If you properly filter your power supplies you needn't worry about RF or AC from the mains with the low impedance that you're using. However, you should probably load the inputs with a small capacitor (~1 nF) and definitely use a Schmidt-trigger inverter such as 40106/74C14. That will slow the switch closure and minimize the effect of bounce, perhaps even swallowing it completely.

Unfortunately, you've not explained what the noise source is and why you're concerned about it. Unless the ambient noise is quite low-impedance, you shouldn't see it in the circuit you're using, particularly if you capacitor-damp the schmidt-triggered input.

Keep in mind, too, that the 40106 will happily work at 12 volts, and will tolerate a very slow rise time, as would result from a capacitor-damped input. I would attenuate the inputs before capacitor-damping them, and feed that to Schmidt-triggers operating at the local supply voltage, however.

It matters quite a bit where the noise is originates, i.e. is it from the power supply, or the cable, on-board logic, or the environment. If that's not known, you can find it easily enough by means of a spectrum analyzer, which will show you the relative power in each frequency band. With a low-resistance 40 cm jumper, you can find the source at the boundaries of your PCB, and that same jumper can help you find appropriate sites for bypass to suppress that noise on your board. Inline ferrite beads can help with noise on your cables.

RE