[Error: Irreparable invalid markup ('<q?only>') in entry. Owner must fix manually. Raw contents below.]
<q?Only ten years ago when it was still cutting edge. :)</q>
Same here, only I'd phrase that as "Only ten years ago when it still really sucked".
I would really like to hope the large vocab stuff has come a long way since then, but my exposure to voice-enabled car navi units at my current job leads me to seriously doubt it. )-:
Actually, it was more like 15 years ago... And yes, it really sucked. It sucked like a 15 solar mass black hole. It's probably down to about 9 solar masses by now.
Actually I do, I did the speech part of our robot project term. I looked a bit at the time, this was two years ago, but I settled on the software suite called Sphinx from Carnegie Mellon University. The link to their main site is: http://www.speech.cs.cmu.edu/
On that site you'll find a ton of links, and perhaps if you contact them they might refer you to others who have done specialization towards medical speech recognition. It is after all, a much smaller subset to cater a program for a specific and contained jargon than for the entire English Language.
DragonSpeaking Pro I am guessiing, although maybe they changed the name. I didn't like it. I found ViaVoice a little easier to train but still a pain. I have not yet used iListen so dunno about that one but I think it's mac-only.
Hmm, do you train it or does it train you? Back when I worked at Dragon, you could tell how long any given person had been working there, because after a while they all sound exactly the same.
It was Dragon Dictate I played with way back when it only ran on the DEC box. You trained it and trained it and trained it and prayed that you didn't catch a cold because you'd have to train it again...
You don't know me and I know this post was old, but if you are still looking for a medical voice recognition software, try Dragon... but make sure you buy the medical module for it. I work at a hospital and we are looking for a new information system for our emergency department. I have seen some compelling demonstrations at a couple of other hospitals who use it. Good luck.
no subject
Date: 2005-05-17 10:37 am (UTC)no subject
Date: 2005-05-17 11:38 am (UTC)But RGU has a medical AI group who do such stuff. I'll email you with a contact.
no subject
Date: 2005-05-17 03:51 pm (UTC)Same here, only I'd phrase that as "Only ten years ago when it still really sucked".
I would really like to hope the large vocab stuff has come a long way since then, but my exposure to voice-enabled car navi units at my current job leads me to seriously doubt it. )-:
Good luck...
no subject
Date: 2005-05-17 05:08 pm (UTC)no subject
Date: 2005-05-17 12:22 pm (UTC)no subject
Date: 2005-05-17 01:20 pm (UTC)http://www.speech.cs.cmu.edu/
On that site you'll find a ton of links, and perhaps if you contact them they might refer you to others who have done specialization towards medical speech recognition. It is after all, a much smaller subset to cater a program for a specific and contained jargon than for the entire English Language.
no subject
Date: 2005-05-17 07:02 pm (UTC)no subject
Date: 2005-05-17 10:01 pm (UTC)no subject
Date: 2005-05-17 10:56 pm (UTC)no subject
Date: 2005-05-18 04:12 am (UTC)p.s. "It's hard to wreck a nice beach".
no subject
Date: 2005-05-18 09:57 am (UTC)no subject
Date: 2005-05-18 01:24 pm (UTC)no subject
Date: 2005-08-21 01:54 pm (UTC)