Human Action Recognition from Depth Videos Using Pool of Multiple Projections with Greedy Selection.
Chien-Quang LeSang PhanThanh Duc NgoDuy-Dinh LeShin'ichi SatohDuc Anh DuongPublished in: IEICE Trans. Inf. Syst. (2016)
Keyphrases
- action recognition
- human actions
- spatio temporal interest points
- human activities
- static images
- action classification
- video dataset
- recognition of human actions
- human object interactions
- depth sensors
- motion history images
- motion capture data
- bag of words
- human movements
- motion features
- recognizing human actions
- ucf sports
- activity recognition
- action detection
- recognizing actions
- body parts
- view invariant
- computer vision
- human detection
- depth cameras
- space time interest points
- action recognition in videos
- spatial temporal
- human activity recognition
- human pose
- depth information
- spatio temporal
- bag of features
- mid level features
- human motion
- action primitives
- view invariant action recognition
- machine learning
- three dimensional
- d objects
- multi view
- human body
- sensor data
- depth images
- complex events