People with a hearing impairment have a hard time taking part in conversations with other people with good ears. As a rule most people are not abble to use sign language. Also people with severe hearing problems often aren't good at using proper intonation due to a lack of significant feedback. This results in strongly diminished potential to take part in e.g. round table discussions.
In theory speech technology can solve these issues. The perfect program would offer the following:
The speech recognition part should be aware of the speech synthesis part.
It would not be realistic to think that this can be accomplished by one student in one smmer, but it might be interesting to create a small part of it.
Does open source voice recognition even exist?
Anyway enough challenges in just one 'little' application.
The amount of voxforge wiki's makes it hard to search information in it. Also it's still using old logo. About speech verification, the most advanced project is Alize:
Link to download ALIZE:
>Also it's still using old logo.
I still kind of like the old logo, and figured it would be OK for the Trac dev wiki .... :)
I know I'm commenting 4 years later, but in that space of time the Alize software has gone offline. The links are no good, and I can find no other place to download it.
If you still have a copy downloaded, would you upload it somewhere and share the link? That would be tremendous.
I'm a student trying to learn about this sort of thing. Thanks!