

8  Sequence Diagram.. 35

9  References 36

9.1         Microsoft Azure. 36

9.2         Specifications. 36

 

 

1       Overview

This guide describes how to configure and use the Microsoft Azure Bot Service (ABS) plugin to the UniMRCP server. The document is intended for users having a certain knowledge of Microsoft Azure Speech APIs and UniMRCP.

 

1.1      Installation

For installation instructions, use one of the guides below.

·         Deb Package Installation (Debian / Ubuntu)

1.2      Applicable Versions

Instructions provided in this guide are applicable to the following versions.

 

UniMRCP 1.7.0 and above

UniMRCP ABS Plugin 1.0.0 and above

 

2       Supported Features

This is a brief check list of the features currently supported by the UniMRCP server running with the Azure SR plugin.

2.1      MRCP Methods

ü  DEFINE-GRAMMAR

ü  RECOGNIZE

ü  START-INPUT-TIMERS

ü  STOP

ü  SET-PARAMS

ü  GET-PARAMS

2.2      MRCP Events

ü  RECOGNITION-COMPLETE

ü  START-OF-INPUT

2.3      MRCP Header Fields

ü  Input-Type

ü  No-Input-Timeout

ü  Recognition-Timeout

ü  Speech-Complete-Timeout

ü  Speech-Incomplete-Timeout

ü  Waveform-URI

ü  Media-Type

ü  Completion-Cause

ü  Confidence-Threshold

ü  Start-Input-Timers

ü  DTMF-Interdigit-Timeout

ü  DTMF-Term-Timeout

ü  DTMF-Term-Char

ü  Save-Waveform

ü  Speech-Language

ü  Cancel-If-Queue

ü  Sensitivity-Level

2.4      Grammars

ü  Built-in speech transcription grammar

ü  Built-in/embedded DTMF grammar

ü  SRGS XML (limited support)

2.5      Results

ü  NLSML

ü  JSON

 

3       Configuration Format

The configuration file of the ABS plugin is located in /opt/unimrcp/conf/umsazurebot.xml. The configuration file is written in XML.

3.1      Document

The root element of the XML document must be <umsazurebot>.

Attributes

 

Name

      Unit

Description

license-file

File path

Specifies the license file. File name may include patterns containing '*' sign. If multiple files match the pattern, the most recent one gets used.

subscription-key-file

File path

Specifies the Microsoft subscription key file to use. File name may include patterns containing '*' sign. If multiple files match the pattern, the most recent one gets used.

 

Parent

None.

Children

 

Name

Unit

Description

<streaming-recognition>

String

Specifies parameters of streaming recognition employed via Microsoft Speech WebSocket protocol.

<results>

String

Specifies parameters of recognition results set in RECOGNITION-COMPLETE events.

<speech-contexts>

String

Contains a list of speech contexts.

<speech-dtmf-input-detector>

String

Specifies parameters of the speech and DTMF input detector.

<utterance-manager>

String

Specifies parameters of the utterance manager.

<rdr-manager>

String

Specifies parameters of the Recognition Details Record (RDR) manager.

<monitoring-agent>

String

Specifies parameters of the monitoring manager.

<license-server>

String

Specifies parameters used to connect to the license server. The use of the license server is optional.

 

Example

This is an example of a bare document.

 

< umsazurebot license-file="umsazurebot_*.lic" subscription-key-file="azbot.subscription.key">

</ umsazurebot>

 

3.2      Streaming Recognition

This element specifies parameters of Microsoft WebSocket streaming recognition.

Attributes

 

Name

Unit

Description

language

String

Specifies the default language to use, if not set by the client.

max-alternatives

Integer

Specifies the maximum number of speech recognition result alternatives to be returned. Can be overridden by client by means of the header field N-Best-List-Length.

alternatives-below-threshold

Boolean

Specifies whether to return speech recognition result alternatives with the confidence score below the confidence threshold.

start-of-input

String

Specifies the source of start of input event sent to the client (use "service-originated" to rely on service-originated startDetected event and "internal" for an event determined by plugin).

skip-unsupported-grammars

Boolean

Specifies whether to skip or raise an error while referencing a malformed or not supported grammar.

skip-empty-results

Boolean

Specifies whether to implicitly initiate a new gRPC request if the current one completes with an empty result.

