UE3:SpeechRecognition (UDK)

Object >> SpeechRecognition

Package: 

Engine

This class in other games:

UT3

Properties

Property group 'SpeechRecognition'

ConfidenceThreshhold

Type: float

Threshhold below which the recognised word will be ignored

Default value: 50.0

Language

Type: string

Language to recognise data in

Default value: "INT"

Vocabularies

Type: array<RecogVocabulary>

Array of vocabularies that can be swapped in and out

Internal variables

bDirty

Type: bool

Modifiers: duplicatetransient, transient

Whether this object has been altered

bInitialised

Type: bool

Modifiers: duplicatetransient, transient

Whether the object was successfully initialised or not

FnxVoiceData

Type: pointer

Modifiers: duplicatetransient, native, const

Cached pointers to Fonix data

InstanceData

Type: RecogUserData

Array size: 4

Cached user data - max users

UserData

Type: array<byte>

Cached user data

VoiceData

Type: array<byte>

Cached neural net data

WorkingVoiceData

Type: array<byte>

Working copy of neural net data

Structs

RecognisableWord

Modifiers: native

int Id 

string ReferenceWord 

This is the reference word, which is returned upon recognition. e.g. "Loque". Does not need to be unique.

string PhoneticWord 

This is the word string that is passed into the recognition. e.g. "Loke"

RecogUserData

Modifiers: native

int ActiveVocabularies 

Bitfield of active vocabularies

array<byte> UserData 

Workspace for recognition data

RecogVocabulary

Modifiers: native

array<RecognisableWord> WhoDictionary 

Arrays of words that can be recognised - note that words need an ID unique among the contents of all three arrays

array<RecognisableWord> WhatDictionary 

array<RecognisableWord> WhereDictionary 

string VocabName 

Name of vocab file

array<byte> VocabData 

Cached processed vocabulary data

array<byte> WorkingVocabData 

Working copy of vocab data