Activity 3: Alternative Input & Output

Designing interfaces is getting more complex as new technologies evolve.

Speech Recognition (Input)

Speech recognition is getting better all of the time due to better software and more powerful computers. One of the leading pieces of software in this field is Dragon NaturallySpeaking.

The advantages and disadvantages of using Dragon NaturallySpeaking are -

Advantages

•    Type as fast as you talk
•    It doesn’t take long to learn
•    It can probably handle your accent
•    It improves with time
•    You can stroll around the room
•    Some very cool web features

Disadvantages

•    It’s not perfect
•    It refuses to learn some words
•    Performance varies in different applications
•    Surfing the web can be a pain
•    Dragon is Windows only


Dragon NaturallySpeaking is a Windows based solution. Macintosh computers can use MacSpeech Dictate. There is also free speech recognition available in Windows Vista. The following is a comparison between these three choices -

As a longtime Macintosh user and some may say “Fan Boy”, I tend to promote the advantages of Macintosh computers and Apple’s OS10.5 “Leopard” but when it comes to speech recognition I am afraid I must concede, “Windows does it better”.

MacSpeech Dictate is just another in a long line of sub par speech recognition programs on the Mac platform.  The one thing going for it is that the new version is using Dragon’s speech recognition engine. Unfortunately this improvement is restricted to Intel based Macs as PPC based machines do not get any benefit.

It is still, even on an Intel Mac not as accurate as Dragon’s or even Vista’s inbuilt speech recognition software.  What it does have going for it is an increased level of customization when setting it up with preferred workflows and better integration with other Mac applications. But this comes at a price.

Vista’s inbuilt speech recognition software is included with the OS for free, works better than the Mac version but is incompatible for use with your Mac applications.

Dragon’s software may be Windows only but if you like me run VMWare or some similar virtualization software on your Mac to enable you to run Windows, Linux or whatever other OS you want may actually end up being your best choice. Sure it is expensive but if you want the job done well then it is really the only contender.

With the ability to copy and paste both ways between the Host and Guest OS’s it is easy to run Dragon on your Windows side and paste the results to your Mac applications in much less time than doing it solely on a Macintosh, taking into account the need to correct the many more errors you get when using a Mac.

Text to Speech (Output)

As well as recognizing speech there is the option to convert text into speech. This has also improved greatly from the previously available robotic sounding voices.

There is a text to speech demonstration at http://enjoylearningonline.com/HCI/TT

I found this demonstration to be quite impressive. I was especially impressed by how realistic the voice was. Unfortunately you cannot really interact on this site.

An interactive demonstrations can be found at  http://tts.loquendo.com/ttsdemo/default.asp?page=id&language=en which I found


the advantages and disadvantages of text to speech are -


Advantages

This feature is built into multiple Operating Systems for free
Great for people with disabilities
Great for young children unable to read yet

Disadvantages

Text read out aloud may be embarrassing
No real Australian dialects
Voices can tend to sound artificial
Poor logic handling when translating non-standard formatted text e.g. tables


Three situations where text to speech might be useful are -

     * Helping people with reading difficulties or those unable to read
     * Useful for converting transcripts that should accompany videos - 508 Compliant
     * Useful for new immigrants to learn how to pronounce words properly



Marketing | Multiple Formats| Alternative Input & Output