Single Image Action Recognition by Predicting Space-Time Saliency.
Marjaneh SafaeiHassan ForooshPublished in: CoRR (2017)
Keyphrases
- single image
- space time
- action recognition
- human actions
- spatio temporal
- multiple images
- spatial and temporal
- bag of words
- video sequences
- image pairs
- motion patterns
- human detection
- light source
- shape from shading
- d scene
- body parts
- super resolution
- computer vision
- visual attention
- activity recognition
- dynamic scenes
- saliency map
- bag of features
- human activities
- view invariant