Multimodal Human Emotion Recognition
MULTIMODAL HUMAN EMOTION RECOGNITION
- Development of fast and accurate human emotion recognition with multimodal inputs
- Hybrid margin-based region of interest (MROI) for fast face detection, fallback template matching for robustness, visual and audio nets, hybrid deep neural network and convolutional neural network
- Technologies: computer vision, face detection, facial landmarks detection, audio features extraction, deep learning Collaboration: University of Prince Edward Island, Canada
- Funding: UBD FIC Grant
- Output: 4 publications
