Mining Mid-Level Features for Action Recognition Based on Effective Skeleton Representation.
Pichao WangWanqing LiPhilip OgunbonaZhimin GaoHanling ZhangPublished in: DICTA (2014)
Keyphrases
- action recognition
- mid level features
- recognizing human actions
- recognition of human actions
- action recognition in videos
- bag of features
- human actions
- bag of words
- action classification
- view invariant
- mid level
- activity recognition
- computer vision
- low level features
- image representation
- high level features
- human activities
- machine learning
- probabilistic model
- high level semantics
- active learning
- image retrieval
- pairwise
- high level