In principle, you could have a system designed to image your surroundings (using cameras, LIDAR, etc) and perhaps some kind of machine learning algorithm to predict what kind of sound would be expected if the events around you were occurring in atmosphere. I imagine this could work well for simple things like a tool hitting a piece of metal, but would be probably run into issues when the events are affected by the lack of atmosphere or give little or no visual indication that they are occurring. And, of course, you wouldn’t be able to “hear” anything outside of the view of your imaging system.