transcription-grammar

String

Specifies the name of the built-in speech transcription grammar. The grammar can be referenced as builtin:speech/transcribe or builtin:grammar/transcribe, where transcribe is the default value of this parameter.

word-info

Boolean

Specifies whether to return word-level time offset information. Can be overridden by client.

inter-result-timeout

Time interval [msec]

Specifies a timeout between interim results containing transcribed speech. If the timeout is elapsed, input is considered complete. The timeout defaults to 0 (disabled).

endpoint-id

String

Specifies the custom endpoint identifier, if used. Can be overridden by client.

http-proxy-host

String

Specifies the host name of HTTP proxy, if used.

http-proxy-port

Integer

Specifies the port number of HTTP proxy, if used.

http-proxy-username

String

Specifies the username employed for HTTP proxy authentication, if used.

http-proxy-password

String

Specifies the password employed for HTTP proxy authentication, if used.

appid

String

The application id of the LUIS model, if used. Can be overridden by client.

intents

String

A comma-separated list of intents to be used from the specified LUIS model, if any. Can be overridden by client.

sdk-log

Boolean

Specifies whether to enable SDK logging. Available since 1.1.0.

 

Parent

<umsazurebot>

Children

None.

Example

This is an example of streaming recognition element.

 

   <streaming-recognition

       language="en-US"

       max-alternatives="1"

       alternatives-below-threshold="false"

       start-of-input="service-originated"

       skip-unsupported-grammars="true"

       transcription-grammar="transcribe"

   />

3.3      Results

This element specifies parameters of recognition results set in RECOGNITION-COMPLETE events.

Attributes

 

Name

Unit

Description

format

String

Specifies the format of results to be returned to the client (use "standard" for NLSML and "json" for JSON).

indent

Integer

Specifies the indent to use while composing the results.

replace-dots

Boolean

Specifies whether to replace '.' with '_' in the parameter names, used while composing an XML content. The parameter is observed only if the format is set to standard.

replace-dashes

Boolean

Specifies whether to replace '-' with '_' in the parameter names, used while composing an XML content. The parameter is observed only if the format is set to standard. Available since GDF 1.15.0.

confidence-format

String

Specifies the format of the confidence score to be returned. The parameter is observed only if the format is set to standard. Use one of:

·         auto for a format based on protocol version,

·         mrcpv2 for a float value in the range of 0..1,

·         mrcpv1 for an integer value in the range of 0..100

Available since GDF 1.7.0.

tag-format

String

Specifies the format of the instance element to be returned. The parameter is observed only if the format is set to standard. Use one of:

·         semantics/xml  for query result represented in XML [default]

·         semantics/json for query result represented in JSON

·         swi-semantics/xml for query result set in an inner <SWI_meaning> element represented in XML

·         swi-semantics/json for query result set in an inner <SWI_meaning> element represented in JSON

Available since GDF 1.9.0.

 

Parent

<umsgdf>

Children

None.

Example

This is an example of results element.

 

   <results

      format="standard"

      indent="0"

      replace-dots="true"

      confidence-format="auto"

      tag-format="semantics/xml"

   />

 

3.4      Speech Contexts

This element specifies a list of speech contexts.

Attributes

None.

Parent

<umsazurebot>

Children

<speech-context>

Example

The example below defines a speech contexts directory.

 

   <speech-contexts>

      <speech-context id="directory" speech-complete="true" enable="true">

         <phrase>call Steve</phrase>

         <phrase>call John</phrase>

         <phrase>dial 5</phrase>

         <phrase>dial 6</phrase>

      </speech-context>

   </speech-contexts>

 

3.5      Speech Context

This element specifies a speech context.

Attributes

 

Name

Unit

Description

id

String

Specifies a unique string identifier of the speech context to be referenced by the MRCP client.

enable

Boolean

Specifies whether the speech context is enabled or disabled.

speech-complete

Boolean

Specifies whether to complete input as soon as an interim result matches one of the specified phrases.

language

String

The language the phrases are defined for.

scope

String

Specifies a scope of the speech context, which can be set to either hint or strict.

 

Parent

<speech-contexts>

Children

<phrase>

Example

This is an example of speech context element.

