Analysis of Yawning Behaviour in Spontaneous Expressions of Drowsy Drivers

Abstract

Driver fatigue is one of the main causes of road accidents. It is essential to develop a reliable driver drowsiness detection system which can alert drivers without disturbing them and is robust to environmental changes. This paper explores yawning behaviour as a sign of drowsiness in spontaneous expressions of drowsy drivers in simulated driving scenarios. We analyse a labelled dataset of videos of sleep-deprived versus alert drivers and demonstrate the correlation between handover-face touches, face occlusions and yawning. We propose that face touches can be used as a novel cue in automated drowsiness detection alongside yawning and eye behaviour. Moreover, we present an automatic approach to detect yawning based on extracting geometric and appearance features of both mouth and eye regions. Our approach successfully detects both hand-covered and uncovered yawns with an accuracy of 95%. Ultimately, our goal is to use these results in designing a hybrid drowsiness-detection system.

Sequential Nonlinear Encoding; A Low Dimensional Regression Algorithm with Application to EEG based Driving Fatigue Detection

“…The questionnaires are usually filled before and after a tedious task where the performance differences can reveal the extent of the fatigue state. (b) video based techniques that utilize such physical symptoms as yawning, pattern motion of eyelid, eye, and head as well as facial and eye expression [17,18,19,20]. (c) cognitive tasks that use the reaction time or the error rate of the responses to a set of visual stimuli [21,22,23].…”

“…sleep, drowsy, alert, near alert, etc.) [18]. More recently, few studies have addressed the problem of continuous fatigue detection (in the framework of a regression problem).…”

Yawn Based Driver Fatigue Level Prediction

“…In addition, our model some time cannot identify the difference between a long-time open mouth and a yawn. But compared to [19], [20] , [21] and [22],it shows that these errors happened because of some drivers unconsciously hide their mouth by putting the hand in the mouth. This suggests that Resnet is good at learning temporal relations on high level features.…”

“…The phenomena of yawning are commonly associated with tiredness and fatigue [7]. It is signified by the unintentional opening of the mouth.…”

“…Accuracy Dataset [19] 92% YawDD [20] 75% YawDD [21] 88.6% YawDD [5] 99.80 % Self-collect data [22] 94.63% YawDD The second stage (the binary yawn detector), in this step the frame is being classified into yawn to continue to the next step or non-yawn, to navigate back to the previews step. The third step is to count the yawns in one-minute time to output the level of: Alert, when the driver does not yawn and then goes back to step one.…”