The Word Cloud

• Sign Up Now!

No topics yet...

0 Photos

3 Member

3 Wall Posts

• 

  by Steve 2 years ago

  Thank you, Brian. I will have a look at Dragon, on your recommendation.
  
  I was hoping that a USB link (or some such) from an mp3 dictaphone would be able to provide a more stable reading source than simply playing the dictaphone into the computer microphone as if it were me speaking. I don't even know if what I'm suggesting is at all possible.
• 

  by Brian 2 years ago

  The technology has been worthwhile for many years now in my view. I am using a system I bought about 6 years ago and it is more than adequate for writing WORD documents.
  I have not tried direct translation from MP3 but my guess is that speech recognition would be more difficult in the absence of training. I think you need to try the system first. MP3 could be stage two, three or even not at all.
  Suggest you buy DragonDictate for about £50 from Amazon, install it and have a go. You need at least 1Gb RAM and a good microphone. It is very very easy to install.
• 

  by Steve 2 years ago

  Excellent. Just the Group I've been waiting for. I had a good look at speech recognition software a few months ago after noticing that the technology had finally reached a point of being worthwhile. I read a review that said the software bundled with Vista was as good as any on the market you could pay hundreds for. Are you able to comment on this claim?
  
  I'm also looking for an mp3 dictaphone possibility which would allow remote input to then be converted to text in some way. Are you aware of any such trickery that is proven to work?
  
  I have been waiting 10 years for this technology to catch up. I used to work for Ericsson, and one of the projects I was on was to develop a virtual secretary system for SoHo businesses. One of its functions was a telephone answering service that could recognise full speech, respond to callers in an unstilted manner, take the verbal message given, convert it to text and bung it to your email or wherever you wanted it. I had a lot of fun testing it (even with funny accents and slang), and it worked a treat. Frustratingly the project got canned just before launch for no reason any of us could work out, and no more was made of the speech recognition technology. The speech dedicated software that followed a few years later (IBM etc.) was utterly pants by comparison, and even now I'm not convinced it's up to the mark. I understand you still have to train it to recognise your voice and speech patterns?
  
  Anyway, I'd love to discuss any of this. Drop me a message in my inbox when you've replied here.
  
  Thanks,
  Steve.