Tutorial-6-sb

Description

SpeechbuilderTutorial MIT 6.893; SMA 5508 Spring 2004 Larry Rudolph Lecture Introduction Speaker Independent;Domain DependentWhat is a domain?a vocabulary (words)sentencesHow to define words?English spelling and pronunciation How to define sentences}Grammar2 MIT 6.893; SMA 5508 Spring 2004 Larry Rudolph Lecture Introduction SpeechbuilderGalaxy is the speech recognition systemSpeechbuilder is a tool to develop a domain for galaxyReal speech recognizers take a lot of work and detailed knowledge of all the components.Speechbuilder is great for prototyping3 MIT 6.893; SMA 5508 Spring 2004 Larry Rudolph Lecture Introduction Galaxy’s ComponentsLanguageGenerationHttpSpeech DialogueManagement SpeechBuilder ApplicationSynthesisServer (cgi)DatabaseAudio Hub ServerTCP SocketFrame Relay ApplicationServer (Python, Java, ...)ContextSpeechResolutionRecog.LanguageProcessing4 MIT 6.893; SMA 5508 Spring 2004 Larry Rudolph Lecture Introduction Speechbuilder APIGalaxy meaning representation provided through frame relayApplications connect via TCP socketsAPI provided in Python, Java, PerlGalaxy Frame RelayTCPSocketgalaxy.frame.Frame methods:galaxy.server.Server methods:getAction()Constructor(machine,port,ID)getAttribute(attr_name) Python classconnect()getText() galaxy.server ...

Informations

• Publié par: Sour
• Nombre de lectures: 29
• Langue: English

Extrait

Speechbuilder

Tutorial

 MIT 6.893; SMA 5508 Spring 2004 Larry Rudolph Lecture Introduction

T 6.893;  MI   8    AMS 055 4 00    riSp 2ngduloyrR L ra                ph  nI   erutceL      2 ontiucodtr

Grammar

How to define sentences}

What is a domain?

Speaker Independent; Domain Deendent

sentences

a vocabulary (words)

English spelling and pronunciation

How to define words?

e urctLeodtrIn   3  noitcury R Larph  udol                

Galaxy is the speech recognition system

Speechbuilder

Speechbuilder is a tool to develop a domain for galaxy

Real speech recognizers take a lot of work and detailed knowledge of all the components.

Speechbuilder is great for prototyping

MA S3;896.T MI          004 ng 2pSir     8  5 05

Speech Synthesis

4

Audio

0

Hub

 

Database Server

  

Galaxy’s Components

 

Dialogue Http Management SpeechBuilder Application Server (cgi)

 

Language Generation

     

Speech Recog.

a

Language Processing

L

Context Resolution

r

TCP Socket Frame Relay Application Server (Python, Java, ...)

r       Spring 208.6  ;39 AMS8055 IT Mn  4 tcoiorudI tner  tuec L                 hploduR y

  IMIn  e urtiucodtr        tceL    no 5    8     riSp 2ng.6 T;398AMS 055 ry Rudolph      00 4        L ra

Speechbuilder API

Application

Galaxy meaning representation provided through frame relay Applications connect via TCP sockets API provided in Python, Java, Perl

TCP Socket galaxy.frame.Frame methods: getAction() getAttribute(attr_name) Python class t g o e S t t T ri e n x g t( ( ) ) galaxy.server.Server Python class Python galaxy.frame.Frame API

Galaxy Frame Relay

galaxy.server.Server methods: Constructor(machine,port,ID) connect() processMessage(blocking) disconnect()

    pSirgn2 00 4893; SMA 5508   lodu  hp                ar L Rryrtdo  nIno 6  cuit        ure Lect  MIT 6.

would

a sample sentence: B A A A

<nt-2> == <nt-1> | NULL

<nt-1> B | <nt-2> A ==

Can you explain this to Grandma?

nt-1 --> nt-2 A --> nt-1 A --> nt-2 A A --> nt-1 A A ...

Grammar

a set of rules or productions

A, B, ...

a set of terminals

What is a grammar?

  IM T.6 5508   893; SMA 4002 gnirpS     Rryar L                hp  duloure Lect        trod  Inon  7 ucti

These are all “objects”

A set of terminals: lights, microwave, toaster, vcr, tv

So, “object” would be an attribute Another example

actually, a non-terminal that goes to a ter minal For example

Attributes think of them as: ter minals

dining room, living room, kitc hen

“room” is the attribute

Speechbuilder’s Grammar

“ “turn” is an action”

“on” is an example of an “onoff” attribute  

Each action is an example sentence

Sentence contains

Speechbuilder calls them “actions”

No complicated productions

What does a rule look like?

E.g. Turn on the lights

zero or more attributes

optional words

“lights” is an example of an “object” attribute

an “action” terminal

        40   g02h   dolpy RuLarr AMS ;398.6 TIM inpr S      0855tcoi n 8     L           nt Idurotuec  re 

        erutceL ront I  n ioctdu 9 398.MS ;55 A  80    pr Sg in0420          aLrr yRudolph         TI6  M 

Example after reduction All sentences for action turn   What gets sent to application

oitcudor 01  nectu   L Intre       h          uR yplod    rraL  04    inpr20g       S MS A5580 6.893;   MIT

<class name="object" type="Key">  <entry>(television | tv) {television}</entry>  <entry>lights</entry>  <entry>microwave</entry>  <entry>toaster</entry>  <entry>v c r {VCR}</entry> </class>

Domain XML example

ceuter  I tnorudction  11

Domain XML example <class name="onoff" type="Key">  <entry>lit {on}</entry>  <entry>off</entry>  <entry>on</entry> </class> <class name="turn" type="Action">  <entry>[can you] [please] turn all the lights off</entry>  <entry>[can you] [please] turn off all the lights</entry>  <entry>[can you] [please] turn off the (living room lights | lights in the living room)</entry>  <entry>[can you] [please] turn the (living room lights | lights in the living room) off</entry> </class> <class name="status" type="Action">  <entry>([can you] [please] tell me | do you know) (what | which) lights are on</entry>  <entry>([can you] [please] tell me | do you know) if the (lights in the kitchen | kitchen lights) are on</entry>  <entry>(is | are) the (dining room television | tv in the living room) On or Off</entry>  <entry>(is | are) the (dining room television | tv in the living room) on</entry> </class> <class name="good_bye" type "Action"> =  <entry>good bye</entry>  <entry>later</entry> </class> <class name="room" type="Key">  <entry>dining room</entry>  <entry>kitchen</entry>  <entry>living room</entry> </class>

aLrr yuRodpl h                 L8055    S   nirp20g   04          6.8 MITSMA 93;