Voice Recognition

Speech recognition in ITAS

In summary

When using both Windows Speech Recognition and Dragon Naturally Speaking 9 it was possible to carry out all actions that could be executed using a mouse or keyboard. Some actions took longer to perform when compared to using a mouse but all programs do have full functionality.

Windows Speech Recognition is included in all versions of Windows Windows. Dragon Naturally Speaking is the only real alternative.  Naturally Speaking 9 is regarded in reviews as the best voice recognition software on the market however Windows Speech Recognition is a very close match and it is initially a more user friendly and easier method of interacting with the computer than Dragon. Naturally Speaking does have many more options for training the computer to adapt to your voice and how you say certain words. With extended use and the willingness to learn specific commands the software could prove itself a more accurate solution than the Windows software.

Full review

Windows Speech Recognition is included in all recent versions of Windows. (2009).
In Windows, speech recognition can be started by clicking Start – Settings – Control Panel and opening Speech Recognition.
Click ‘Set up microphone’ and follow the instructional wizard. Clicking ‘Take Speech Tutorial’ the tutorial takes around 30 minutes and guides you through the common speech recognition commands.
Speech recognition can be configured to run at start-up to prevent having to use the mouse to turn it on each time you start your session.
To tell the computer to start listening the “Start Listening” command is used. The microphone icon in the lower right hand corner of the screen will turn from grey to blue.
When you have finished with speech recognition use the “Stop Listening” command to stop the computer listening to what you say.

When speech recognition is running, the command “What can I say?” brings up the Windows Speech Recognition Reference Card, which is a list of common commands so you know what commands you can use if you do get stuck or forget what commands you can say.
As well as dictation Windows Speech Recognition allows you to control windows, start programs switch between windows, work with menus and click buttons using only your voice.
To close, maximize, restore or minimize the active window the following commands are used: “Close that” “Maximize that” “Restore that” “Minimize that”.
To switch between open applications the “Switch” command is used followed by the name of the application you want to switch to. For example “Switch to Internet Explorer”.

Windows Speech Recognition also gives you the ability via speech to press any key or combination of keys as though you were using a keyboard. This is done by saying the “Press” command followed by the keyboard key or keys you wish to press. For example you could say “Press a”, “Press Shift a”, “Press Alt Tab” etc.
Certain keys can be pressed without using the “Press” command. For example, Delete, Backspace, Enter, Page Up, Page Down, Home, End, Tab.
When navigation through forms and application such as ITAS I have found the quickest methods are buy using voice commands to press certain keys on the keyboard.
Navigating up and down lists can be done using “Press Down Arrow”, “Press Up Arrow”, “Page Down” and “Page Up” commands.
Navigating up and down forms can be done by using “Tab” to go down a form and “Press Shift Tab” to go back up a form.
Expanding combo boxes/drop down lists can be done with the command “Press F4”.
Check boxes can be ticked by the command “Press spacebar”.
Items can be opened using the “Enter” command.
Windows Speech Recognition includes a feature it calls ‘Say what you see’. This work well and allow you to open the File, Edit, View, Insert etc menus buy simply saying their name.
Saying the command “Show Desktop” minimises all open application and allows you to then click icons on the desktop with commands such as “Double Click” followed by the icon you want to open/run. For example “Double click My Computer”.
One of the most user friendly features of Windows Speech recognition is the command “Show Numbers”. When said, numbers will appear over everything on the screen and you can say the number to click the corresponding button, link or textbox. This feature makes navigation very quick and effective and allows a beginner to speech recognition to be able to navigate without knowing any complicated commands.
Windows has a number of commands to cleverly navigate around applications however one of the simplest is with the use of the “Mousegrid” command. This feature allows you to move the cursor anywhere on the screen as if you had control over the mouse.
Saying “Mousegrid” splits your screen into a grid of 9 numbered sections. Saying a number then splits that section of the grid into a further smaller grid of 9 numbered sections. This continues until you have navigated the cursor sufficiently accurately to the point of the screen you wish to click on. Using the commands “Click” “Double click” or “Right click” control the mouse buttons and performs the action as if you were using the mouse.
Using Windows Speech Recognition, I found that all tasks that could be performed by a mouse or keyboard could be achieved by using only voice commands so all programs did have full functionality. Some actions as expected took longer to perform compared to having the use of a mouse, however some actions like opening programs were quicker by voice than having to look for icons on the desktop or having to navigate through the start menu and clicking with a mouse.
Sometimes commands do have to be repeated and corrections made to text but with continued use and practice I’m sure this could be reduced to a minimum. It does now appear that speech is a real alternative to using the mouse and keyboard.
The only other real alternative to Windows Speech Recognition is Dragon Naturally Speaking 9 which is regarded in reviews as the best voice recognition software, with potential for 99.8 percent accuracy.
Naturally speaking can be purchased around £50 and £100 (microphone included) depending on which version is desired, standard or preferred. The standard version does not work in Microsoft Excel and does not have a feature that allows you transcribe from portable voice recorders but both have the same speech recognition engine.
After initially using Naturally Speaking software it didn’t appear as user friendly as Windows Speech Recognition. Many of the commands used are the same, both have the “MouseGrid” command (naturally speaking uses commands “Mouse Click” “Mouse Double Click” and “Mouse Right Click” commands to control mouse buttons) so any function that can be performed via the mouse can still be controlled by voice.
It felt as though the Naturally Speaking software was designed to use a combination of keyboard, mouse and voice to interact with the computer, as a way of getting text on the screen faster. Whereas Windows Speech Recognition attempt to do away with the mouse and keyboard altogether which it does in a simpler and more obvious way than Naturally Speaking, making navigating around the whole windows environment more straight forward.
Naturally Speaking does not have the “Show numbers” command that Windows has which is one of the features that makes Windows speech recognition so easy to new users and those who do not want to remember lots of specific commands.
However Naturally Speaking did have many more options for training the computer to adapt to your voice and how you say certain words. So with extended use and the willingness to learn specific commands the software could prove itself a more accurate solution than the Windows software.
Naturally speaking does appear to be a very sophisticated product but speech recognition seems to be an area that Microsoft is investing heavily in and is already a very good match for Naturally Speaking. If you have windows Windows installed or are planning to buy a new computer it is well worth trying Windows Speech Recognition before spending the extra on Naturally Speaking.

Was this helpful?
Thanks for your feedback

Recently viewed