 

      <speech-context id="directory" speech-complete="true" enable="true">

         <phrase>call Steve</phrase>

         <phrase>call John</phrase>

         <phrase>dial 5</phrase>

         <phrase>dial 6</phrase>

      </speech-context>

 

3.6      Phrase

This element specifies a phrase in the speech context.

Attributes

 

Name

Unit

Description

tag

String

Specifies an optional arbitrary string identifier to be returned as an instance in the NLSML result, if the transcription result matches the phrase.

 

Parent

<speech-context>

Children

None.

 

This is an example of a speech context with phrases having tags specified.

 

      <speech-context id="boolean" speech-complete="true" scope="strict" enable="true">

         <phrase tag="true">yes</phrase>

         <phrase tag="true">sure</phrase>

         <phrase tag="true">correct</phrase>

         <phrase tag="false">no</phrase>

         <phrase tag="false">not sure</phrase>

         <phrase tag="false">incorrect </phrase>

      </speech-context>

 

3.7      Speech and DTMF Input Detector

This element specifies parameters of the speech and DTMF input detector.

Attributes

 

Name

Unit

Description

vad-mode

Integer

Specifies an operating mode of VAD in the range of [0  ... 3]. Default is 1.

speech-start-timeout

Time interval [msec]

Specifies how long to wait in transition mode before triggering a start of speech input event.

speech-complete-timeout

Time interval [msec]

Specifies how long to wait in transition mode before triggering an end of speech input event. The complete timeout is used when there is an interim result available.

speech-incomplete-timeout

Time interval [msec]

Specifies how long to wait in transition mode before triggering an end of speech input event. The incomplete timeout is used as long as there is no interim result available. Afterwards, the complete timeout is used.

noinput-timeout

Time interval [msec]

Specifies how long to wait before triggering a no-input event.

input-timeout

Time interval [msec]

Specifies how long to wait for input to complete.

dtmf-interdigit-timeout

Time interval [msec]

Specifies a DTMF inter-digit timeout.

dtmf-term-timeout

Time interval [msec]

Specifies a DTMF input termination timeout.

dtmf-term-char

Character

Specifies a DTMF input termination character.

speech-leading-silence

Time interval [msec]

Specifies desired silence interval preceding spoken input.

speech-trailing-silence

Time interval [msec]

Specifies desired silence interval following spoken input.

speech-output-period

Time interval [msec]

Specifies an interval used to send speech frames to the recognizer.

 

Parent

<umsazurebot>

Children

None.

Example

The example below defines a typical speech and DTMF input detector having the default parameters set.

 

   <speech-dtmf-input-detector

      vad-mode="2"

      speech-start-timeout="300"

      speech-complete-timeout="1000"

      speech-incomplete-timeout="3000"

      noinput-timeout="5000"

      input-timeout="10000"

      dtmf-interdigit-timeout="5000"

      dtmf-term-timeout="10000"

      dtmf-term-char=""

      speech-leading-silence="300"

      speech-trailing-silence="300"

      speech-output-period="200"

   />

 

3.8      Utterance Manager

This element specifies parameters of the utterance manager.

Attributes

 

Name

Unit

Description

save-waveforms

Boolean

Specifies whether to save waveforms or not.

purge-existing

Boolean

Specifies whether to delete existing records on start-up.

max-file-age

Time interval [min]

Specifies a time interval in minutes after expiration of which a waveform is deleted. Set 0 for infinite.

max-file-count

Integer

Specifies the max number of waveforms to store. If reached, the oldest waveform is deleted. Set 0 for infinite.

waveform-base-uri

String

Specifies the base URI used to compose an absolute waveform URI.

waveform-folder

Dir path

Specifies a folder the waveforms should be stored in.

file-prefix

String

Specifies a prefix used to compose the name of the file to be stored. Defaults to 'umsazurebot-', if not specified.

use-logging-tag

Boolean

Specifies whether to use the MRCP header field Logging-Tag, if present, to compose the name of the file to be stored.

 

Parent

<umsazurebot>

Children

None.

Example

The example below defines a typical utterance manager having the default parameters set.

 

   <utterance-manager

      save-waveforms="false"

      purge-existing="false"

      max-file-age="60"

      max-file-count="100"     

      waveform-base-uri="http://localhost/utterances/"

      waveform-folder=""

   />

3.9      RDR Manager

This element specifies parameters of the Recognition Details Record (RDR) manager.

