In one embodiment, one or more systems may receive input from a user identifying an interactive region of a physical environment. The one or more systems may determine a location of the interactive region relative to a depth sensor and monitor, at least in part by the depth sensor, the interactive region for a predetermined event. The one or more systems may detect, at least in part by the depth sensor, the predetermined event. In response to detecting the predetermined event, the one or more systems may initiate a predetermined action associated with the predetermined event.