Improving multi-view human action recognition with spatial-temporal pooling and view shifting techniques.
Tuan-Dung LeThi-Oanh NguyenThanh-Hai TranPublished in: SoICT (2017)
Keyphrases
- spatial temporal
- action recognition
- multi view
- multiple views
- human actions
- single view
- multiple viewpoints
- view invariant
- human activities
- bag of words
- activity recognition
- three dimensional
- computer vision
- depth map
- d objects
- body parts
- bag of features
- surface reconstruction
- spatio temporal
- viewpoint
- range images
- video database
- motion capture
- visual hull
- view synthesis
- co occurrence
- learning algorithm
- multimedia
- keywords
- machine learning
- text categorization
- semi supervised