Nuance Dragon NaturallySpeaking improves accessibility for all

Written By: Peter Abrahams
Published:
Content Copyright © 2008 Bloor. All Rights Reserved.

The major users of early versions of speech recognition technology where people who could not use a keyboard. Even though it was slow and made mistakes it was a boon to the user as they could, with a little patience, produce high quality written communications.

Now, the technology can reliably transcribe at 120 words per minute. This compares very well with the average user who struggles to type 40 words per minute accurately. So it is no surprise that many users including lawyers, doctors and court reporters are using the technology. The widening of the potential user base has encouraged Nuance to invest in continuous improvements in the basic voice recognition engine as well as the functionality of the user interface. This investment has brought benefits to the original group of users who have no other viable input method.

Here are some of the new features of Dragon NaturallySpeaking (DNS) that will help users who have limited or no use of the keyboard or mouse and are voice-only users.

Shortcuts

“Send email to Peter Abrahams and Steve Chambers,”: Dragon opens the user’s email program, creates a new email and puts the appropriate contact names into the “to:” box.

“Search maps for Italian restaurants in Boston, Massachusetts,” or “Search the Web for Bloor article on DNS”: Dragon opens the default search engine and Web browser to enter the spoken words in the search field and return the results. Users can also search specific sites by voice including Google, Yahoo!, eBay, Wikipedia and YouTube.

In previous versions, navigating to the correct window and input field would have required several spoken commands; users who could use a mouse tended to use it instead of voice commands. The shortcuts will greatly enhance the experience of voice-only users.

Quick Voice Formatting

Quick Voice Formatting commands allow users to issue a single voice command, such as “Bold Quick Voice Formatting,” to make formatting text faster than ever. Or, to use Word styles say “set style header 2”. In addition, an extensive set of natural language commands means users don’t have to memorize complex instructions. In Microsoft Word, for example, a user can say ‘make that bigger,’ and the font size increases instantly.

Again this is faster than previous version when at least two voice commands would be necessary.

Bluetooth headset

A special DNS 10 edition is bundled with a Bluetooth headset allowing users to move further from the workstation whilst continuing to dictate. This will be a boon to users who are unable to put on a headset unaided as they will be able to leave their workstation and return without needing to ask for help.

Digital voice recorders

DNS 10 can transcribe dictation from voice recorders and PDAs; this enables a user to record notes anywhere and then upload them to their workstation.

More Flexible Enrolment for younger speakers and users with certain speech challenges

Self-paced reading is available for speakers who have difficulty with the existing interactive text passages. This makes it easier for younger users and users with certain speech challenges to train the product to understand their voice.

Some medical conditions that make it difficult to use a keyboard also affect the user’s speech. This new enrolment process should ensure that more users who need voice recognition software can train it efficiently.

Better control of commands versus dictation on the web

In previous version it was sometimes difficult to ensure that DNS could distinguish between a word such as ‘click’ being used in dictation or as a command. DNS 10 provides various options to reduce this confusion. This should remove some frustration felt by voice-only users.

Section 508

DNS 10 is fully certified for Section 508 ensuring that all functions can be used without the need to use a keyboard or mouse. This is essential for voice-only users.

Data collection

Nuance is keen to continue to improve its product and to do this it needs to understand how it is used in real life environments. The user is asked if they would be willing to participate in a program to collect this data.

Conclusion

There are further new features in DNS 10 but I have picked out those that have a direct benefit to users who cannot use the keyboard. These features will make it easier and faster for them to work and play on their workstation.