High-Level Geometry-based Features of Video Modality for Emotion Prediction.
Raphaël WeberVincent BarrielleCatherine SoladiéRenaud SéguierPublished in: AVEC@ACM Multimedia (2016)
Keyphrases
- high level
- low level
- video clips
- video sequences
- video analysis
- prediction accuracy
- video content
- feature space
- mid level
- co occurrence
- video retrieval
- temporal information
- key frames
- video images
- semantic concepts
- prediction error
- video data
- higher level
- multi modal
- feature set
- facial expressions
- image features
- feature vectors