Leveraging Motion Priors in Videos for Improving Human Segmentation.
Yu-Ting ChenWen-Yen ChangHai-Lun LuTingfan WuMin SunPublished in: CoRR (2018)
Keyphrases
- motion capture data
- static images
- motion cues
- motion segmentation
- motion based segmentation
- perceptual information
- object segmentation
- temporal segmentation
- video objects
- independently moving objects
- image sequences
- periodic motion
- human motion
- image segmentation
- rigidly moving
- video sequences
- input video
- temporal coherence
- video stabilization
- motion features
- multibody
- dynamic scenes
- moving camera
- segmentation algorithm
- video scene
- camera motion
- parametric models
- video frames
- segmentation method
- motion model
- key frames
- motion capture
- moving observer
- optical flow estimation
- space time
- human activities
- motion history images
- level set
- motion estimation
- motion analysis
- recognition of human actions
- moving objects
- action recognition
- medical images
- optical flow
- human arm
- visual motion
- successive frames
- prior information
- energy function
- spatial coherence
- video content
- motion field
- shape prior
- image motion
- motion parameters
- prior model
- spatio temporal
- humanoid robot
- traffic scenes
- labeling problems
- dynamic textures
- video analysis