Multi-Action Recognition via Stochastic Modelling of Optical Flow and Gradients.
Johanna CarvajalConrad SandersonChris McCoolBrian C. LovellPublished in: MLSDA@PRICAI (2014)
Keyphrases
- action recognition
- optical flow
- motion features
- computer vision
- human actions
- bag of words
- activity recognition
- human detection
- spatial temporal
- bag of features
- view invariant
- motion field
- human activities
- image sequences
- recognizing human actions
- recognition of human actions
- flow field
- body parts
- motion model
- video dataset
- static images
- recognizing actions
- depth sensors
- action classification
- mid level
- motion estimation
- human activity recognition
- action recognition in videos
- action primitives
- low level
- view invariant action recognition
- three dimensional
- machine learning