TomFi: Small Object Tracking Using Commodity WiFi

Rodent infestation has always been one of the most severe threats to humans, solve which consumes massive manpower and resources. People usually use traps or poisons treat rat infestation. Such passive countermeasures are inefficient. Damages often occur when rats finally trapped killed, not mention potential risk injure humans cause pollution. In this paper, we propose a WiFi-based active small object tracking system, namely TomFi. The core components TomFi include number deep learning techniques that bridge locations/motions WiFi signal variations (represented by channel state information). first employs detection model detect appearance rat, then localize it based on two-branch localization model. Through design model, information loss distortion problem. We conduct extensive experiments in both laboratory environment real-world kitchen scenarios. results show can achieve success rate \(99\%+ \) centimeter-level accuracy real time.