Before I go ahead and buy a Vista box to try out Vista speech recognition and the WSR toolkit, I have a couple questions.  First, my needs are:

1) to use speech recognition to parse .wav files into text.  

 

We currently have a few thousand transcribed urology reports, and their associated .wav files.  These .wav files and 16 bit, 11 Khz mono (future .wav files could be captured with a higher sampling rate if needed).  They're recorded on a philips speech mic. 

 

2) We would like to be able to train / teach a speech recognition profile based on the .wav files in conjunction with the word documents. 

 

3) ideally we'd like to be able to setup multiple speech recognition profiles (one for each physician).

 

I know Dragon Medical 10 in its various forms will do this, but we'd rather not spend the $10,000 or so to get started.  Can the WSR toolkit help us accomplish at least some of the above?

 

Thanks!

The Transcribe From Wave File function of the WSRToolkit says:

"Using this feature will allow you to dictate to a portable voice recorder, then after putting the wave file on your computer, you can transcribe the file into text which can be copied and pasted into any application. Note: You'll get better transcription accuracy if you use "Train From Text" and "Add From File" before transcribing your file. Also: The dictated phrases 'new line' and 'new paragraph' will not appear in the text box below but will be displayed perfectly when the text is pasted into your favorite word processor.

Choose 'Browse' to locate the wave (.WAV) file, then choose 'Begin Transcription' to begin."

 

If you follow the above suggestions for Train From Text and Add From File,  and your .wav files were created as a physician would using Dragon (enunciating each word clearly, speaking in phrases and dictating punctuation) you should achieve high accuracy.

 

Marty Markoe, eMicrophones, Inc.