Flexible human action recognition in depth video sequences using masked joint trajectories.
Antonio Tejero-de-PablosYuta NakashimaNaokazu YokoyaFrancisco Javier Díaz PernasMario Martínez-ZarzuelaPublished in: EURASIP J. Image Video Process. (2016)
Keyphrases
- action recognition
- human actions
- motion capture data
- video sequences
- view invariant
- static images
- depth sensors
- human movements
- body parts
- human activities
- space time interest points
- video dataset
- human detection
- depth cameras
- activity recognition
- bag of words
- computer vision
- moving objects
- recognition of human actions
- spatial temporal
- action classification
- motion trajectories
- human pose
- motion recognition
- motion history images
- recognizing human actions
- human object interactions
- bag of features
- human motion
- depth map
- spatio temporal
- action detection
- depth information
- view invariant action recognition
- action primitives
- recognizing actions
- machine learning
- motion estimation
- optical flow