Home   Documentation   Demonstrations   Download

AMICO: Documentation / AMICO Speech Modules

 

Text-to-speech (TTS) and speech recognition services. TTS modules register for AMICO notifications, and pronounce text derived from AMICO variables using TTS templates (to see an example of a TTS template file click here.). Speech recognizers detect phrases said by the user, and update AMICO variables with recognized text.

 

 

Module How to run (from AMICO root; this root is also used in AMICO process runner). Description
English text-to-speech engine (FreeTTS) java -jar %AMICO_HOME%/modules/speech/free-tts/modules-freetts.jar <tts-template-file>

To see an example of a TTS template file click here.

Text-to-speech engine based on FreeTTS library. Pronounces requested text in English.
English, German and Tibetan text-to-speech engine (Mary)

java -jar %AMICO_HOME%/modules/speech/mary-tts-adapter/modules-mary-tts.jar <tts-template-file>

To see an example of a TTS template file click here.

Note: you must also run a Mary server.

Text-to-speech engine based on MARY Text-to-Speech system. Pronounces requested text in German, English and Tibetan.

Dutch text-to-speech engine (Nextens)

java -jar %AMICO_HOME%/modules/nextens/nextens-tts-adapter/modules-nextens-tts.jar <tts-template-file>

To see an example of a TTS template file click here.

Note: you must also run a Festival server.

Text-to-speech engine based on NeXTeNS Text-to-Speech synthesizer for Dutch. Pronounces requested text in Dutch.

English speech recognizer (Sphinx-4)

java -jar -Xmx100m %AMICO_HOME%/modules/amico-%AMICO_HOME%/modules/speech/sphinx-4/sensors-speech.jar <update-variable> <message> <gui> <config-file>

-Xmx100m parameter is used to create bigger heap space within Java virtual machine. Without this setting, OutOfMemoryError exception is often thrown.

All parameters are optional. Default values are (respectively): "speech-command", "", "nogui", "file:<current-directory>/conf/modules/sphinx-4/speechcommander.config.xml"


Example
:

java -jar -Xmx100m speech-command "Available speech commands: in, out, left, right, up, down" %AMICO_HOME%/modules/speech/sphinx-4/sensors-speech.jar gui conf/speech/sphinx-4/speechcommander.config.xml

English speech recognizer based on Sphinix-4 library. Send recognized speech phrases.

 

Validation Logos