Drowsiness is a major cause of accidents, in particular in road transportation. It is thus crucial to develop robust drowsiness monitoring systems. There is a widespread agreement that the best way to monitor drowsiness is by closely monitoring symptoms of drowsiness that are directly linked to the physiology of an operator such as a driver. The best systems are completely transparent to the operator until the moment he/she must react. In transportation, cameras placed in the passenger compartment and looking at least at the face of the driver are most likely the best way to sense physiology related symptoms such as facial expressions and the fine behavior of the eyeballs and eyelids. We present here the new database 1 called DROZY that provides multiple modalities of data to tackle the design of drowsiness monitoring systems and related experiments. We also present two novel systems developed using this database that can make predictions about the speed of reaction of an operator by using near-infrared intensity and range images of his/her face.

Drowsiness is a major cause of fatal accidents, in particular in transportation. It is therefore crucial to develop automatic, real-time drowsiness characterization systems designed to issue accurate and timely warnings of drowsiness to the driver. In practice, the least intrusive, physiology-based approach is to remotely monitor, via cameras, facial expressions indicative of drowsiness such as slow and long eye closures. Since the system’s decisions are based upon facial expressions in a given time window, there exists a trade-off between accuracy (best achieved with long windows, i.e., at long timescales) and responsiveness (best achieved with short windows, i.e., at short timescales). To deal with this trade-off, we develop a multi-timescale drowsiness characterization system composed of four binary drowsiness classifiers operating at four distinct timescales (5 s, 15 s, 30 s, and 60 s) and trained jointly. We introduce a multi-timescale ground truth of drowsiness, based on the reaction times (RTs) performed during standard Psychomotor Vigilance Tasks (PVTs), that strategically enables our system to characterize drowsiness with diverse trade-offs between accuracy and responsiveness. We evaluated our system on 29 subjects via leave-one-subject-out cross-validation and obtained strong results, i.e., global accuracies of 70%, 85%, 89%, and 94% for the four classifiers operating at increasing timescales, respectively.