Avatar-mediated face tracking and lip reading for human computer interaction
Proceedings of the 12th annual ACM international conference on Multimedia
Toward a decision-theoretic framework for affect recognition and user assistance
International Journal of Human-Computer Studies - Human-computer interaction research in the managemant information systems discipline
Robust facial feature tracking under varying face pose and facial expression
Pattern Recognition
Using EEG spectral components to assess algorithms for detecting fatigue
Expert Systems with Applications: An International Journal
Driver alert state and fatigue detection by salient points analysis
SMC'09 Proceedings of the 2009 IEEE international conference on Systems, Man and Cybernetics
Comparing combinations of EEG activity in train drivers during monotonous driving
Expert Systems with Applications: An International Journal
Facial event classification with task oriented dynamic Bayesian network
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
An automated face reader for fatigue detection
FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition
Estimation of driver's fatigue based on steering wheel angle
EPCE'11 Proceedings of the 9th international conference on Engineering psychology and cognitive ergonomics
Kernelized Fuzzy Rough Sets Based Yawn Detection for Driver Fatigue Monitoring
Fundamenta Informaticae - Knowledge Technology
Hi-index | 0.00 |
The vision-based driver fatigue detection is one of themost prospective commercial applications of facial expressionrecognition technology. The facial feature tracking isthe primary technique issue in it. Current facial trackingtechnology faces three challenges: (1) detection failure ofsome or all of features due to a variety of lighting conditionsand head motions; (2) multiple and non-rigid objecttracking; and (3) features occlusion when the head is inoblique angles. In this paper, we propose a new active approach.First, the active IR sensor is used to robustly detectpupils under variable lighting conditions. The detectedpupils are then used to predict the head motion. Furthermore,face movement is assumed to be locally smooth sothat a facial feature can be tracked with a Kalman filter. Thesimultaneous use of the pupil constraint and the Kalmanfiltering greatly increases the prediction accuracy for eachfeature position. Feature detection is accomplished in theGabor space with respect to the vicinity of predicted location.Local graphs consisting of identified features are extractedand used to capture the spatial relationship amongdetected features. Finally, a graph-based reliability propagationis proposed to tackle the occlusion problem andverify the tracking results. The experimental results showvalidity of our active approach to real-life facial trackingunder variable lighting conditions, head orientations, andfacial expressions.