Attributes

 

Name

Unit

Description

save-records

Boolean

Specifies whether to save recognition details records or not.

purge-existing

Boolean

Specifies whether to delete existing records on start-up.

max-file-age

Time interval [min]

Specifies a time interval in minutes after expiration of which a record is deleted. Set 0 for infinite.

max-file-count

Integer

Specifies the max number of records to store. If reached, the oldest record is deleted. Set 0 for infinite.

record-folder

Dir path

Specifies a folder to store recognition details records in. Defaults to ${UniMRCPInstallDir}/var.

file-prefix

String

Specifies a prefix used to compose the name of the file to be stored. Defaults to 'umsazurebot-', if not specified.

use-logging-tag

Boolean

Specifies whether to use the MRCP header field Logging-Tag, if present, to compose the name of the file to be stored.

 

Parent

<umsazurebot>

Children

None.

Example

The example below defines a typical utterance manager having the default parameters set.

 

   <rdr-manager

      save-records="false"

      purge-existing="false"

      max-file-age="60"

      max-file-count="100"     

      waveform-folder=""

   />

3.10 Monitoring Agent

This element specifies parameters of the monitoring agent.

Attributes

 

Name

Unit

Description

refresh-period

Time interval [sec]

Specifies a time interval in seconds used to periodically refresh usage details. See <usage-refresh-handler>.

 

Parent

<umsazurebot>

Children

<usage-change-handler>

<usage-refresh-handler>

Example

The example below defines a monitoring agent with usage change and refresh handlers.

 

   <monitoring-agent refresh-period="60">

 

      <usage-change-handler>

        <log-usage enable="true" priority="NOTICE"/>

      </usage-change-handler>

 

      <usage-refresh-handler>

        <dump-channels enable="true" status-file="umsazurebot-channels.status"/>

      </usage-refresh-handler >

 

   </monitoring-agent>

 

3.11   Usage Change Handler

This element specifies an event handler called on every usage change.

Attributes

None.

Parent

<monitoring-agent>

Children

<log-usage>

<update-usage>

<dump-channels>

Example

This is an example of the usage change event handler.

 

      <usage-change-handler>

        <log-usage enable="true" priority="NOTICE"/>

        <update-usage enable="false" status-file="umsazurebot-usage.status"/>

        <dump-channels enable="false" status-file="umsazurebot-channels.status"/>

      </usage-change-handler>

3.12  Usage Refresh Handler

This element specifies an event handler called periodically to update usage details.

Attributes

None.

Parent

<monitoring-agent>

Children

<log-usage>

<update-usage>

<dump-channels>

Example

This is an example of the usage change event handler.

 

      <usage-refresh-handler>

        <log-usage enable="true" priority="NOTICE"/>

        <update-usage enable="false" status-file="umsazurebot-usage.status"/>

        <dump-channels enable="false" status-file="umsazurebot-channels.status"/>

      </usage-refresh-handler>

 

3.13  License Server

This element specifies parameters used to connect to the license server.

Attributes

 

Name

Unit

Description

enable

Boolean

Specifies whether the use of license server is enabled or not. If enabled, the license-file attribute is not honored.

server-address

String

Specifies the IP address or host name of the license server.

certificate-file

File path

Specifies the client certificate used to connect to the license server. File name may include patterns containing a '*' sign. If multiple files match the pattern, the most recent one gets used.

ca-file

File path

Specifies the certificate authority used to validate the license server.

channel-count

Integer

Specifies the number of channels to check out from the license server. If not specified or set to 0, either all available channels or a pool of channels will be checked based on the configuration of the license server.

http-proxy-address

String

Specifies the IP address or host name of the HTTP proxy server, if used.

http-proxy-port

Integer

Specifies the port number of the HTTP proxy server, if used.

 

Parent

<umsazurebot>

Children

None.

Example

The example below defines a typical configuration which can be used to connect to a license server located, for example, at 10.0.0.1.

 

   <license-server

      enable="true"

      server-address="10.0.0.1"

      certificate-file="unilic_client_*.crt"

      ca-file="unilic_ca.crt"

   />

 

For further reference to the license server, visit

 

http://unimrcp.org/licserver

4       Configuration Steps

This section outlines common configuration steps.

