SkeleMotion: A New Representation of Skeleton Joint Sequences based on Motion Information for 3D Action Recognition.
Carlos CaetanoJessica SenaFrançois BrémondJefersson A. dos SantosWilliam Robson SchwartzPublished in: AVSS (2019)
Keyphrases
- action recognition
- recognizing human actions
- recognition of human actions
- view invariant
- action recognition in videos
- bag of features
- human actions
- activity recognition
- human detection
- spatial temporal
- body parts
- bag of words
- motion history images
- computer vision
- human activities
- static images
- motion capture data
- action classification
- video dataset
- independent subspace analysis
- action primitives
- space time interest points
- recognizing actions
- depth sensors
- viewpoint
- human pose
- text classification