Non-Volume Preserving-based Feature Fusion Approach to Group-Level Expression Recognition on Crowd Videos.
Kha Gia QuachNgan LeKhoa LuuChi Nhan DuongIbsa JalataKarl RicanekPublished in: CoRR (2018)
Keyphrases
- feature fusion
- expression recognition
- group level
- facial expression recognition
- facial expressions
- feature extraction
- multiple features
- individual level
- face detection
- group formation
- video sequences
- face images
- video frames
- face identification
- quality prediction
- fusion algorithm
- facial features
- data fusion
- pose variations
- data sets
- data reduction
- similarity measure
- machine learning