you could check if the input level is below a treshold for an amount of time - this only works of the environment isn't too noisy

You can check that mouseX is the same than pmouseX for a given time (say a couple of seconds).

thanks for the ideas. I am hoping to eliminate any need for mice/keyboards in the actual gallery setting.

The input level suggestion from local_guru might work... on a day to day gallery basis this is probably fine, on more crowded days it might be a bit more dodgy. The key might be finding the right mic for the situation.

I'm planning to do some testing this week.

tone generator near the mic to announce a new person (a little buzzer that emit a tone when pressed that can be picked up by minim) - or just one button by the mic (to minimize the input presence) that could be built really easily with a small arduino board and a button... 

Hi,

If part of the concept is to use a microphone as input I would use a program that can deal properly with the sound (like ex. SuperCollider or Pure Data) as audio input, filter the sound to the vocal range, add a gate, measure the amplitude and frequencies, add a lag (easing) to the amplitude signal and then send it as OSC to Processing.

A directional (and thus probably expensive) microphone will save you lots of trouble. I take it that you dont want to capture the sound of the whole room but have an area of interaction. Shotgun mics are great and expensive, but you can also achieve surprisingly  good results with a mic inside a parabola (like a big plastic kitchen bowl).

good luck - and tell us how it goes!
martin

Thank you for all of your suggestions. I am so glad I posted -- this gives me many more options to explore. I'll let everyone know how my tests go!

-rebecca