Action Recognition for Depth Video using Multi-view Dynamic Images.
Yang XiaoJun ChenZhiguo CaoJoey Tianyi ZhouXiang BaiPublished in: CoRR (2018)
Keyphrases
- multi view
- depth video
- action recognition
- view invariant
- three dimensional
- single view
- ccd camera
- depth map
- multiple views
- image data
- dynamic scenes
- d objects
- image retrieval
- input image
- multiple viewpoints
- image registration
- surface reconstruction
- human actions
- computer vision
- low quality
- test images
- range images
- feature points
- similarity measure
- object recognition
- input data
- video sequences
- semi supervised
- filtering method
- view synthesis
- image matching
- lighting conditions
- pose estimation
- particle filter
- time of flight
- feature space