A Survey of EEG-Based Driver State and Behavior Detection for Intelligent Vehicles

The driver's state and behavior are crucial for the driving process, which affect the driving safety directly or indirectly. Electroencephalography (EEG) signals have the advantage of predictability and have been widely used to detect and predict the users' states and behaviors. Accordingly, the EEG-based driver state and behavior detection, which can be integrated into the intelligent vehicles, is becoming the hot research topic to develop an intelligent assisted driving system (IADS). In this paper, we systematically reviewed the EEG-based driver state and behavior detection for intelligent vehicles. First, we concluded the most popular methods for EEG-based IADS, including the algorithms of the signal acquisition, preprocessing, signal enhancement, feature calculation, feature selection, classification, and post-processing. Then, we surveyed the research on separate EEG-based driver state detection and the driver behavior detection, respectively. The research on EEG-based combinations of driver state and behavior detection was further reviewed. For the review of these studies of driver state, behavior, and combined state and behavior, we not only defined the related fundamental information and overviewed the research on single EEG-based brain-computer interface (BCI) applications, but also further explored the relevant research progress on the EEG-based hybrid BCIs. Finally, we thoroughly discussed the current challenges, possible solutions, and future research directions.