|
speech
|
The module based on Microsoft Speech API. It is made to recognize speech based on runtime grammar in different formats: XML W3C specification "custom" simple format closed to the one used in the CSLU RAD Toolkit. The class RobotGrammarManager maintains a dictionary of various vocabulories usefull in HRI (Objects, Agents, Actions...) which can be expanded at runtime and can be used in grammars.
Microsoft speech sdk http://msdn.microsoft.com/en-us/library/hh361572(v=office.14).aspx
–grammarFile specify the name of the grammar file to load at startup –talkback if specified the program will use TTS to repeat every sentence recognized
If iSpeak is running then all text to speech commands will be forwarded to iSpeak.
/modulename/recog/continuous:o broadcast commands recognized by the asynchronous grammar (see rpc commands "asynrecog" ) /moduleName/rpc handles rpc commands of type "tts" "sentence to say" Say a sentence
"asyncrecog"
"addGrammar"...
Add a simple grammar (uses current state of RGM) to the asynchronous grammar
"clear"... Reset the asynchrnous grammar to its original state
"recog"
"timeout" "ms"
Set the timeout for all recognition actions
"choices" "choice1" "choice2"... Run a recognition over the list of choices passed in arguments. Returns the word/sentence recognized
"dictation" Run a recognition of open dictation
"grammarXML" "xml formatted grammar" Run the recognition of a grammar specified in the W3C SRGS xml format (http://www.w3.org/TR/speech-grammar/)
"grammarSimple" "grammar to recognize" Run a recognition on a custom format using RobotGrammarManager, refer to RobotGrammarSpecification.txt
"RGM" "vocabulory" "get" "vocabuloryName" [NOT IMPLEMENTED YET] return a string containing all the vocabulory words
"add" "vocabuloryName" "word" Expand a given vocabulory with a specific word
"addAuto" "vocabuloryName" Expand a given vocabulory by putting the system in dictation mode and asking confirmation of what is being recognized
None.
None.
None.
Windows 7.
1.9.1