Home » Posts tagged 'speech recognition'
Tag Archives: speech recognition
I first used Dragon Speaking Naturally, the voice recognition software, when it first came out in 1997. The CEO of Unipart Group of Companies John Neil was trialing it. This and the Internet were his passions (indeed all new technologies he insisted on exploiting for their potential).
Between 1994 and 1998 I spent between one and three weeks a month working here as an outside supplier to the communications team directing regular video news and training ‘films’ that were distributed internally, to suppliers and to shareholders i.e. all their ‘stakeholders.’
Today I find I am still drawing upon the insights I gained on Learning & Training Development at that time. Unipart, lead by John Neil, were the advance guard, pushing employee development further than any other business I was aware of.
Back to Speech Recognition
At the time my only use for it was to read back scripts. Am I talking about the same thing? There was something on my Apple too that by writing phonetically you could have it speaking in a Geordie accent. I recall spending some time trying to ‘teach’ voice recognition software to understand me. My spoken voice is, despite being brought up on Tyneside, recognizably RP … even dare I say Public School / Oxbridge (so no doubt meeting the requirements of my highly aspirational middle class parents). I think it was still ‘tuned in’ to an American intonation.
Why I gave up?
I find the delay (there is some) between the thought and my fingers on the keyboard works, whereas voice recognition was taking it raw from my mind. Stream of consciousness at this Proustian ‘volume’ would then require editing and interpretation, which rather defeated its purpose. My late father on the other hand, a solicitor by training, would dictate letters word perfect, first time. (His mentality and training).
A decade on I am taking it seriously.
I record notes into a digital recorder that could be podcast content for my swim coach blog, by using Dragon Speaking Naturally this can be quickly converted into text and images added.
Its called reader choice
The idea is that poolside a coach needs to have their eyes on the swimmers (rather like a driver having their eyes on the road). If I have offered some simple spoken guidelines on a set, key points and tips, these can be reviewed in situ. Notes on specific swimmers too.
I was woken by the dog, otherwise even I wouldn’t be up this early. I ought to be reading a book rather than doing this … there are ways to ease yourself back to sleep. I tell you, my mind is going like the clappers. I should be mining my dreams right now. Much of the time I am tussling with the content of this blog, the 1200 pieces I’ve ‘dumped’ in my OU eportfolio MyStuff and a desk strewn in white papers, reports, catalogues, directories, hand outs and leaflets from Learning Technologies 2011.
Earlier (see LT2.1) I wondered about a walk around Learning Technologies in the company of a camera on a steadicam. Easier still would have been an informed walk around with audio, as a podcast, with the Floor Plan and pictures.
When am I going to offer this as a podcast? Is there something to be gained in this? Something lost?
Meanwhile I click through 34 voice recordings labelling and deleting.
There is a period here where I deliberately record all, or most of , a swim coaching session, both notes to myself in situ, and my instructions to the swimmers. I’ve done this as an exercise to understand what value can be gained from ‘recording everything.’ That very fact that I am listening to this three weeks after the event indicates one problem. The next is ‘chunking’ the content into manageable pieces, something I did in part while recording (to protect the identities of people I am working with). In practice I can see that such ‘chunking’ should be done at the time … rather like stop starting a dictaphone.
So I learnt something. Whether I record, verbatim, other chunks of my life is quite another matter.
You record 8 hours of material in a day, how many days is this going to take to process.
It reminds me of a story of a diarist who appeared to spend his day writing about writing his diary.
Remind you of anyone?
P.S. Renewing my relationship with Unipart a decade on by clicking through their website I decided to apply directly to their Human Resources department to see about joining them as a consultant.
- How to create accessible e-learning for students with disabilities (mymindbursts.com)
- From Nuance, Siri in your next car (reviews.cnet.com)
- Intel: Voice Recognition Will Phase Out Touch (tomshardware.com)
- Fortemedia Announces World’s Leading Voice Processing Solution for Reliable Voice Recognition Targeting “Hands-free” Mode in Smartphones (virtual-strategy.com)
- Review: Dragon NaturallySpeaking lets you take your hands off the keyboard (pcworld.com)