My team is thinking on making a Voice Controlled Metronome, instead of turning the nob to set the beats, you say the number.
Our current plan is use a voice chip recognition...which connect to the arduino and the output will connect the audio amp/potentiometer(for volume control) setup for the sound.
instead of turning the nob to set the beats, you say the number.
output will connect the audio amp/potentiometer(for volume control)
What are you trying to control my voice commands?
Our current plan is use a voice chip recognition
What chip?
the problem with a Voice Controlled Metronome is getting the number out of the voice recognition system. This is especially difficult in that the number will contain either two or three digits. Voice recognition on a chip is not as good as you might hope it to be.
I would pick something that you stood a chance of completing.
On the "make it easier" tack ...
How about clapping the beat, and have it repeat?
Still not trivial, but easier. Might be more fun too 
GB
What chip?
We were looking at the HM 2007 one that was pre bulit for us, so we could focus on the code
What are you trying to control my voice commands?
When you say "60", the metronome will output 60 beats(in sound)
On the "make it easier" tack ...
How about clapping the beat, and have it repeat?
Still not trivial, but easier. Might be more fun too Smiley
GB
that does sound nice
When you say "60", the metronome will output 60 beats(in sound)
You do realize that the HM2007, when used in speaker-independent mode, has a vocabulary of only 10 words, right?
So - you wouldn't be able to say "60" (unless that is a common BPM?), but you could say "six-zero" (and anyone else could, as well).
The only problem here is that while you could get it to be speaker-independent for 10 words (numbers 0-9), you wouldn't have any "mark" or "command" words to tell it to "go" (you would probably have to use a timeout or something; but then again, you wouldn't have any way to tell it to "start", either).
The 40 word vocabulary of the HM2007 is only available for speaker-dependent use; if only one person -ever- will use this, and you don't mind re-training the metronome between users (or each user would have their own), then the system could potentially be more versatile, at the expense of needing to be trained before usage.
Maybe you should go for a hybrid approach; use those 10 speaker independent words for commands: "set beat", "play beat", etc - then have it listen for clapping for the beat via a separate system (which would be similar to a door knock detector) activated by the command spoken.
You do realize that the HM2007, when used in speaker-independent mode, has a vocabulary of only 10 words, right?
So - you wouldn't be able to say "60" (unless that is a common BPM?), but you could say "six-zero" (and anyone else could, as well).
The only problem here is that while you could get it to be speaker-independent for 10 words (numbers 0-9), you wouldn't have any "mark" or "command" words to tell it to "go" (you would probably have to use a timeout or something; but then again, you wouldn't have any way to tell it to "start", either).
The 40 word vocabulary of the HM2007 is only available for speaker-dependent use; if only one person -ever- will use this, and you don't mind re-training the metronome between users (or each user would have their own), then the system could potentially be more versatile, at the expense of needing to be trained before usage.
Maybe you should go for a hybrid approach; use those 10 speaker independent words for commands: "set beat", "play beat", etc - then have it listen for clapping for the beat via a separate system (which would be similar to a door knock detector) activated by the command spoken.
Thank you for the insight on the HM 2007, do you know of any chips that can be used they way i stated, or do you think the "door knock" method is better :-/
thanks for the help everyone