4.1      Using Default Configuration

The default configuration should be sufficient for the general use.

4.2      Specifying LUIS Model

A LUIS model is referenced by the corresponding App ID.

 

The App ID can be specified globally in the configuration file umsazurebot.xml by means of the parameter appid in the element <streaming-recognition>. For example:

 

   <streaming-recognition

      interim-results="true"

      start-of-input="service-originated"

      language="en-US"

      max-alternatives="1"

      appid="abcdefgh-ijkl-mnop-qrst-vwxyz12345678"

   />

 

The App ID can also be specified per individual MRCP RECOGNIZE request using one of the alternate methods listed below.

Built-in grammar

builtin:speech/transcribe?appid=abcdefgh-ijkl-mnop-qrst-vwxyz12345678

 

Vendor-Specific-Parameters

Vendor-Specific-Parameters: appid=abcdefgh-ijkl-mnop-qrst-vwxyz12345678

 

SRGS XML grammar

<grammar mode="voice" root="transcribe" version="1.0"

            xml:lang="en-US"

            xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>

    <meta name="appid" content="abcdefgh-ijkl-mnop-qrst-vwxyz12345678"/>

    <rule id="main"> <one-of/ > </rule>

</grammar>

4.3      Specifying Intents

A particular intent or a list of intents in the LUIS model can optionally be specified in the configuration file umsazurebot.xml by means of the parameter intents in the element <streaming-recognition>. For example:

 

   <streaming-recognition

      interim-results="true"

      start-of-input="service-originated"

      language="en-US"

      max-alternatives="1"

      appid="abcdefgh-ijkl-mnop-qrst-vwxyz12345678"

      intents="BookFlight, RoomReservation"

   />

 

The intents can also be specified per individual MRCP RECOGNIZE request using one of the alternate methods listed below.

Built-in grammar

builtin:speech/transcribe?intents=BookFlight, RoomReservation

 

Vendor-Specific-Parameters

Vendor-Specific-Parameters: intents=BookFlight, RoomReservation

 

SRGS XML grammar

<grammar mode="voice" root="transcribe" version="1.0"

            xml:lang="en-US"

            xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>

    <meta name="intents" content="BookFlight, RoomReservation"/>

    <rule id="main"> <one-of/ > </rule>

</grammar>

4.4      Specifying Recognition Language

Recognition language can be specified by the client per MRCP session by means of the header field Speech-Language set in a SET-PARAMS or RECOGNIZE request. Otherwise, the parameter language set in the configuration file umsazurebot.xml is used. The parameter defaults to en-US.

The recognition language can also be set by the attribute xml:lang specified in the SRGS XML grammar. For example:

 

<?xml version="1.0" encoding="UTF-8"?>

<grammar mode="voice" root="transcribe" version="1.0"

          xml:lang="en-AU"

          xmlns="http://www.w3.org/2001/06/grammar">

   <meta name="scope" content="builtin"/>

   <rule id="transcribe"><one-of/></rule>

</grammar>

 

4.5      Specifying Sampling Rate

Sampling rate is determined based on the SDP negotiation. Refer to the configuration guide of the UniMRCP server on how to specify supported encodings and sampling rates to be used in communication between the client and server.

4.6      Specifying Speech Input Parameters

While the default parameters specified for the speech input detector are sufficient for the general use, various parameters can be adjusted to better suit a particular requirement.

·         speech-start-timeout

This parameter is used to trigger a start of speech input. The shorter is the timeout, the sooner a START-OF-INPUT event is delivered to the client. However, a short timeout may also lead to a false positive. Note that if the start-of-input parameter in the ws-streaming-recognition is set to service-originated, then a START-OF-INPUT event is sent to the client at a later stage, upon reception of a speech.startDetected response from the service.

·         speech-complete-timeout

This parameter is used to trigger an end of speech input. The shorter is the timeout, the shorter is the response time. However, a short timeout may also lead to a false positive.

Note that both events, an expiration of the speech complete timeout and a speech.endDetected response delivered from the service, are monitored to trigger an end of speech input, on whichever comes first basis. In order to rely solely on an event delivered from the speech service, the parameter speech-complete-timeout needs to be set to a higher value.

