advice needed on speech recognition software

quietlyleaving

I'm looking into voice recognition software. Has anyone here tried any. The only one I've heard of is Dragon software. Is this the best?

HintOfLime

I tried Dragon about 7 years ago, and it was pretty good at that time for simple dictation and voice commands.

I don't know if they still offer it, but when I got mine their big selling point was that it included a good Plantronics headset with a good noise-cancelling microphone in it (getting a clear sample is the first challenge of voice rec). It was indeed a really good headset, and I still have and use it today with skype - it's quality is awesome and it's survived far longer than any other headset I've owned without parts breaking/falling off.

EDIT: It seems this may no longer be the case, according to the cnet review:

The bad: The pricey license of Dragon NaturallySpeaking 10 serves only one user; rocky installation on Windows Vista; no Mac option; poor audio quality on included headset led to poor results.

Bummer. :/

- Lime

fokyc

IF you really need to use it, disability etc. they are OK

Takes a lot of patience and a quiet setting to work reasonably well.

I have kept mine in case I get a problem and need it.

Mine was an IBM early version mid 90's

fpkyc

quietlyleaving

thanks lime and folkyc

I can understand why it would need a quiet setting but I don't have much time and patience. I wonder if the newer versions are more efficient?

teel

Have you tried the built-in voice recognition in Windows? Not saying it's the best, but it's the cheapest On my Win 7 desktop I got acceptable results with commands - not dictating though, but that's mostly because my english is very accented, not being native english speaker.

HintOfLime

I can understand why it would need a quiet setting but I don't have much time and patience. I wonder if the newer versions are more efficient?

I bet realistically it's about the same (maybe a tad better thanks to faster processor speed and some marginal improvements). A lot seems to depend on your voice. I don't have a strong accent, so it could take dication from me at a normal speaking pace (Dragon uses some grammatical analysis to figure out words you said.. so it'll get it wrong initially, but as you complete the sentence it may correct itself.) It always requires some proof-reading, which frankly I think is done better in a second pass with the keyboard, rather than using the dumb voice commands while creating the draft. Usually you want to fix grammer and stuff afterwards anyway.

When I used it.. it tended to be a bit of a novelty - it could eventually get the job done, but not without some pain. In the end, I could still type things faster and with better accuracy, so unless the technology has vastly improved, or you really need to be hands-free, I wouldn't bother. (With a cordless headset, it might be nice if it could take notes/dictation while I pace around the room brainstorming or work on the whiteboard.)

The price does seem to be an issue now.. $200? I think it was about $80 when I got it (version 5 I think).

- Lime

quietlyleaving

I don't mind parting with the cash - its just that I want it to work reasonably well as I need to be more handsfree than I am now.

quietlyleaving

btt

where is gubbering when you need him

fokyc

Have you looked at e-speaking?

http://www.e-speaking.com/index.htm

it looks reasonably priced; pay $14 if you like it after 30 days free.

fokyc

dssynergy

I work with disability accommodations, and I can say that most people use Dragon for the speech-to-text software. If you have a Mac, it has a good built in system.

But no matter what software a person chooses, they can expect to spend a few hours training the software to recognize their speech patterns. And, it only will recognize one voice at a time. So it is not something that can used for meetings, classes, or other times when different people are taking turns speaking, or when many people are speaking at once. Don't believe the accuracy rating that they give themselves. They achieve those results in a controlled environment. Also, as far as I know, you will have to speak like this:

Hi[exclamation point] [new paragraph]

Thanks so much [comma] for stopping by the other day [period] I really enjoyed our visit[period] I can't believe how big John has gotten[exclamation point][new paragraph] .....

Speech-to-text software really is amazing, and I think people for the most part are happy with the results, IF they have realistc expectations of what is involved in using it and what results to expect.

Just my 2 cents. Feel free to PM for a more in depth discussion.

DS

advice needed on speech recognition software

Share this