[MV] Voice recognition for transcription: Opinions sought

[Chuck Rogers]

Rena (and everyone else):

Simply put, no speech recognition can do what you ask, and it will  
probably be many years before computers have the processing power  
necessary to do so. The problem will be that the speaker is not  
speaking his or her punctuation, and not speaking in an environment  
with a consistent noise level, nor will they be using a noise- 
canceling microphone that is in a consistent position in relation to  
their mouth. All of these factors will introduce enough inaccuracy in  
the transcribed text to make it not worth the effort.

We have many people using our transcription solution and what they do  
is re-speak the audio in their own voice, inserting punctuation as  
they go. This produces much more reliable transcription and still  
saves about 30% of what it would take to type in the text manually.



Best Regards,

Chuck Rogers, Chief Evangelist
MacSpeech, Inc.




On Feb 21, 2007, at 10:31 PM, Rena Bussinger wrote:

> I am a Mac user and completely new to voice recognition software on  
> any platform. I do a lot of transcription of voice recordings, and  
> am interested in experimenting with using software to help minimize  
> my workload. I understand that the software must be trained to  
> recognize a speaker's unique voice and that ongoing training of the  
> software is necessary to improve the accuracy of the translation  
> over time.
>
> Here's what I'm wondering: How involved is the initial training  
> process? I am producing my voice recordings from personal history  
> interviews with subjects who are active and willing participants.  
> Would it be feasible to create a speaker profile for each of my  
> interview subjects, have them go through the initial voice training  
> process (on my laptop) and then playback their recorded audio with  
> iListen or DNS/Parallels active? If anyone has any experience using  
> either program for a similar purpose or to transcribe your own  
> voice memos/speeches/performances/etc. I'd be very interested in  
> the details of your process and how successful it is. My  
> transcriptions are primarily for reference/editing purposes so  
> complete accuracy is not as important as just having a rough text  
> version of the audio that I then edit extensively anyway.
>
> Any feedback is appreciated.
>
> Thanks,
> Rena Bussinger
>
> The Story Foundry
> We Make History. (Yours.)
> http://www.thestoryfoundry.com
> 206.349.8391
>
> _______________________________________________
> MacVoice mailing list
> MacVoice at listserver.themacintoshguy.com
> http://listserver.themacintoshguy.com/mailman/listinfo/macvoice
>
> Listmom is trying to clean out his closets! Vintage Mac and random  
> stuff:
>          http://search.ebay.com/_W0QQsassZmacguy1984