Darran's Voice Recognition Pages

More up-to-date information on voice stuff can now be found here.

These web pages are the incidental result of spending the last four years of my life talking to computer hardware instead of pressing keys on the keyboard.

Voice recognition systems have only recently come of age. This has not only happened due to technical advances in the speech arena but more so in the ever-increasing processor power available at PC prices. Today’s voice recognition systems are extremely usable. All that is needed is a tiny bit of user effort, patience and some development work to achieve a solid speech interface to your user environment. There is also the added bonus in knowing that swearing or shouting at your machine is not a complete waste of time, as long as you have set up the appropriate macros. The idea behind these pages is to provide a resource for information regarding speech input systems. I hope it helps!

Speech Recognition Software:

Here are the main players in the speech arena, each company has a product that allows you to talk to your computer in a natural manner. Each product provides a different degree of command and control, which basically means that you can control the Windows environment and Windows applications by voice. For example you can lunch Notepad by voice and then completely control the Notepad application by voice as well.

• Naturally Speaking from Dragon Systems
• Via Voice from IBM
• Voice Xpress from Lernout & Hauspie
• FreeSpeech from Philips

Susan Fulton has a detailed comparison chart on the features of each of the continuous speech programs at her web site.

Here are some hardware recommendations that you should be aware of before you start to use NaturallySpeaking.

If you wish to browse the web by voice then there are a number of products available:

• Converse Web
• FreeSpeech Browser
• With some additional macros you can control both Netscape and IE by using Web Keyboard.
• Both NaturallySpeaking and ViaVoice have browsing facilities built into the software.

Jonathan Epstein has written set of macros for controlling WebKeyboard with Dragon NaturallySpeaking.

Here is my view on the browsing product available.

My Macros:

• Macros for using NaturallySpeaking handsfree
• NatSpeak macros for using pine to send mail.
• NatSpeak macros for writing HTML code.
• My DragonDictate Macros

NatMerge is a utility from Joel Gould that will allow you to merge NaturallySpeaking macros into your existing macro set. You can also download the Natural Command utility from Dragon Systems that does the same job as NatMerge.

Unix Links:

I'm a Unix System Admistrator; this means I need to speak Unix to various Unix machines. Here are a few documents that may help you if you need to do the same:

How I talk to Unix boxes

  [The following links will be available shortly]

• An additional vocabulary for Unix words for NaturallySpeaking(useful words like Solaris and Red Hat etc.)
• General Unix macros
• Solaris specific macros
• Redhat specific macros
• A guide to installing Solaris handsfree
• Solaris System Administration By Voice

Programming by voice:

• Writing and debugging Java code by voice by Johnathan Espstein.

You don't want to run a Windows-like operating system, I understand:

• ViaVoice is now available for Linux.
• ViaVoice is also available for the MAC

Links:

• Joel's Unofficial Natspeak Pages
• Susan Fulton's Computing Out Loud pages.
• Loads and loads of NatSpeak macros from Chad.
• Excellent NatSpeak macros and documentation from Scriven.
• Speech Technology Magazine

Darran Bryant dnb1002@cus.cam.ac.uk