·         vad-mode

This parameter is used to specify an operating mode of the Voice Activity Detector (VAD) within an integer range of [0 … 3]. A higher mode is more aggressive and, as a result, is more restrictive in reporting speech. The parameter can be overridden per MRCP session by setting the header field Sensitivity-Level in a SET-PARAMS or RECOGNIZE request. The following table shows how the Sensitivity-Level is mapped to the vad-mode.

 

Sensitivity-Level

Vad-Mode

[0.00 ... 0.25)

0

[0.25 … 0.50)

1

[0.50 ... 0.75)

2

[0.75 ... 1.00]

3

 

4.7      Specifying DTMF Input Parameters

While the default parameters specified for the DTMF input detector are sufficient for the general use, various parameters can be adjusted to better suit a particular requirement.

·         dtmf-interdigit-timeout

This parameter is used to set an inter-digit timeout on DTMF input. The parameter can be overridden per MRCP session by setting the header field DTMF-Interdigit-Timeout in a SET-PARAMS or RECOGNIZE request.

·         dtmf-term-timeout

This parameter is used to set a termination timeout on DTMF input and is in effect when dtmf-term-char is set and there is a match for an input grammar. The parameter can be overridden per MRCP session by setting the header field DTMF-Term-Timeout in a SET-PARAMS or RECOGNIZE request.

·         dtmf-term-char

This parameter is used to set a character terminating DTMF input. The parameter can be overridden per MRCP session by setting the header field DTMF-Term-Char in a SET-PARAMS or RECOGNIZE request.

4.8      Specifying No-Input and Recognition Timeouts

·         noinput-timeout

This parameter is used to trigger a no-input event. The parameter can be overridden per MRCP session by setting the header field No-Input-Timeout in a SET-PARAMS or RECOGNIZE request.

·         input-timeout

This parameter is used to limit input (recognition) time. The parameter can be overridden per MRCP session by setting the header field Recognition-Timeout in a SET-PARAMS or RECOGNIZE request.

4.9      Specifying Vendor-Specific Parameters

The following parameters can optionally be specified by the MRCP client in SET-PARAMS, DEFINE-GRAMMAR and RECOGNIZE requests via the MRCP header field Vendor-Specific-Parameters.

 

Name

Unit

Description

start-of-input

String

Specifies the source of start of input event sent to the client (use "service-originated" for an event originated based on a first-received interim result and "internal" for an event determined by plugin).

alternatives-below-threshold

Boolean

Specifies whether to return speech recognition result alternatives with the confidence score below the confidence threshold.

speech-start-timeout

Time interval [msec]

Specifies how long to wait in transition mode before triggering a start of speech input event.

interim-result-timeout

Time interval [msec]

Specifies a timeout between interim results containing transcribed speech. If the timeout is elapsed, input is considered complete. The timeout defaults to 0 (disabled).

appid

String

The application id of the LUIS model, if used.

intents

String

A comma-separated list of intents to be used from the specified LUIS model, if any.

 

All the vendor-specific parameters can also be specified at the grammar-level via a built-in or SRGS XML grammar.

 

The following example demonstrates the use of a built-in grammar with the vendor-specific parameters alternatives-below-threshold and speech-start-timeout set to true and 100 correspondingly.

 

builtin:speech/transcribe?alternatives-below-threshold=true;speech-start-timeout=100

 

The following example demonstrates the use of an SRGS XML grammar with the vendor-specific parameters alternatives-below-threshold and speech-start-timeout set to true and 100 correspondingly.

 

<grammar mode="voice" root="transcribe" version="1.0" xml:lang="en-US" xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>

    <meta name="alternatives-below-threshold" content="true"/>

    <meta name="speech-start-timeout" content="100"/>

    <rule id="transcribe">

        <one-of ><item>blank</item></one-of>

    </rule>

</grammar>

 

4.10 Maintaining Utterances

Saving of utterances is not required for regular operation and is disabled by default. However, enabling this functionality allows to save utterances sent to the service and later listen to them offline.

The relevant settings can be specified via the element utterance-manager.

·         save-waveforms

Utterances can optionally be recorded and stored if the configuration parameter save-waveforms is set to true. The parameter can be overridden per MRCP session by setting the header field Save-Waveforms in a SET-PARAMS or RECOGNIZE request.

·         purge-existing

This parameter specifies whether to delete existing waveforms on start-up.

