Jennifer Lee Final Automated Detection of Human Emotion
Goal To be able to identify emotions using a low quality camera (webcam). To be able to identify emotions using a low quality camera (webcam). Applications Applications Human-Computer Interaction Human-Computer Interaction Alternate Reality Alternate Reality Product Testing Product Testing
Past Research Very good results Very good results About 80-90% accuracy About 80-90% accuracy Generally have access to high quality cameras Generally have access to high quality cameras Two visual-based processes Two visual-based processes Marker based Marker based Shadow based Shadow based AngerSadnessHappinessNeutral Anger Sadness Happiness Neutral Analysis of Emotion Recognition using Facial Expressions, Speech and Multimodal Information (2004)
Development Python (OpenGL, PIL) Python (OpenGL, PIL) Webcam Read each image Identify Markers Produce Tracking Image Analyze Tracking Image GUI Lighting Adjustments Feature Isolation Head Shift Adjustments
Progress Python webcam integration Python webcam integration Tracking Tracking Basic marker identification Basic marker identification Basic marker tracking Basic marker tracking Head movement compensation Head movement compensation Detailed marker tracking Detailed marker tracking Identification Identification Basic Identification Basic Identification Learning Learning GUI GUI