Speech recognition

November 30, 2006

I had to laugh at the video of the Vista voice recognition launch and the failure of the program to recognise even simple speech. Surprised because my current Microsoft voice recognition software is surprisingly accurate, but also reminded of an experience I had as a child.

In 1962, my family was on its way back to Australia following some time living in New Jersey and we visited family in Seattle during the time that the World’s Fair was on. I don’t remember much of the Fair in detail except for a visit to the IBM pavilion.

IBM were showing off their latest research into the development of speech recognition for their adding machines. Anyone remember the adding machine?

Simple to operate. Enter the numbers and pull the handle to add. The IBM World’s Fair version was the same except for the addition of a microphone, which from memory fitted into a box at the back which must have also had a mechanical connection to the arm. I haven’t yet found any information online about this object.

What I do remember is queuing up to have a turn with the calculator. When my turn came, I took the microphone and spoke in my clearest voice something like “2 plus 5 equals” and then waited. In a moment the mechanism chattered, the handle moved and the machine printed out the answer – 8. The IBM man flustered, fiddled with some tone controls and asked me to try again. Same result. Must be my Australian accent, he reasoned. He then went on to explain that this was very new technology and admitted that it had some way to go before it could be relied on. Don’t worry, we will have voice control mastered in a few years time, he assured.

As a result of this experience, I have been patently waiting for his prophesy to come true. Over the years I have listened to demonstrations of various voice recognition programs such as Dragon as they got better and better. Not perfect, but this year for the first time I have started using the software supplied as part of Microsoft Office 2003. It works reasonably well and I have started to think that the bold future offered by the demonstrator at the IBM pavilion at the Seattle World’s Fair has finally arrive – only it took 44 years and the latest Vista demonstration seems to suggest that we are still not there yet.


4 Responses to “Speech recognition”

  1. Lawrie Zion Says:

    Hi, this is very interesting as I’m writing a doco about the Australian accent and am looking at voice systems that incorporate good Australian speech engines – forgive my ignorance, but what is the Vista launch to which you refer? thanks, Lawrie

  2. cyberslacker Says:

    Sorry, Lawrie. The hyperlink colour is not very clear here. The link to the video is http://www.youtube.com/watch?v=2Y_Jp6PxsSQ&NR

  3. Rena Says:

    Awesome article.

  4. […] 2006, I was amused by the apparent lack of progress in voice recognition. And has it got any better? Well, yes and no. The ‘yes’ applies to Google voice which […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: