Ubuntu Wiki

(2011-03-04) Lightning talk on SpeechControl for Ubuntu Developer Week

intro:

Hello everyone,

I welcome you all for this short introduction to SpeechControl.

If you have questions, please wait a while, I will be brief, thank you.

This project started in November 2010 with a vision from wiki.ubuntu.com/hajour. I am Keimpe de Jong, better known as wiki.ubuntu.com/UndiFineD and her partner. We have 4 girls with either a form of AD(H)D, Dyslexia or poor sight.

SpeechControl is an accessibility project. wiki.ubuntu.com/SpeechControl It aims to make controlling your computer easier. In the distant future we hope to reach Star Trek like capabilities.

How it began ...

We began by writing capable people in this area, explaining hajour her vision and asked if they would be willing to contribute to it. Then we asked to be placed under the accessibility team and ubuntu beginners team flags.

• Ubuntu Accessibility

• Ubuntu Beginners Team

And the roundup of currently capable developers with an interest amazed us. We are grateful for all our team members that help us and we welcome new ones with open arms.

Key features:

• Speech recognition: Using the available tools at our disposal, SpeechControl will be able to comprehend your spoken word.

• Preconfigured Command Execution: knowing what to do what you ordered it to.
• Virtual Assistant: Repetitive task guidance improve for faster and smoother process execution.
• Speech Synthesis: More commonly known at TTS (Text-To-Speech), the voice that communicates with the end-user will be as natural and inviting as technology permits.

So how is this done ?

After some research we found Simon to currently be the best capable open source software capable in controlling a computer. www.Simon-Listens.org

Simon is / was not in the Debian repositories, due to licensing issues. We believe that is cleared up by now and Debian can take on Simon, as it uses the same license type as SSH.

So what is left to be done ?

Well there are several disabilities, and we wish to make it good for everyone. In order to do that we need to:

• extend Simon (API work),
• talk to all the system busses (dbus, at-spi2, ...),
• analyse Speech-to-text and make some context (Wintermute),
• execute commands based upon input given,
• help to reduce repeating work,
• communicate back to the user (via Text-to-Speech).

Blueprints:

A lot of blueprints need work, we would love to gather all your input and make things better.

Request come in for simple but related tasks, like remind a user to take their medicine, or read out an e-book.

Besides this, it bothered us we could not properly have an accessible meeting. So there is work being done on a Speech capable chatcliënt plugin.

The possibilities are open and endless, and would make using your computer so much easier. To prevent a wild bloom of small projects, our team currently focuses on specifying blueprints and defining the path to take. Our initial goal is a proof of concept and optimize later on.

Progress:

In a few months good work has been done; the team is slowly growing and could use extra help writing specifications. After that we will develop all the libraries and applications. If you would like to learn more feel free to come and visit us on #SpeechControl read more about us on wiki.ubuntu.com/SpeechControl our team is located here: launchpad.net/~speechcontrolteam

I would like to thank everyone in our team, for the great work they have done so far.

If you have questions feel free to ask them now, but for the sake of the sessions I recommend longer talks to be held in #SpeechControl.