Human action recognition using shape and CLG-motion flow from multi-view image sequences.
Mohiuddin AhmadSeong-Whan LeePublished in: Pattern Recognit. (2008)
Keyphrases
- multi view
- image sequences
- flow field
- three dimensional
- motion analysis
- dynamic scenes
- optical flow
- single view
- motion estimation
- rigid objects
- multiple views
- motion field
- scene flow estimation
- motion parameters
- motion capture
- motion model
- depth map
- moving camera
- motion segmentation
- visual hull
- rigid motion
- image motion
- velocity field
- moving objects
- human motion
- photometric stereo
- video sequences
- d objects
- camera motion
- ego motion
- robust tracking
- object tracking
- optical flow estimation
- shape model
- optic flow
- semi supervised
- feature points
- camera calibration
- motion patterns
- multiple cameras
- computer vision
- articulated objects
- structure from motion
- multi frame
- range images
- light field
- video objects
- space time
- multi view learning
- background subtraction
- vector field
- smoothness constraint
- scene structure
- shape parameters
- camera parameters