Eric wrote:
why but I get significantly worse results when dictating directly into open office versus something like DragonPad.
For speech recognition both of them require significantly (at least triple) the stated requrements, for adequte performance.
My suspicion is that open office takes way more resources than simple editors and this causes problems with how speech recognition programs read data from the audio stream.
That doesn't only cause problems with speech recognition. It also causes problem with speech output. [This in one reasion why I say that the minimum acceptable system for running OOo 2.0.x is a 3GHZ machine with 4GB RAM. That way you have a fighting chance of getting acceptable performance.]
ViaVoice the ever helpful monopoly on desktop speech recognition) would work with open office and make a high accuracy, friendly to use interface but I'm not holding my breath.
a) What happened to the promise by IBM to opensource ViaVoice? b) What about using Sphinx? c) There is a group in Sun that is working on a Java Speech Recognition Engine. Either b or c should, in theory, be able to either be dropped into OOo, or work in tandem with OOo.
I think Microsoft would be a better group to work with because they seem to be actively reaching out to the handicapped community for not just speech recognition but the
You have got to be joking. The major reason that handicapped "like" Microsoft, is because of the third party support. The secondary reason is that the organizations that purchase the hardware and software neither fund, nor use them. xan jonathon -- Ethical conduct is a vice. Corrupt conduct is a virtue. Motto of Nacarima --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
