Category: Geeks r Us
Hello friends, I am currently have my thesis. My group is planning to develop a speech synthesizer for LINUX operating system as well as it's office. What do you think? I made this plan because I want to help my fellow visually impaired in our country who can't aford to buy an expensive screen reader available from the market. I really need your suggestion. Please send me a private message or reply to this topic. Thank you very much.
Evangeline
Hi,
is your goal to develop a synthesizer or screen reader? In either case, I would suggest that rather than reinventing the wheel, you would better serve the community by concentrating your efforts on making a contribution to one of the projects that are already out there. If your goal is a speech synthesizer, then look at doing some work on the Espeak project. If you want to do screen reader development, then consider giving the Orca project a look.
My opinion is that with Espeak and Orca, we already have a very good foundation in terms of accessibility for years to come, and the addition of more projects to this space would just delude the pool of developers for everyone involved.
Thanks for the advice. Well, our goal is to have a local screen reader. That means to have a screen reader that have a Philippine English accent, and also to speek Filipino as well.
hmm... there are various kinds out there already; however the goal as it was already mentioned is to contribute to what it is out there. As far as accents... you might want to develop your own synth as well. i nice thingwould be to try and mirror some of the features from oralux; (one of them being the system will begin to speak as soon as it is booted) . It'd be nice to see a bios that would speak to you as well; however that is beyond the scope of linux. Cheers
Give me something the size and quolity of espeak but the speed of something like orfeyus and i'll be happy.
Evangeline,
You might look at the FreeTTS project at Carnagie Mellon University in the US.
The Sourceforge URL is http://freetts.sourceforge.net/docs/index.php.
They work with the speech synthesis folks in Edinburgh.
The reason I suggest this is that they specifically address the creation of additional voices. Apparently, it is not easy, but it is possible, with their system. To get a new voice would not be a big programming issue. More of an issue of getting various programs to work together and pushing a bunch of files around.
We use FreeTTS in the blind accessible games we develop.
It runs fine on Linux and Mac since it's written in Java.
John Bannick
Chief Technology Officer
7-128 Software
www.7128.com
Well, thank you very much for your advice. I will check it out all your suggestion, and if it is okay to you guys, I will send you a private message for more information. I really appreciate your help.
Evangeline