[jitsi-dev] Speech to Text Feature


vote for google speech api

Dne 20.5.2013 19:44, Justus Wolbert napsal(a):

Hello Developers at Jitsi,

Wouldn't it be great if Jitsi offered Speech to Text as one of it's features? This would enable people who are hearing impaired to follow a conference call. Or help people make notes during a conversation. It can also be used for the Voice Control feature.

But let me introduce myself. My name is Justus and I'm a student at the faculty for Art, Media & Technology of the School of the Arts in Utrecht. System development for audio & music is one of my specialisations and I would really like to contribute to Jitsi. My idea is to implement CMUSphinx <http://cmusphinx.sourceforge.net/> in Jitsi to allow people who are hearing impaired join in on conference calls without having one person typing out what is being said. It can also enable you to make Voice Control possible for platforms with great Speech recognition (96%!).

Currently I am still a newbie in coding but I have some experience in Java and C++. If you like the idea and other developers want to work at it in a fast pace it is fine by my, it would probably be a great learning experience for me. But if you'd like the idea but want me to do the development by myself then that's also great but please keep in mind it might take me a while to develop.

At last I would like to complement you for implementing about Spatial Audio which is often an overlooked issue in these platforms. Hope to hear from you soon.

Marek Cervenka