Yes, they can hear the sounds simply by looking at the bag, even through a sound-proof glass door, no sound recording needed whatsoever.
"That's the results of the astounding visual microphone," project done by researchers at MIT, reports by Sarah Lewin for IEE Spectrum's blog.
It works like this: All sound creates vibrations. Those vibrations move your eardrum and that's why you can hear.
Sound vibration can also microscopically move other objects in the room. The researchers used a high-speed camera to film such objects and wrote an algorithm to translate the tiny movements captured on film back into the sounds that created them.
Some objections captured visual sounds better than other. For instance a bag of chips, great; a potted plant, mediocre; a soda can, bad.