Hello everyone and welcome back!
Today's topic of interest is Dragon NaturallySpeaking software.
For those of you who are unfamiliar, voice recognition software allows you to put words to paper simply by using your voice. It sounds simple enough, right? Not necessarily. It requires a lot of training and persistence on the part of the user. Training allows the computer to adapt to a person's voice and particular speech patterns. This process is essential to the user's success with dictation as well as computer accuracy. There are two primary ways to proceed while training. The first is through introductory training directly after installation. In this process, a person is asked to create a name for the voice files to be saved under. (Note: there can be more than one user.) The program then checks whether your computer has adequate sound quality in memory to support it; it does this by asking users to read aloud what is provided on the computer screen. If input is normal or above average, training can continue. If not, the computer may ask you to backtrack, providing you a few options. For example, checking your source of audio input (i.e. the headset microphone provided with the software or a regular tabletop microphone)
Users are then asked to continue the training process by reading aloud from a variety of book excerpts. They cover a variety of topics and cater to specific ages and skill levels of the user and what the product is used for, some excerpts may be more appropriate than others. That way, the software gets to know the type of vocabulary that is unique to you and will be more accurate at guessing what you were saying when you dictate the future.
The second of training is called on the spot training. It's not really called that or anything, but that's the best way I can to explain that. This type of training is ongoing, where the computer tries to guess what the user is saying. If correct, the user may continue dictation without interruption. If incorrect, they must go back and correct the error. This is a necessity for two reasons. First and probably most obvious, people want to produce documents without errors. But actually there is a reason more important than that. Correcting an error improves the accuracy of recognition and adapts the vocabulary to your way of speaking. If mistakes are made without correction, the computer assumes it's accurate and adapts itself accordingly, influencing other words in your vocabulary .
Correction can occur in several ways, but I'm going to highlight only two of them. First,if the computer dictates incorrectly, the user can simply say "Correct" and then the word it spelled wrong. Then, a list of possible choices of sound alike words will appear. From there, the user can simply say just "Choose" and the number corresponding to the correct word. If it does not show up, they use will probably have to spell it manually by saying "Spell That "by using the handy command and alphabet list as users begin to navigate and understand the software.
As you can see, there is too much to discuss in one post and quite honestly, I am getting tired of talking. Yes, I am using the voice recognition software as we speak to write this. So stay tuned for part two of what I decided to make a series.
Part two will be called Voice Recognition: Commanding Your Computer.