An improved valence-arousal emotion space for video affective content representation and recognition.
Kai SunJunqing YuYue HuangXiaoqiang HuPublished in: ICME (2009)
Keyphrases
- emotional state
- affect recognition
- emotion recognition
- multimedia
- space time
- facial expressions
- recognition rate
- online video
- video data
- temporal domain
- video sequences
- physiological signals
- object recognition
- recognition accuracy
- human activities
- feature extraction
- affective states
- video streams
- affective computing
- multimedia data
- real time
- video content
- character recognition
- video analysis
- video frames
- image representation
- object representation
- textual descriptions
- low dimensional
- web content
- video footage
- appearance manifolds
- human emotion
- lecture videos
- temporal structure
- user generated
- temporal information
- activity recognition
- action recognition
- human computer interaction
- computer vision