Wednesday, 1 May 2013

I can...use my voice to write!


Using your voice to write is becoming increasingly popular.  There are a variety of voice recognition options that are built into your computer, tablets and phones.  For kids with disabilities, this can mean greater independence for writing anything from essays and emails, to putting a reminder in a calendar.  In this issue, we will explore various types of voice recognition solutions and the skills required to make them valuable tools. 

What is voice recognition?
Voice recognition (or speech to text) involves a person speaking aloud into a microphone attached to a computer that translates what's been said into text on the screen.

Readiness for Voice Recognition
Voice recognition can be a fantastic tool for some kids. However, it may not be the best fit for everyone.  There are some skills that help make voice recognition work more effectively.  Take a look through the chart below.  The more often you answer "yes", the more success your child might have using this technology.

YesNoVoice recognition software works best when your child can...


Speak smoothly and clearly
  • If not, is the articulation consistent?  


Speak without being out of breath
  • Breath support issues may result in noisy breath or difficulty finishing sentences


Control his/her rate of speech
  • Able to say words all together for commands such as “Go to sleep”, “End of line” 
  • Able to articulate whole words together “Computers” vs. “com…puters”



Read at a Grade 4 level or above.  If not, at what grade level?
  • Read text predicted by the software
  • Recognize the correctly spelled word from a list of similar words



Generate their own writing or have supports/strategies in place to help them
  • Able to compose text aloud (aka dictate)
  • Able to edit written work - recognizing errors and able to correct them 


Visually scan through text on a screen
  • Horizontally to read text on the screen
  • Vertically to read a list of words in a correction box

Voice Recognition Software:
Here is a summary of some commonly used speech recognition options and some comments from our ACS clinicians who have experience using them.

Dragon Naturally Speaking is a windows based software program made by Nuance: http://www.nuance.com/for-individuals/by-product/dragon-for-pc/index.htm
  • This is a commonly used software that allows someone to use only their voice to control and dictate to the computer.  For kids with physical disabilities, this software allows them to control the computer and write without using their hands.
  • This software works best when your child has trained his/her voice, which involves reading a passage to the computer so it can learn how your child speaks.
  • It's important to maintain the child's voice file by having the child correct errors that are made during dictation.  This allows the software to learn more about how your child speaks and this improves the software's overall recognition.  It also means that it's important to speak in sentences and avoid "testing" the recognition with random words.
  • There are 3 different versions of this software.  The majority of our clients use the home version, however, if your child requires customized commands or other features, it may be useful to explore the other versions. 
  • Nuance training videos and resources: http://www.nuance.com/for-individuals/by-product/dragon-for-pc/existing-customers/dragon-training/index.htm 
Dragon Dictate is the mac based software program by Nuance: http://www.nuance.com/for-individuals/by-product/dragon-for-mac/index.htm
WordQ/SpeakQ is a windows based software program by GoQSoftware : http://www.goqsoftware.com/product-details/wordq-and-speakq/

  • You can use this software in two modes: "speak and select" and "continuous"
  • When using speak and select, your child speaks a sentence and the program uses word prediction to give your child options from which to choose.  The choices can be read aloud to your child and then he/she can select one to put into their document.
  • When using continuous, your child simply speaks and the program inputs it's best guess into the document.  Your child then corrects his/her work.  In this mode, the recognition is not improving as the the software is not getting feedback.  
  • Often kids use speak and select to help the program learn how they speak and then switch over to continuous depending on the writing task and accuracy of prediction.
  • This software requires the user to use a mouse or mouse alternative to make a "selection" and does not have voice commands to control the computer. 
Other Voice Recognition Options:

Microsoft Speech Recognition is voice recognition software that is built into Windows operating systems that are Vista or higher.  You can use this voice recognition software to control your computer as well as for dictation.  For more information:  http://windows.microsoft.com/en-ca/windows7/set-up-speech-recognition
  • It works best with Windows based products such as MS Word etc.
Siri is a voice recognition platform included on:  iPhone 5, iPhone 4S, iPad (3rd generation) and iPod touch (5th generation). You can ask Siri to schedule appointments, check information or simply for dictation.  It uses natural speech so you don't have to train it by reading a passage or using commands.  For more information about Siri in Canada see: http://www.apple.com/ca/ios/siri/siri-faq/

Apps:  There are a variety of apps that can be used for speech to text purposes.  When looking through all the possibilities, keep in mind that not all programs will recognize a person's voice in the same way.  
Microphones - There are many different options when it comes to microphones.  When choosing a microphone for use with voice recognition, it's important to be aware of a few different features:

USB Microphones:   One of the advantages of the USB microphone is that it is directly connected to the computer.  This direct connect may allow for improved transmission of information impacting the recognition.

Noise Cancelling Microphone - this feature helps to reduce the background noise picked up by the microphone. 

Wireless/Bluetooth - A wireless headset uses bluetooth technology to send information from the microphone  to the computer. If recognition of speech is challenging, keep in mind that bluetooth technology compresses the size of the information (speech file) when sending to the computer and may slightly impact the recognition.

Using a idevice as a microphone - if you have an idevice, you can download an app and use it as a microphone.   http://www.nuance.com/dragon/remote-microphone/index.htm

Manufacturer Recommendations: The company that produces your software may have some recommendations on the types of microphones recommended.
  • For Nuance, check out: http://support.nuance.com/compatibility/ and choose Dragon Naturally Speaking or Dragon Dictate
  • For GoQ, check out: http://www.goqsoftware.com/wp-content/uploads/2010/11/SpeakQ-Microphone-Transcript-for-goQ.pdf 

Vocal Hygiene: using your voice to write can be taxing on your vocal cords.  Here are some suggestions to help keep vocal cords healthy:
  • Speak in your natural voice (avoid speaking in a monotone voice)
  • Breathe in through your nose to keep air moist
  • Humidify room if possible
  • Keep dust to a minimum
  • Drink water frequently - not only when taking a break
  • Limit use when not feeling well
  • Take a break every 30 to 60 minutes
  • Avoid smoking, caffeine and antihistamines if possible as they dry out the throat
  • Avoid excessive throat clearing/coughing  - try a gentle swallow instead
How do I try out Voice Recognition?
  • Talk with your child's teacher or school personnel to discuss the possibility of using voice recognition.  
  • If your child has an Occupational Therapist, speak with him or her about your interest in trying voice recognition.
  • You can also learn more about these programs online.  There are some free demos and webinars that may be helpful.  
    • Free Trials: http://www.goqsoftware.com/trial/
    • Video Demos from Nuance:although there are no free trials, you can look at online videos to get an idea of how their software works.  http://www.nuance.com/for-individuals/by-product/dragon-for-pc/home-version/index.htm 
  • ATN Access for Persons with Disabilities Inc, a company based in London, ON, offers fee-for-service software or assistive technology training including voice recognition. 



 

1 comment:

  1. Wow, this 'I Can' is awesome! The voice recognition readiness article answered a lot of our questions (and points we hadn't thought of). It gave us a better understanding of potential strengths and challenges of this, for our daughter. THANK YOU!
    Matt and Gill

    ReplyDelete