Boosting particle filter-based eye tracker performance through adapted likelihood function to reflexions and light changes

Keyphrases

• Eye Tracker 100%
• Lighting Changes 100%
• Particle Filter 100%
• Likelihood Function 100%
• Difference Image 100%
• Pupil Image 100%
• Filter Method 100%
• Bright Pupil 100%
• Head Movement 50%
• Tracker 50%
• Image Sequence 50%
• Prior Selection 50%
• Initialization Method 50%
• Likelihood Ratio 50%
• Background Modeling 50%
• AdaBoost 50%
• Foreground Modeling 50%
• Cascade Classifier 50%
• Log-likelihood Ratio 50%
• Long Head 50%
• Static Threshold 50%
• Dark Pupil 50%
• Eye Image 50%

Engineering

• Difference Image 100%
• Likelihood Function 100%
• Particle Filter 100%
• Good Performance 50%
• Eye Image 50%
• Log-Likelihood Ratio 50%
• Adaboost 50%
• Test Subject 50%
• Image Sequence 50%