·         max-file-age

This parameter specifies a time interval in minutes after expiration of which a waveform is deleted. If set to 0, there is no expiration time specified.

·         max-file-count

This parameter specifies the maximum number of waveforms to store. If the specified number is reached, the oldest waveform is deleted. If set to 0, there is no limit specified.

·         waveform-base-uri

This parameter specifies the base URI used to compose an absolute waveform URI returned in the header field Waveform-Uri in response to a RECOGNIZE request.

·         waveform-folder

This parameter specifies a path to the directory used to store waveforms in. The directory defaults to ${UniMRCPInstallDir}/var.

4.11 Maintaining Recognition Details Records

Producing of recognition details records (RDR) is not required for regular operation and is disabled by default. However, enabling this functionality allows to store details of each recognition attempt in a separate file and analyze them later offline. The RDRs ate stored in the JSON format.

The relevant settings can be specified via the element rdr-manager.

·         save-records

This parameter specifies whether to save recognition details records or not.

·         purge-existing

This parameter specifies whether to delete existing records on start-up.

·         max-file-age

This parameter specifies a time interval in minutes after expiration of which a record is deleted. If set to 0, there is no expiration time specified.

·         max-file-count

This parameter specifies the maximum number of records to store. If the specified number is reached, the oldest record is deleted. If set to 0, there is no limit specified.

·         record-folder

This parameter specifies a path to the directory used to store records in. The directory defaults to ${UniMRCPInstallDir}/var.

5       Recognition Grammars and Results

5.1      Using Built-in Speech Transcription

For generic speech transcription, having no speech contexts defined, a pre-set identifier transcribe must be used by the MRCP client in a RECOGNIZE request as follows:

 

builtin:speech/transcribe

 

The name of the identifier transcribe can be changed from the configuration file umsazurebot.xml.

 

Speech contexts are defined in the configuration file umsazurebot.xml and available since Azure SR 1.5.0. A speech context is assigned a unique string identifier and holds a list of phrases.

 

Below is a definition of a sample speech context directory:

 

      <speech-context id="directory" speech-complete="true">

         <phrase>call Steve</phrase>

         <phrase>call John</phrase>

         <phrase>dial 5</phrase>

         <phrase>dial 6</phrase>

      </speech-context>

 

Which can be referenced in a RECOGNIZE request as follows:

 

builtin:speech/directory

 

The prefixes builtin:speech and builtin:grammar can be used interchangeably as follows:

 

builtin:grammar/directory

 

5.2      Using Built-in DTMF Grammars

Pre-set built-in DTMF grammars can be referenced by the MRCP client in a RECOGNIZE request as follows:

 

builtin:dtmf/$id

 

Where $id is a unique string identifier of the built-in DTMF grammar.

 

Note that only a DTMF grammar identifier digits is currently supported.

 

Built-in DTMF digits can also be referenced by metadata in SRGS XML grammar. The following example is equivalent to the built-in grammar above.

 

<grammar mode="dtmf" root="digits" version="1.0"

          xml:lang="en-US"

          xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>

    <rule id="digits"><one-of/></rule>

</grammar>

 

Where the root rule name identifies a built-in DTMF grammar.

5.3      Retrieving Results

Results received from the Azure service are transformed to a certain data structure and sent to the MRCP client in a RECOGNITION-COMPLETE event. The way results are composed can be adjusted via the <results> element in the configuration file umsazurebot.xml.

NLSML Format

If the format attribute is set to standard, which is the default setting, then the header filed Content-Type is set to application/x-nlsml and the body of a RECOGNITION-COMPLETE event is set to an NSLML result composed as follows.

<input>

The <input> element in an NLSML result is set to the transcribed text.

<instance>

By default, the <instance> element in the NLSML result is composed based on an XML representation of the returned intent. This behavior can be adjusted via the tag-format attribute, which accepts the following values.

·         semantics/xml

The default setting. The intent is represented in XML.

·         semantics/json

The intent is represented in JSON.

·         swi-semantics/xml

The intent is set in an inner <SWI_meaning> element being represented in XML.

·         swi-semantics/json

The intent is set in an inner <SWI_meaning> element being represented in JSON.

JSON Format

If the format attribute is set to json, then the header filed Content-Type is set to application/json and the body of a RECOGNITION-COMPLETE event is set to a JSON representation of the intent.

