Category: 

Does Voice Recognition Software Really Work?

Article Details
  • Written By: R. Kayne
  • Edited By: O. Wallace
  • Last Modified Date: 26 July 2014
  • Copyright Protected:
    2003-2014
    Conjecture Corporation
  • Print this Article
Free Widgets for your Site/Blog
Bill Clinton met John F. Kennedy when he was 16.  more...

September 2 ,  1666 :  The "Great Fire of London" burned down more than 13,000 buildings, including St. Paul's   more...

Voice recognition software, like any software package, varies in quality depending on the coding. But does the technology work? The short answer is yes, especially if you ask those using the near-undisputed champion, Dragon NaturallySpeaking™. Dragon is compatible with Windows™ operating systems. Various sources indicate Macintosh users tend towards iListen.

This type of software translates the spoken word into text for hands-free document creation. Kits generally come with a microphone and training materials used to "familiarize" the software with a user’s voice, speech patterns and enunciation. Initial training requires about 15 minutes of reading prepared text into the microphone, but extended training is encouraged to increase accuracy.

By many accounts, a good voice recognition software package can be as accurate as 95% after the initial training period, reaching close to 99% after extended training. It can also be used to send commands to the computer, including mouse movements and even macros. For the uninitiated, a macro is a series of movements required to carry out an operation, condensed and recorded to a shortcut key combination.

While voice recognition software can be very accurate, it does take getting used to. Sentences are dictated with the punctuation spoken, including pauses. The time it takes to invest in extra training sessions to make the software more accurate pays off in the long run, as correcting language isn’t always a matter of running a quick spellcheck. “Being” for example, might sound like “bean,” though context checkers should catch most mistakes.

Ad

Some people who benefit from voice recognition include those who suffer from disabilities or conditions like carpel tunnel syndrome. Translators and medical transcribers can also find it invaluable. Business executives who dictate notes can use the software to generate memos, compile "to do" lists and create correspondence, as some packages are compatible with portable digital voice recorders.

There are various packages of voice recognition software at differing price. In some cases, the voice engine is the same among packages and therefore the accuracy, but features differ. Advanced packages for professionals are geared towards the needs of each market with targeted functionality. Some operating systems include a free version of the software.

Low-quality voice recognition software will be slow to translate, have a lower level of accuracy, and be less compatible with third-party packages. It might be fun to try, but if you want to truly use it, a better quality package — whether free or purchased — will save you time and headaches. Just be sure the package does what you require, whether that’s sending emails, preparing spreadsheets, or generating notes with portable devices.

Ad

More from Wisegeek

You might also Like

Discuss this Article

anon335710
Post 10

Is the software compatible with travel agency booking systems?

anon325228
Post 9

@anon50308: There is a lot of software for disabilities. I think the best one is Voice Finger.

anon306877
Post 8

How does voice recognition software work with programs like quickbooks?

anon132300
Post 5

can these systems be used to voice in type in an online pokerstars chat bow?

anon50308
Post 4

My wife had a stroke and is unable to use one hand. She wants to take an online college course. Can she use voice recognition software to fill in online forms, tests, etc.?

anon24696
Post 3

These complex and sophisticated programs use built-in audio-to-text dictionaries that match the sound of a word to the written word. In other words (no pun) every dictionary word in the program has two parts: the digital acoustic fingerprint of the spoken word, and the textual counterpart. When you speak into a mic, the software matches the acoustic analog print to the digital print, identifying the word. If the word can be one of several words that all sound alike, (I, eye, aye), the software looks at the surrounding words using a grammar checker to see what the likely intended word is. When you consider some people do not enunciate, have accents and dialects, you can see how complex these programs really are. That's why a good program is necessary for good results, and it helps to speak clearer than you normally might when dictating to a program like this. It can reduce errors, as long as you are consistent. If you speak clearly one time, and more naturally the next, you won't train the software efficiently.

anon23579
Post 2

i want to know how this software actually works. how it is able to recognize our voice and is able to perform the task?

malena
Post 1

Absolutely it works! Being used to emails from my brother that were comprised of one to two very carefully worded sentences, I was amazed when I received a page long email! When I asked him what gives, it all made sense -- voice recognition software! It's a godsend to those holdouts that refused to learn how to type with more than one finger!

Post your comments

Post Anonymously

Login

username
password
forgot password?

Register

username
password
confirm
email