I think the number of commands limited by two reason:
time of "recognition" is directly proportional to this number;
probability of "false detection" also proportional to size of vocabulary.
If both this factors are not critical to your application, you could try arduino as a voice recognition w/o external shield http://arduinovoicerecognition.blogspot.se/
, but you'd still need external memory chip, spi or i2c eeprom / flash / sd, 1kb per 1 word.
direct99:
Are there any better Voice Recognition modules?
You used "better" a couple of times now. "Better" is entirely relative. What is you want to be able to do? That's the only way to answer what is "better".
Do you mean one or two word command sequences ('Play music' or 'Lights on', etc), or are after you natural speech recognition ('Arduino - close the curtains, set the lighting level to thirty percent and turn the tv on to channel 4')?
Do you mean one or two word command sequences ('Play music' or 'Lights on', etc), or are after you natural speech recognition ('Arduino - close the curtains, set the lighting level to thirty percent and turn the tv on to channel 4')?
Having a natural speech recognition throught Arduino is Great but is it possible though?
I just want my Arduino to recognize four or five word command sequences.
I also wants it to recognize as many commands as possible.
Please stop posting the same question - including the ones in the recycle bin, that's five times you've posted the same thing.
See your thread in "microcontrollers".
@OP: People had already taken their time and effort and had a go at answering your question when you dumped the thread in the recycle bin.
Don't you think that a little rude?
I do.
(but don't worry, I recovered it from the bin, dusted it off a bit, and merged it with this thread)