Adaptive multi-view graph convolutional networks for skeleton-based action recognition.
Xing LiuYanshan LiRongjie XiaPublished in: Neurocomputing (2021)
Keyphrases
- multi view
- action recognition
- view invariant
- single view
- multiple views
- human actions
- depth map
- three dimensional
- computer vision
- camera calibration
- activity recognition
- d objects
- view synthesis
- multiple cameras
- surface reconstruction
- body parts
- range images
- multiple viewpoints
- human activities
- motion capture
- semi supervised
- spatio temporal
- principal component analysis
- training set
- visual hull