LARNet: Latent Action Representation for Human Action Synthesis.
Naman BiyaniAayush Jung RanaShruti VyasYogesh RawatPublished in: BMVC (2021)
Keyphrases
- human actions
- recognizing human actions
- action recognition
- view invariant
- recognition of human actions
- spatio temporal
- human activities
- action classification
- human motion
- action recognition in videos
- space time
- activity recognition
- action sequences
- video sequences
- recognizing actions
- spatio temporal interest points
- visual features
- motion recognition
- bag of words
- human pose
- motion primitives
- computer vision
- single view
- sensor data
- viewpoint
- bag of features
- latent variables
- action detection
- human object interactions