Hello VoxForge folks
I write here because I see the project as regards the Italian language does not proceed. This thing makes me sad .... In Italy there is no alternative open to the speech recognition software, and also in terms of proprietary software the situation is not convenient. Src software is a useful technology in many fields, especially to help the disabled but also to develop applications more frivolous. Italian developers do not have open source speech recognition engines is very hard work under these conditions. But Voxfoge have created excellent acoustic models, to create the Italian one, you need a lot of human resources to create the Italian model. I ask voxforge staff permission to create facebook group to use other social networks and other channels to promote the collection italian recording hours of Italian language. I am just geek but I'd be happy to help other Italian enthusiasts and developers to have a good SRS software open and running. greetings from Rome Carlo
Hello Etrurian, glad to hear you are a supporter of open source and voice. I don't speak on behalf of Voxforge, except in the sense that there is only one person who manages the VF resources (who is not me) and by extension Voxforge is us, the readers and contributors.
There are about 7,000 languages in use in the world; one person can't be expected to coordinate and build except in a remote, supervisory sense.
It's our job to do as you point out, provide avenues for people to contribute audio. I don't think you need anyone's permission to do that. Just go ahead and do it. Anyone can submit at any time.
One of the best things you can do is to both encourage submissions in your language and make sure that when the package is submitted to the VF collection that it is as clean, accurate and complete as possible. Then the Italian (and other) collection of audio will gain a reputation for ease of absorption into individual models. People will see it as being worth contributing to, and in a professional manner. Provided you have the audio and the metadata you don't need anyone to create a model for you.
Given this approach, what do you think other VF members apart from the guy who manages the servers can do to help you?
i'm working with Julius 4.1.5.. I'm working for University of Unisannio (Benevento, Italy).
My project is to develop a tool that can allow deaf students to follow university courses more easily.
If you want we can work together to make it more accessible to Italian users using Julius
This is my contact https://www.facebook.com/carlo.galliano