ofxSpeech

May 30th 2009 - by Ruben Duque (latrokles) Updated - Nov 4 2013 - Jeff Crouse (jefftimesten)

This addon is a wrapper around the Apple speech recognition and speech synthesis APIs. Ergo, it is Mac only.

Check out this video to see the "zoo" example in action. This will give you a realistic expectation of the accuracy of the library.

Installation:

1. Place ofxSpeech in the addons folder.

Adding to a Project:

1. Add the ofxSpeech to your project in the IDE (Xcode presumably). 2. Add #include "ofxSpeech.h" in your testApp.h

ofxSpeechRecognizer usage:

1. testApp.h Add a function to your testApp that takes a string by reference, this function will get called when a word is recognized. Also declare a ofxSpeechRecognizer object. void speechRecognized(string & wordRecognized); ofxSpeechRecognizer recognizer;

1. setup In the testApp.cpp setup() function, add the function mentioned above as a listener to a speechRecognized event and init the recognizer.

ofAddListener(recognizer.speechRecognizedEvent, this, &testApp::speechRecognized);

recognizer.initRecognizer();

You can load a list of words to recognize from a file in your data directory containing each word in a separate line.

recognizer.loadDictionaryFromFile("dictionary.txt");

Alternatively you can have a vector of words and add them that way

vector<string> words; words.push_back("red"); words.push_back("green"); words.push_back("black"); recognizer.loadDictionary(words);

startListening for words

recognizer.startListening();

1. speechRecognized In the function defined in testApp.cpp place code to be executed once a given word is detected

void testApp::speechRecognized(string & wordRecognized) { if(wordRecognized == "red") ofBackground(255, 0, 0);

if(wordRecognized == "green") ofBackground(0, 255, 0); }

the recognized can be stopped

recognizer.stopListening();

ofxSpeechSynthesizer usage:

//define and create an ofxSpeechSynthesizer object ofxSpeechSynthesizer synthesizer;

//list available voices in the system synthesizer.listVoices();

//initalize synthesizer with voice, if no voice is given the default system voice is used. synthesizer.initSynthesizer("Whisper");

//speak a word or sentence synthesizer.speakPhrase("hello, this a test of the synthesizer");

//to turn on spelling character by character synthesizer.setCharacterByCharacter(true);

//to turn off spelling character by character synthesizer.setCharacterByCharacter(false);

//to turn on digit by digit for numbers synthesizer.setDigitByDigit(true)

//to turn off digit by digit for numbers synthesizer.setDigitByDigit(true)

//to pause speaking synthesizer.pauseSpeaking();

//to resume speaking synthesizer.continueSpeaking();

//to stop speaking completely synthesizer.stopSpeaking();