VASE: Object-Centric Appearance and Shape Manipulation of Real Videos.
Elia PeruzzoVidit GoelDejia XuXingqian XuYifan JiangZhangyang WangHumphrey ShiNicu SebePublished in: CoRR (2024)
Keyphrases
- object shape
- appearance features
- appearance information
- target object
- object shapes
- object appearance
- global shape
- appearance cues
- geometric information
- appearance variations
- appearance model
- visual appearance
- real objects
- spatial configuration
- object representation
- video objects
- spatial configurations
- arbitrary viewpoint
- object geometry
- shape reconstruction
- shape constraints
- visual input
- invariant properties
- object motion
- reference object
- object localization
- bounding box
- d objects
- shape variations
- object tracking
- video sequences
- d mesh
- shape representation
- viewing angle
- shape model
- pictorial structure
- motion cues
- spatial arrangement
- moving objects
- video data
- shape descriptors
- object location
- dynamic scenes
- shape features
- object model
- shape prior
- shape analysis
- partial occlusion
- object classes
- video frames
- particle filter
- surface shape
- dynamic textures