Temporal-Relational CrossTransformers for Few-Shot Action Recognition.
Toby PerrettAlessandro MasulloTilo BurghardtMajid MirmehdiDima DamenPublished in: CVPR (2021)
Keyphrases
- action recognition
- spatial temporal
- motion history images
- human detection
- human actions
- computer vision
- bag of words
- activity recognition
- recognizing human actions
- spatio temporal
- action classification
- static images
- body parts
- bag of features
- depth sensors
- temporal information
- recognition of human actions
- action recognition in videos
- action primitives
- video sequences
- human pose
- mid level
- temporal relations
- video content
- video dataset
- human activities
- view invariant
- independent subspace analysis
- atomic actions
- temporal structure
- key frames
- space time
- visual features
- low level
- keywords