I'm planning to develop a voice recognition technology that requires minimal effort for programmers to integrate into their mobile applications (Minimal to no coding at all).
The technology will also have a user adaptation model that allows it to customise itself to the user's voice and speech patterns.
Are there other problems you would like solved or features you want implemented in voice recognition technologies?
Do help me complete a short survey (9 questions only!!!) to share your views?
All survey participants will get a free trial of the product if this project comes to fruition.
Thank you for your valuable feedbacks.
Just make it work right and don't release it until it's surprisingly good to all who use it. No one needs another mediocre speech recognition system.
Remember when everyone thought computers were supposed to make our lives easier. I think it's time to try to make that happen.
Though probably more appropriate in the Joke Thread, imagine sitting at a computer with great voice recognition, and uttering "well, kiss my grits [insert word of choice here].
Every time that I drive my wife's new car, with voice recognition, I have to watch my language. That dang car has so many voice-activated devices, that I worry a lot.
Bill, may we go ahead and use this thread to discuss voice recognition software already available? I mean, that sure is what I thought the thread title was referring to. "Survey" should have been in the title to make the purpose of the OP clear, but it is now what it is, so maybe we can use it to launch a discussion?
Or is there another thread around here where the discussion has been going for a bit?
And I'm asking you, Bill, because you seem to be the ranking member thus far in the thread.
Except that rank is weird, if it is intended to use the vocabulary "most" because there seems to be a lot of mosts around these parts and each new most reduces the value of the mosts that came before. But sliding off-topic, yes? Best stop with that line of thought.
EDIT: Just looked at DOC-2327 and see that "most" is part of that rank. So I want to know who is the mostest?
Sorry for the late response.
I think that a discussion of existing VR software would be quite fine. The OP might pick up some ideas, based on that discussion.
I will even start. We have had Mercedes with VR software for certain functions, for several models and the last 10 years. While the older versions were a bit limited in their commands, the newer versions seem more sensitive. My wife's new car will not let me verbally enter in a telephone number, where her older one, had no issues. Maybe it is a matter of "talking" to it often enough, to get a voice print. I do not mumble, nor do I have a speech impediment, and even do a fair amount of Voice-Over work, but the danged car cannot understand me, where it has no issues with my wife.
Just got a new Lexus with VR for many things. Again, it seems to be spending a lot of time memorizing MY voice. My wife has not spent much time in it, but I will be curious how it treats her. As both respond with a female voice, I wonder if they just like to mistreat males?
"Open the pod bay door HAL... "
As for "senior member," that is just because I am the oldest living Adobe user on Earth...
Good luck to the OP, and as Noel is a software developer, he probably has some useful ideas for you.
I've contributed my piece up above.
But Bill brings up good examples of why speech recognition needs to be exceptional.
One day we will have computer systems that are utterly and completely better and listening and speaking than even the best of us. For now, we still have cars that do the wrong things half the time, regardless of what commercials show. God help us when someone inadvertently/prematurely wires something physically important to a computer that can be controlled by voice (thinking along the lines of "Blue Screen Of Death" taking on a whole new meaning).
Thank you for the response, Bill. I haven't exactly been Cpt. Quick myself here in responding. Everybody's busy, right?
Anyway, I am very interested to find out what voice recognition product is worth buying or trying. Maybe even a good freeware product.
I installed, with some difficulty, the Microsoft product that is availbale for Vista Ultimate, and it's just not going to make the grade. I worked very hard at "teaching" it the bla-bla-bla that was being asked of me by the M_soft engineers that made that product and it's just way too buggy.
I even started testing it just for research purposes and could trick it into writing what I wanted, but it was way too much trouble.
What I am looking for is a product that I can create memorandums with. In other words, I talk and the nice voice recognition software gets most of what I say into a text format and then I can go through it and edit what needs editing.
I don't expect it to be perfect. Far from it. But I don't want to have to be spending all my time choosing this or that word or letter or number etc. I just want to talk in a nice smooth manner and in a very clear voice, which I can easily do, and the nice software gets even just 70% of what I say without constantly bugging me for input as I am talking.
Of course, what I am asking here may be beyond any software out there right now. But that's whay I asked if we could discuss this business in this thread.
AND this was not generated by a voice recognition system!
In fact, it is generated by two fingers that really need to wear glasses, but they refuse to. They are constantly making mistakes to make me look bad.
I did not read all the replies (time), but I will say a few years back we were doing a chapter on this topic in a book on game design. After talking to large game companies like EA and Blizzard, they recommended AT&T technology at that time.
My children recognize my voice only when I say something they want to hear.
As a total outsider of the voice recognition field, I had a pleasant experience with Vonage's message taking and converting voice to text.
Europe, Middle East and Africa