Online transcription service using SR?
Submitted by davidorban on Mon, 12/03/2007 - 15:58.
A friend of mine periodically has hour long interviews to transcribe, and I told her that using speech recognition she'd be quicker. She would like to use a service that could do the job, and I want to point her to one or more, that would be more convenient, because of its use of SR.
Can you recommend some?
David
David Orban
www.davidorban.com
PS: sorry if this topic has been already covered previously. I tried to use the search function on the forum, withour success.



davidorban wrote: A friend
A friend of mine periodically has hour long interviews to transcribe, and I told her that using speech recognition she'd be quicker. She would like to use a service that could do the job, and I want to point her to one or more, that would be more convenient, because of its use of SR.
Is there any chance your friend would be interested in saving money and doing the transcription herself? There is a method called Echo Dictation when a person listens to the interview in earphones, stops the recording, and then dictates into Dragon. If your friend needs to reread and edit the transcription, it may be a watch in the amount of time they spend and eventually save some money even with the cost of Dragon and some of the hardware/software.
What we recommend for recording is a software called SoniClear recording software. This software allows a person to use their notebook computer to record an interview. They can even use bookmarks or an agenda to make separate tracks of different parts of the interview. The software also has a very easy to use transcription software applet to stop, start and rewind the recording during echo dictation. A Footpedal makes the transcription even easier.
Here are some tips you can pass on to your friend for making the transition to Dragon:
1. Read "Key Steps to High Speech Recognition Accuracy"
2. Go to the Links/Articles page of the eMicrophones web site. Listen to the three sound files on How to Sound and How Not to Sound. Essentially, it may be continuous speech, but it is not conversational speech you need to master. Each word must be enunciated clearly. Speaking in phrases, helps the system because the system is listening not only for the sounds of words but comparing each word to words before and after for context clues.
3. CLICK HERE to go to a link in our Dragon NaturallySpeaking FAQ or frequently asked questions section. It tells you the proper way to create a new user profile including when to add your typical documents (not during the training phase).
4. When you begin to make corrections, make corrections within the context of at least one other word. This will improve the speech profile as relates to your particular style of speaking and will increase the probability that the word will be recognized correctly the next time you speak it.
--
Martin Markoe, eMicrophones, Inc.
The best microphones for Speech Recognition
Thanks for the resource
Thanks for the resource links, and the advice. I don't think that my friend would want and pick the technology up at this point. She wants it done for her!
We also sell
We also sell digital
recorders and NaturallySpeaking
but it probably doesn't matter whether or not the transcription service
utilizes speech recognition because they're more than likely going to charge
the same price. We supply NaturallySpeaking
and digital recorders to a number of transcription companies. They're still going to have to input the
interview in a word processor; by Echo Dictating or typing. Since this question isn't really relevant to
speech recognition per se, you might want to google for the best
transcription pricing. Pakistan, India
and even China typically charge less than US counterparts but usually have
more errors. You may also find a
bargain on the transcription end-user forums from someone looking to pick up
a little money on the side. See www.transcriptionessentials.com/forums
Lunis
Orcutt - Developer of KnowBrainer &
Host
of the KnowBrainer Speech Recognition Forum
A Nuance Gold Certified Endorsed Vendor
ALWAYS Ask If Your Speech Recognition Vendor Is
Nuance Certified
Hi Lunis, I actually start
Hi Lunis, I actually start from the preconception that it should matter if the service uses DNS! My assumption is that technology should enable a better performing setup to come out with competitive pricing, whether US based, or not.
KnowBrainer wrote: We
You may also find a bargain on the transcription end-user forums from someone looking to pick up little money on the side. See www.transcriptionessentials.com/forums
Slightly off topic but I think you might find the people on
trancriptionessentials are professional transcribers, not looking for 'a little money on the side'.
{Admin fixed quote formatting}
At one time...
At one time there was such a service being advertised. It's been a while since I've seen the ad and I cannot remember their name or URL. The transcription was from a telephone dictation session if I recall correctly. This may have been on the VoiceGroup list. The archives are still searchable there.
With the newer programs (DNS 9.5) I would think that if this person can dictate into a recorder they should be able to dictate to a computer. It could then be formatted by someone else if needed.
Yes, well, it is an
Yes, well, it is an interview, so there are multiple voices, and a direct transcription from the recording, on a recorder or the computer, won't be possible...
davidorban wrote: Yes,
Yes, well, it is an interview, so there are multiple voices, and a direct transcription from the recording, on a recorder or the computer, won't be possible...
just for the heck of it, take a look at the following:
http://www.customspeechusa.com/speechmax.html
This is only one source, but you may find what they have to offer interesting, if not inexpensive.
Chuck Runquist
Former Dragon NaturallySpeaking SDK & Senior Technical Solutions PM for DNS
"Everything takes longer and costs more." -- Jerry Pournelle