SOSense aims to save lives and prevent irreversible harm by improving the safety of swimmers, especially children, in private and public pools. The SOSense system constantly listens to underwater pool sounds, learning each pools unique environmental noises (e.g. cleaning robots, water purifiers, and pumps) in order to filter out those sounds for better human detection. The system continuously monitors streamed data from the pool through SOSenses local processor and remote cloud. The system enters Pool Entry Mode when no human sounds are recognized, and an alarm will sound when a human (as opposed to wind, pets, or objects) is detected in the water. In Drowning Detection Mode, the system recognizes humans in the pool and monitors for bodily distress signals.