An enhanced multi-view human action recognition system for virtual training simulator.
Beom KwonJunghwan KimSanghoon LeePublished in: APSIPA (2016)
Keyphrases
- multi view
- view invariant
- human actions
- virtual training
- single view
- action recognition
- virtual environment
- virtual reality
- virtual humans
- multiple views
- human motion
- three dimensional
- e learning
- spatio temporal
- d objects
- depth map
- dynamic scenes
- motion capture
- video sequences
- space time
- activity recognition
- visual features
- semi supervised
- feature extraction
- human activities
- multiple viewpoints
- range images
- surface reconstruction
- computer vision
- multi view learning
- pairwise
- low level