 

The format attribute can be specified by the MRCP client per individual MRCP RECOGNIZE request as a query input attribute to the built-in speech grammar. For example:

 

builtin:speech/transcribe?format=json

 

The format attribute can also be specified in SRGS XML grammar. For example:

 

<grammar mode="voice" root="transcribe" version="1.0"

            xml:lang="en-US"

            xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>

    <meta name="format" content="json"/>

    <rule id="main"> <one-of/ > </rule>

</grammar>

 

6       Monitoring Usage Details

The number of in-use and total licensed channels can be monitored in several alternate ways. There is a set of actions which can take place on certain events. The behavior is configurable via the element monitoring-agent, which contains two event handlers: usage-change-handler and usage-refresh-handler.

 

While the usage-change-handler is invoked on every acquisition and release of a licensed channel, the usage-refresh-handler is invoked periodically on expiration of a timeout specified by the attribute refresh-period.

 

The following actions can be specified for either of the two handlers.

6.1      Log Usage

The action log-usage logs the following data in the order specified.

·         The number of currently in-use channels.

·         The maximum number of channels used concurrently.

·         The total number of licensed channels.

The following is a sample log statement, indicating 0 in-use, 0 max-used and 2 total channels.

 

[NOTICE] AZUREBOT Usage: 0/0/2

 

6.2      Update Usage

The action update-usage writes the following data to a status file umsazurebot-usage.status, located by default in the directory ${UniMRCPInstallDir}/var/status.

·         The number of currently in-use channels.

·         The maximum number of channels used concurrently.

·         The total number of licensed channels.

·         The current status of the license permit.

·         The license server alarm. Set to on, if the license server is not available for more than one hour; otherwise, set to off. This parameter is maintained only if the license server is used.

The following is a sample content of the status file.

 

in-use channels: 0

max used channels: 0

total channels: 2

license permit: true

licserver alarm: off

 

6.3      Dump Channels

The action dump-channels writes the identifiers of in-use channels to a status file umsazurebot-channels.status, located by default in the directory ${UniMRCPInstallDir}/var/status.

 

7       Usage Examples

7.1      Speech Transcription

This examples demonstrates how to perform speech recognition by using a RECOGNIZE request.

 

C->S:

 

MRCP/2.0 336 RECOGNIZE 1

Channel-Identifier: 6e1a2e4e54ae11e7@speechrecog

Content-Id: request1@form-level

Content-Type: text/uri-list

Cancel-If-Queue: false

No-Input-Timeout: 5000

Recognition-Timeout: 10000

Start-Input-Timers: true

Confidence-Threshold: 0.87

Save-Waveform: true

Content-Length: 25

 

builtin:speech/transcribe

 

S->C:

 

MRCP/2.0 83 1 200 IN-PROGRESS

Channel-Identifier: 6e1a2e4e54ae11e7@speechrecog

 

 

S->C:

 

MRCP/2.0 115 START-OF-INPUT 1 IN-PROGRESS

Channel-Identifier: 6e1a2e4e54ae11e7@speechrecog

Input-Type: speech

 

 

S->C:

 

MRCP/2.0 498 RECOGNITION-COMPLETE 1 COMPLETE

Channel-Identifier: 6e1a2e4e54ae11e7@speechrecog

Completion-Cause: 000 success

Waveform-Uri: <http://localhost/utterances/utter-6e1a2e4e54ae11e7-1.wav>;size=20480;duration=1280

Content-Type: application/x-nlsml

Content-Length: 214

 

<?xml version="1.0"?>

<result>

  <interpretation grammar="builtin:speech/transcribe" confidence="0.777">

    <instance>

      <object><string name="query">book a room</string>

        <object name="topScoringIntent">

          <string name="intent">RoomReservation.Reserve</string>

          <number name="score">0.77710616600000004</number>

        </object>

        <array name="entities"></array>

      </object>

    </instance><input mode="speech">book a room</input>

  </interpretation>

</result>

 

 

8       Sequence Diagram

The following sequence diagram outlines common interactions between all the main components involved in a typical recognition session performed over MRCPv2.

9       References

9.1      Microsoft Azure

·         LUIS

·         Recognize Speech, Intents and Entities

 

9.2      Specifications

·         Speech Recognizer Resource

·         NLSML Results