Meet my new pet: a sweet little dragon!
This is my first test recording with my new toy Dragon NaturallySpeaking — after the Microsoft speech recognition software didn’t work out so well I had to try something new.
When I saw on eBay an auction for the software, I was unhappy about the fact that it was a German version. To my surprise, there where also auctions with “Dragon NaturallySpeaking Preferred” in a US version available. Needless to say that this was my preferred choice.
This sentence was supposed to display house. Out Dragon PDAs over Microsoft’s speech-recognition
[This sentence was supposed to display how superious Dragon is over Microsoft’s speeck recognition … well. Let’s try that again!]
The reason why I am eager to train the software onto my voice is simple:
if I have trained it well it will be much faster than me typing and I can produce more postings. Hey! I did say a blog posting.
No, I did not say “a” blog posting – I just said blog posting.
So far, I am quite surprised — pleasantly — that the software actually knows blog. Okay, it still needs some training between blog and block.
And the command for a new line (especially when I want two of them) should really be “enter enter” and not “new line new line”.
As far as punctuation goes I think it is easier to keep my hands on the keyboard and insert those – no need to do every thing with voice.
What will be more difficult for the system rather than for myself? The fact that I don’t keep to what the software probably considers to be good word composition – we will see if we can train him on that! :-) (did you know that they had a “smiley face” comand built in?)
One of the things I did not like was during installation was when it was trying to get a feel for how i write. It said “should I go through emails and documents” but then it just went through my personal folder and indexed everything. For normal user this might be okay, it is just that I keep every kind of document in there, for example saved Web pages with hotel reservations, a collection of spam messages for later analysis, and many more things which definitely should not be considered my writing style.
But those are the first baby steps into getting this software up and running. The reason I bought Dragon was simple: even if it just takes most of what I run to write down and I have to do manual corrections, it still should buy me considerable time when writing blog postings. And isn’t blogging at about a conversational style anyhow?
So using any speech-recognition software seems to be the perfect approach for podcast also. I tried to transcribe one of my Useful Sounds but that did not go too well … let’s just say that my river of thoughts is best to be listened to and not to be read in transcribed form. ;))
So from the first minutes of trying out this software it actually seems that I can do some work with this. It needs a lot of training but I can manage that. I will just throw in the training center some of my blog postings and read them out loud. (I might want to check if it knows words like put costing podcasting before I start typing – see what I mean? Where is the list of new media names when you need one?)
One way in the future of telling that I did use the speech-recognition software? A proper usage of the apostrophe on words like don’t / can’t / haven’t here as well as correctly written words like German, English, French with a capital letter. ;)
Other than that the other software I tried out before this program was much more fun. When I sayed “computer” it did “wake up” but this one just spells out computer – boo-hoo!
Hm. I need to train the software to recognize the expression hm. And it should allow me to have a technorati comand which will insert the tag including the HTML code into the current blog posting. Interesting!
I see a new Armada of commands coming like “ping technorati”, “go to my space”, or even “pay my rent in Second Life”.
After all this is 2006 and we should finally makes use of what technology is offering us.
= the (that was supposed to be my name — let’s try that again)
Nicole Simon, who now has to train English terms of punctuation …
yes this took long oh then just typing it down … but then again, I did not have to stress my hands with typing. Imagine you lying in bed and slowly dictating your blog posts one by one …
Technorati Tags: speech recognition, msspeech, dragon natural speaking
Tags: tools
I have tried text recognition, but actually, I hate using it. I simply can’t speak texts that are meant for publishing in whole sentences. It’s just not how my mind works.
The above four sentences have been heavily edited, restructured, another sentence inserted, etc. All in the time that others I know would use to speak the same sentence (yes, I can type reasonably fast). Well, maybe not quite so fast, but fast enough for me to not get noticed when chatting in IRC :-).
Yes, I’m sometimes a bit of an awkward speaker, when prompted to speak freely suddenly. Either I keep loosing words in a sentence, or leave some sentences hanging, while my mind changes subject, or I speak deliberately a bit more slowly (and coherently, which is taxing my own patience.
But it’s a nightmare when using text recognition.
Besides, I feel silly speaking to my computer.
I already do podcast, so the silly part is already taken care of. ;)
As for the rest – I had a transcript, horrorfull. And I am trying to think if the way I am typing for example this answer could be ‘prewritten’ and then modified afterwards. ;)
this comment was written 100% by hand. ;)
Hi
I resell Dragon NaturallySpeaking in the UK and am a trainer. I have recently posted some Dragon demonstrations on my website http://www.speechempoweredcomputing.co.uk. You might be interested in viewing the demo in MS Word to see the level of accuracy and response that Dragon NaturallySpeaking (v9) is capable of.
Best wishes
Peter