EmailWhen it comes to voice recognition software, the choice may be less about which speech recognition product to buy and more about whether voice recognition software is more efficient for your use than typing. Most reviews agree that voice recognition software accuracy and ease of use have improved so much that it may finally be worth the investment of time and money. At 90 percent accuracy you'll have to correct one mistake in every 10-word sentence; however, test accuracy for NaturallySpeaking10 and Windows Speech Recognition now ranges between 97 percent and 99 percent. Reviewers say that correction is easy in Dragon NaturallySpeaking 10 and Windows Speech Recognition, but each reviewer expresses an individual level of patience for interrupting flow.
When considering which speech recognition package is right for you, experts say to consider the following:
- System requirements are important. Before you buy any speech recognition software, be sure your hardware and operating system exceed the minimum system requirements.
- Make sure the edition will work with the software programs you use most frequently. For instance, the Standard edition of Dragon NaturallySpeaking 10 won't work with Excel; for that you need the Preferred version. Only the Professional edition works with PowerPoint, Lotus Notes and Outlook.
- Microphones can make a difference. Good microphones take into account a normal level of background noise and deliver clean speech to the software. Some reviews say the Bluetooth headset Dragon is shipping in some bundles doesn't do a great job in that regard, but that wired headsets in NaturallySpeaking 10 (and NaturallySpeaking 9) are much better.
- The type of writing you do may not be well suited to voice recognition. If you're a software engineer writing code or a sports reporter using lots of proper nouns (especially foreign names) and statistics, voice recognition software is not likely to help you. Those who work in a quiet cubicle area (where you can disturb others by talking) or a noisy environment (where the microphone can pick up extra noise) also face challenges.
What is voice recognition software (VRS)?
Some experts say that automatic speech recognition is erroneously called voice recognition, but both terms are accepted for common usage. Voice recognition software (VRS) has two main functions. The first is spoken control of computer functions (e.g., "open Word" or "turn off microphone"). The second is to convert your spoken words to text in a computer program, especially Microsoft Word or Outlook/Outlook Express/Windows Mail. This means that you can create text without typing. When you speak through a microphone (most voice recognition software programs include one), the software translates the sounds into written words.
VRS has always been a blessing for people who are unable to type, as well as for those who have difficulty typing due to a physical condition. The killer application that would allow any of us to dictate a stream of consciousness and flawlessly translate it remains elusive. Early programs used discrete speech technology, which meant you had to pause between words in order for the computer to understand them. The latest generation uses continuous speech technology, which allows you to speak more naturally, although most reviewers find that the software must catch up with them somewhat. With caveats about accuracy, nearly all reviewers say that talking is faster than typing as a means of entering text.
All voice recognition programs but Dragon NaturallySpeaking 10 require an enrollment process, during which you read sample text out loud to train the speech recognition software to understand your voice patterns. While most voice recognition software users want a large vocabulary in their software, larger vocabularies require a more time-consuming training process. Because many words sound similar, you need to train the computer to understand how you pronounce them. The software learns from its mistakes and progressively improves.