Unsupervised Drowsy Driving Detection With RFID

• Published in: IEEE transactions on vehicular technology Vol. 69; no. 8; pp. 8151 - 8163
• Main Authors: Yang, Chao, Wang, Xuyu, Mao, Shiwen
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.08.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Alarm systems; channel state information (CSI); deep learning; Driver fatigue; drowsy driving detection; Electroencephalography; Fatigue; Feature extraction; Frequency hopping; Radio frequency identification; Radio-frequency identification (RFID); RFID tags; Traffic accidents; Traffic safety; unsupervised learning; Vehicle safety; Vehicles
• ISSN: 0018-9545; 1939-9359
• DOI: 10.1109/TVT.2020.2995835

With the increasing number of vehicles and traffic accidents, driving safety has become an important factor that affects human daily life. As the primary cause of driving accidents, driving fatigue could be prevented by a sensing and alarm system built in the vehicle. In this paper, we propose an effective, low-cost driving fatigue detection system to sense driver's nodding movements using commodity RFID. The system measures the phase difference between two RFID tags attached to the back of a hat worn by the driver. To accurately extract nodding features, we propose an effective approach to mitigate the environment noise, the interference caused by surrounding movements, and the cumulative error caused by the frequency hopping offset in FCC-compliant RFID systems. A long short-term memory (LSTM) autoencoder is utilized to detect nodding movements using calibrated data. The highly accurate detection performance of the proposed system is validated by extensive experiments in various real driving scenarios.