ActAR: Actor-Driven Pose Embeddings for Video Action Recognition.
Soufiane LamghariGuillaume-Alexandre BilodeauNicolas SaunierPublished in: CoRR (2022)
Keyphrases
- action recognition
- action detection
- human actions
- human pose
- action classification
- recognizing actions
- spatial temporal
- video dataset
- static images
- recognizing human actions
- motion features
- recognition of human actions
- bag of words
- activity recognition
- human detection
- body parts
- human activities
- computer vision
- space time interest points
- mid level
- pose estimation
- video sequences
- human object interactions
- multimedia
- video frames
- motion history images
- depth sensors
- video streams
- bag of features
- view invariant
- atomic actions
- space time
- event recognition
- visual words
- d objects
- video surveillance
- feature space
- video retrieval
- video clips