I was wondering how to make your own speech and object recognition. I've been working on this project and it uses both of them. Sorry for vagueness but I don't know where to start and thought this might be the best place. I just need like one example to get how it works.
Just program your Artificial Intelligence first and let it write its' own AV recognition code!
Kidding, but you do realize that there are many big brains with million dollar budgets who still haven't successfully solved these issues right? Voice recognition and visual pattern matching are still cutting edge topics. I think your best bet is to check out the robotics hobbyist sites.
BTW, my favorite phrase from an old Scientific American article on why human speech is so hard to translate by a computer program:
"Time flies like an arrow, but fruit flies like a BANANA!"
A couple of months ago, I made an simple AI program (using c++) which learned from user input, and had no native language recognotion skills. I think you could maybe feed it lots of bitmaps and tell it what they are, and then it would recognise one without you saying what it was (it would need a lot of training though).
Is that the sort of thing you are looking to do with your AI?
Hmmn...
Well, I could send it to you if you wanted...
It would need major modifications to read images, though - at the moment it just uses std::cin for input. Also, it is really, really stupid in the beginning (think newborn baby), so don't get impatient.
This is funny, because I am currently trying to design a thing you can speak to, using the same base program that my AI uses.