I would like to see speech synthesis (text to speech), speech recognition and perhaps speech recording in Scratch.
The recognition does not have to be general -- just provide a Case statement.
Wouldn't it be cool to be able to converse with sprites?
Recording for playback might also be useful, but synthesis and recognition would be neat for sure.
Years ago I edited a book series to which Chris Schmandt was a contributor -- might be help?
Larry
Offline
Do you have any idea how complicated that would be? Speech recognition software is only just starting to develop. Text to speech is easy, but actually recognising human speech, a bunch of guys in the MIT media lab can't do that.
Offline
WeirdF wrote:
Do you have any idea how complicated that would be? Speech recognition software is only just starting to develop. Text to speech is easy, but actually recognising human speech, a bunch of guys in the MIT media lab can't do that.
My mac does that. I can say "Log me out" and it will log me out. It is a bit picky though...
Offline
Please think before you post. This is impossible.
Speech recognition will be impossible for many years! Once it gets out, it's not like Scratch will be the first to have it!
It's been said that making speech is like squeezing toothpaste out of a tube. Speech recognition is like fitting it all back in.
Anyway, I'm putting this in my sticky.
Offline
The-Whiz wrote:
WeirdF wrote:
Do you have any idea how complicated that would be? Speech recognition software is only just starting to develop. Text to speech is easy, but actually recognising human speech, a bunch of guys in the MIT media lab can't do that.
My mac does that. I can say "Log me out" and it will log me out. It is a bit picky though...
No, that is teaching it to recognise certain phrases with your voice, what he's talking about is actually the computer understanding the English language. We can do it easily if it recognises your voice.
Offline
WeirdF wrote:
The-Whiz wrote:
WeirdF wrote:
Do you have any idea how complicated that would be? Speech recognition software is only just starting to develop. Text to speech is easy, but actually recognising human speech, a bunch of guys in the MIT media lab can't do that.
My mac does that. I can say "Log me out" and it will log me out. It is a bit picky though...
No, that is teaching it to recognise certain phrases with your voice, what he's talking about is actually the computer understanding the English language. We can do it easily if it recognises your voice.
True, but you need to teach it your way of talking. And people can say words differently, so you'd probably have to speak in a fragmented way.
Offline