Efficient action recognition from compressed depth maps.
Jie MiaoXiaoyi JiaReji MathewXiangmin XuDavid TaubmanChunmei QingPublished in: ICIP (2016)
Keyphrases
- action recognition
- depth map
- depth cameras
- multi view
- depth information
- computer vision
- activity recognition
- d scene
- three dimensional
- human actions
- body parts
- stereo matching
- recognizing human actions
- real scenes
- depth data
- image sequences
- object tracking
- low resolution
- input data
- d objects
- high resolution
- viewpoint
- stereo camera
- image retrieval