Tuesday, April 14, 2015

School Tools: Your wish is my command!

By Kent Remund

If you grew up in the 70’s and 80’s, you dreamed of the possibilities of controlling things through voice.  

From the Jetsons, to Star Trek and Knight Rider, I imagined how cool it would be to be able to speak some commands and have powerful actions taken place. For the most part, those days are happening now.   

Speech recognition for computers has been around for the past 10 -15 years, but it has made significant progress in the past couple of years through smartphones and even the ability to talk to your car.

Speech recognition enables an individual to turn their speech into text to operate their phone, tablet and computer using voice commands, e.g., create documents, send email, text messages and control applications. 

For students who struggle with the writing process because of physical limitations, vision impairments or specific learning difficulties, this software has proven to be a successful assistive technology solution. 

There are typically 4 different methods used to evaluate and assist individuals with voice to text. These include Windows Speech Recognition, Dictation for Mac, and Dragon Naturally Speaking.  

These three programs work best using a USB microphone which allows for a “cleaner” audio input for the computer to process. The fourth program uses tablets or smart phones.   

Let’s look at each in more detail:

Windows Speech Recognition: Microsoft began including speech recognition in their operating system (OS) beginning with Vista and has continued to do so through their OS. This recognition works quite well for speech to text but really excels in providing variety operations using a windows computer through voice for those with limited mobility and limited mouse/keyboard use. 

Windows speech recognition can be launched through the accessibility/ease of access features. Once a short microphone compatibility test is done, Windows Speech Recognition works well. Typically, there's upwards of 80 – 85% accuracy for the average individual with a strong, consistent voice. Microsoft provides good usability and best of all, FREE with all windows computers.
Screen shot of Windows Speech Recognition.
Screen shot of Windows Speech Recognition.

Dictation for Mac: Speech recognition has come a long way over the past couple years from Apple. Yosemite operating system made a huge jump with dictation that comes with every Cac computer. There are many computer commands built in, and it has the ability for the individual to create new dictation commands. 

Accuracy levels are typically above 90 percent with most individuals and are easy to launch by pressing the left function (fn) key twice. Apple provides great usability and best of all it is FREE with every new Apple computer.  
Mac Dictation screen shot.
Mac Dictation screen shot.

Dragon Naturally Speaking: Dragon was one of the first voice recognition programs. It was created in 1997, and has been the leader in voice recognition through the years. Dragon comes as a full software install on a computer. Once an individual trains their voice with dragon, it is extremely accurate at 99 percent. Dragon will continue to learn and grow with the user and their voice commands. Dragon for PC and Mac starts as low as $99.00 and is an excellent choice for someone that uses voice commands on a daily basis. 
Dragon Naturally Speaking screen shot.
Dragon Naturally Speaking screen shot.

Tablets and Smart Phones: Many individuals are simply using their tablets and smart phones for speech to text and other voice activation commands.  Android, Apple and Microsoft added a microphone button to the keyboard when it is present on the screen.  

Simply press this microphone and dictate into an email, text, notes or any area that you would normally enter text. Many other commands are available such as asking the weather conditions, checking stocks and hundreds of other commands through Siri, Ok Google or Cortana. 

The drawback to using this technology is that the user must be connected to the internet or a data plan for their speech to be converted to text. There is also a limit in how much can be spoken into the devices at one time. Typically this is a short paragraph before needing to press the microphone button once again.  
Smart phone speech to text microphone key.
Smart phone speech to text microphone key.

Overall, there are many variables in how accurate each of these systems converts speech to text for each individual. These are a few of the alternate solutions for students to complete coursework that seems to work well for many with learning disabilities and other writing challenges. 

Kent Remund is part of the Utah Center for Assistive Technology serving Utahns with disabilities, and a member of the Utah Assistive Technology Teams serving Utah school districts.  

1 comment: