Thread: Voice control software, what one?

I've been writing this into my own software to try get as close to 100% voice control as possible with a little bit of Ir for backup, currently have pretty much all media playing commands supporting voice, all your albums can be played with 'play <artist> <album>' which then also adds 'play <track>' voice commands for the album.

Have been adding mapping/navigation system so today I decided to add voice street directory support, have only done a few quick tests (because it takes approximately 2.5 hours to compile the ~32,000 streets in my city into a speech engine but after thats been done once for each map loads speech data in about 10 seconds), only added 'find <street> <suburb>' so far but it seems to be working pretty good, the downside with having so many possible matches is it decreases the accuracy, the more words in a phrase and the less phrases to match against the more accurate it is.

Anything dictation based is pretty much useless with todays technology, the only real way to get close to 100% is supply it with a list of known commands to match against instead of trying to turn spoken words into what could be any text.

Sounds good

But what could I use for now?

Yeah, I'm also wondering what can be used. I'm thinking about creating a custom interface (geared for what I would like it to do.. but I'd release it to the public if it actually works). I've had experience with working with VB, and C++.. .and I'm gonna try to write this mofo in C#. I'd like to do as little work as possible, of course ... so if there's already a good program out there that can accept voice inputs, that'd be great. So what progs would people recommend?


My vision: A button on the steering wheel, that you hold down and give the command "load playlist <name of list>" and simple commands such as "fast forward 10 seconds" etc. Windows has voice recognition built in, but I think it'd be easier to just have voice software that'd do it for me, rather than try to code in their API into my attempt. Any ideas are welcome.

It's a fun program to play around with, but unfortunately it doesn't have an option to type out what you say, only a predetermined setting. So I can get everything up to "computer, load M P 3" but I'd have to have the mp3's name pre-scripted. I can't just say the name of the MP3 and have it typed into the Open dialogue. I guess I'll have to figure out on my own how this damn SAPI stuff works. Oy, it's been like half a decade since I've done any programing... =\

True, you can send individual keystrokes... but that's kinda tedious. Plus, I'm not always exactly sure on the precise spelling of some of the mp3s right off the bat. VoiceFlux is great... I just wish one of the actions it had was a "Dictate for x seconds" one, where for x amount of seconds it would just type out whatever you said.

I downloaded the MS SDK, and realized just how much I'd forgotten about programing. =P I can make the front-end GUI and make it interactive, but I'll have to relearn a lot to make the program work with the hardware and to incorporate the speech recognition.