View article

This paper presents a real-time gesture recognition technique based on RFID technology. Inexpensive and unintrusive passive RFID tags can be easily attached to or interweaved into user clothes. The tag readings in an RFID-enabled environment can then be used to recognize the user gestures in order to enable intuitive human-computer interaction. People can interact with large public displays without the need to carry a dedicated device, which can improve interactive advertisement in public places. In this paper, multiple hypotheses tracking is used to track the motion patterns of passive RFID tags. Despite the reading uncertainties inherent in passive RFID technology, the experiments show that the presented online gesture recognition technique has an accuracy of up to 96%.