What is Perlbox Voice Application Framework?
The Perlbox Voice Application Framework is simply a set of libraries designed to facilitate the rapid creation and deployment of voice enabled applications for Unix type systems. This library leverages as many existing open source tools as possible in order to provide stability and robustness.
Perlbox-Desktop was the first offering from perlbox and the first desktop written entirely in Perl/Tk. It was used to test several concept relating to the utilization of Perl in mid to large scale applications, as well as the viability of open source command response systems. Perlbox-Desktop was met with great success, clocking thousands of downloads in the year after it's initial release. Perlbox-Desktop is a separate product, at present, from the Perlbox Voice Application Framework and its related products. In the coming months, Perlbox-Desktop will be reworked to utilize the dynamic framework.
The Sphinx Group at Carnegie Mellon University is committed to releasing the long-time, DARPA-funded Sphinx projects widely, in order to stimulate the creation of speech-using tools and applications, and to advance the state of the art both directly in speech recognition, as well as in related areas including dialog systems and speech synthesis.
In simple terms, Sphinx2 is the program that changes spoken words into machine readable text. Perlbox Voice Application Framwork interfaces Sphinx2 in several ways. In the first way, Perlbox Voice Application Framework translates the output of Sphinx2 into commands that the user sets up and then executes these commands. The second way that Perlbox Voice Application Framework interfaces with Sphinx2 is by dynamically creating new language models (vocabularies) for the Sphinx2 listening agent.
Festival is a general multi-lingual speech synthesis system developed at CSTR. It offers a full text to speech system with various APIs, as well an environment for development and research of speech synthesis techniques. It is written in C++ with a Scheme-based command interpreter for general control.
In simple terms, Festival is a program which takes text and says it in a 'robot' voice. Perlbox Voice Application Framework utilizes Festival to inform the user of the status of the application and other such items.
No. Current installations will work fine, though you may need to set the path for Sphinx2.
The ability to specify vocal responses to commands you make is a part of Perlbox Voice Application Framework's pseudo command subsystem. When you create a new vocabulary in your application, in the command section, specify something like say "Phrase to say". Then when you say whatever you specified in the speech portion of the vocabulary entry, the computer will respond by speaking "Phrase to say". This should be covered in more detail in your applications tutorial.
This should be completely covered in your applications tutorial.
This one is easy! Find the directory $HOME/.perlbox-voice, make sure it exists, then remove it. The next time you start perlbox voice,
the configuration your version shipped with will be restored.
Not currently. Sphinx2 is designed for realtime, untrained listening and is therefore unsuitable. Other projects, like xvoice have tried this with ViaVoice and Sphinx, without a lot of success. We hope to try ourselves using Sphinx3.
No. I often say that you do not train sphinx how to listen, Sphinx trains you how to talk. Speak clearly, try several approaches untill you learn what it responds to.
If you have more questions, please email them to me at perlbox.org.