Unsupervised Coherent Video Cartoonization with Perceptual Motion Consistency.
Zhenhuan LiuLiang LiHuajie JiangXin JinDandan TuShuhui WangZheng-Jun ZhaPublished in: CoRR (2022)
Keyphrases
- temporal consistency
- space time
- temporal filtering
- input video
- video sequences
- object motion
- key frames
- motion features
- moving camera
- video representation
- video frames
- spatial and temporal
- visual data
- motion analysis
- static images
- successive frames
- visual cues
- visual perception
- image sequences
- video data
- video footage
- unsupervised manner
- visual input
- low frame rate
- shot change detection
- motion estimation
- motion model
- image frames
- dynamic scenes
- single frame
- motion detection
- surveillance videos
- motion patterns
- dynamic textures
- temporal continuity
- camera motion
- motion trajectories
- video clips
- optical flow
- periodic motion
- perceptual information
- video analysis
- moving objects
- low level
- temporal coherence
- camera movement
- unsupervised learning
- video objects
- visual information
- video content
- motion segmentation
- video surveillance
- layered representation
- wide area motion imagery
- video summarization
- multiview video
- video signals
- reference frame
- video retrieval
- human motion
- spatio temporal
- motion capture data
- multi view
- human visual system